research paper in sports injury

• Open access
• Published: 27 May 2024

Principles of musculoskeletal sport injuries for epidemiologists: a review

• Chinchin Wang   ORCID: orcid.org/0000-0002-8472-9936 1 , 2 ,
• Steven D. Stovitz 3 ,
• Jay S. Kaufman 1 ,
• Russell J. Steele 4 &
• Ian Shrier 2 , 5  

Injury Epidemiology volume  11 , Article number:  21 ( 2024 ) Cite this article

181 Accesses

1 Altmetric

Metrics details

Musculoskeletal injuries are a common occurrence in sport. The goal of sport injury epidemiology is to study these injuries at a population level to inform their prevention and treatment.

This review provides an overview of musculoskeletal sport injuries and the musculoskeletal system from a biological and epidemiologic perspective, including injury mechanism, categorizations and types of sport injuries, healing, and subsequent injuries. It is meant to provide a concise introductory substantive background of musculoskeletal sport injuries for epidemiologists who may not have formal training in the underlying anatomy and pathophysiology.

An understanding of sport injuries is important for researchers in sport injury epidemiology when determining how to best define and assess their research questions and measures.

Sport and physical activity are crucial to maintaining a healthy lifestyle for people of all ages. Their wide-ranging benefits include prevention of chronic diseases, reduced morbidity and mortality, and improved mental health (Warburton et al. 2006 ). However, participation in sports and physical activity can also result in injury. Referred to as “sport injuries”, they occur most commonly to the musculoskeletal system that allows the human body to move. Sport injuries can result in morbidity, predispose to further injuries, and decrease subsequent activity due to time lost during recovery or reduced desire to be active (Emery and Pasanen 2019 ). By studying the epidemiology of sport injuries, we can inform their prevention and management at a population level.

Epidemiologists conducting research in the sport injury field may not have formal training in the anatomy and pathophysiology of sport injuries and the musculoskeletal system. This review aims to provide a concise introductory overview of musculoskeletal sport injuries for epidemiologists, covering (1) definition of a sport injury from biological and epidemiological perspectives; (2) common categorizations of sport injuries and subsequent injuries; (3) a summary of the musculoskeletal systems and the injuries occurring to specific tissues and organs; (4) principles of healing and rehabilitation of sport injuries.

What is a sport injury?

An injury is generally considered a sport injury when it occurs in relation to participation in sport or physical activity. Sport injuries occur most commonly to the musculoskeletal system (Patel and Baker 2006 ) and as such, we focus on musculoskeletal sport injuries in this review. However, sport participation can also result in injuries to other organ systems such as the neurological system (e.g. concussions, spinal cord injuries and peripheral nerve injuries), cardiovascular system (e.g. arrhythmias), and other systems (Toth et al. 2005 ; Maron and Pelliccia 2006 ). In this section, we provide definitions of injury from both the biological and epidemiological perspectives.

Biological perspective

Biologically, injuries are broadly defined as tissue damage (Bartlett and Bussey 2013 ). When an individual performs activity, their body is exposed to various forces. For simplicity, we will refer to the sum of these forces as load (Bartlett and Bussey 2013 ). Load causes tissues to undergo deformation. Upon deformation, tissue cells try to keep or restore themselves to their original state, causing an internal resistance known as stress (Bartlett and Bussey 2013 ; Gómez-González et al. 2020 ; Bahr et al. 2012a ). Loads beyond a tissue’s load capacity (or load tolerance) will cause excessive stress leading to tissue damage (Fig.  1 A) (Bartlett and Bussey 2013 ; Bahr et al. 2012a ).

Relationship between a musculoskeletal tissue’s load capacity, load, and injury. The red coloured bars represent load on a tissue, while the blue coloured bars represent the tissue’s load capacity. A Tissues exposed to loads lower than their load capacity experience microdamage (panel a ), while tissues exposed to loads greater than their load capacity experience immediate tissue damage and injury (panel b ). B Given sufficient recovery, microdamage from loads below a tissue’s capacity (panel a ) will result in strengthening and increased load capacity (panel b ). C Tissues without sufficient recovery (panel a ) have a decreased load capacity (panel b ) which can lead to immediate tissue damage and injury even with normal loads (panel c )

Tissue damage can occur to varying extents. Large loads beyond a tissue’s load capacity will cause immediate tissue damage. The amount of damage can usually be visualized without a microscope, and may result in physical symptoms (e.g. pain) and limitations of tissue function (Bartlett and Bussey 2013 ; Bahr et al. 2012a ). Healing involves structural changes, where damaged tissue may be repaired or replaced with scar tissue (Baoge et al. 2012 ). Smaller loads that are close to or minimally above a tissue’s load capacity can cause microdamage or microtrauma, referring to microscopic tears or cracks within a cellular membrane that can be viewed using traditional light microscopy (Bartlett and Bussey 2013 ; Bahr et al. 2012a ; Radasch 1999 ; Stauber et al. 2020 ; Lieber 2018 ). While generally considered an injury in a biological sense, when microdamage is isolated, tissues are usually able to restore tissue integrity without scar tissue, and without most of the physical symptoms of injury (Bahr et al. 2012a ). Given sufficient recovery time, structural adaptations in response to microdamage lead to increased tissue strength and load capacity (Fig.  1 B) (Wiesinger et al. 2015 ; Bahr et al. 2012b ). However, repeated microdamage without sufficient recovery can result in injury, limiting tissue function and resulting in physical symptoms and scar tissue formation (Fig.  1 C) (Bartlett and Bussey 2013 ; Bahr et al. 2012a ; Baoge et al. 2012 ).

Damage can also be restricted to the internal structure of a tissue cell, which might be only visible using electron microscopy (Lieber 2018 ). This includes damage at the cell cytoskeleton level (Butterfield 2010 ; Gefen and Weihs 2016 ). The cytoskeleton is responsible for cellular structure and stability, and plays crucial roles in cell movement, division, and intracellular transport (Gefen and Weihs 2016 ). Damage to the cytoskeleton might impair transport of important molecules, leading to decreased cell function (Gefen and Weihs 2016 ). This damage may or may not be considered an injury. For instance, concussion injuries usually involve a decrease in neuronal brain function without visible microscopic damage (Signoretti et al. 2011 ). While the underlying mechanisms for the decreased function are not fully understood, this is consistent with the hypothesis of cell cytoskeleton damage (Park et al. 2022 ).

Finally, fatigue is the loss of tissue strength upon repeated loading and associated microdamage. While not considered an injury in itself, it results in decreased load capacity, which can make a tissue more susceptible to injury (Bartlett and Bussey 2013 ). Muscle fatigue specifically refers to a reduction in the maximal force or ability to maintain maximal force of the muscles due to repeated use, and is reversible with rest (Enoka and Duchateau 2008 ). It occurs due to impairments in the contractile proteins and structures that allow muscles to generate force, which can be considered as damage at the cell cytoskeleton level (Westerblad et al. 1993 , 1990 ).

Epidemiologic perspective

The International Olympic Committee consensus definition of injury for surveillance and epidemiologic studies is “tissue damage or other derangement of normal physical function due to participation in sports, resulting from rapid or repetitive transfer of kinetic energy” (Bahr et al. 2020 ). This is distinct from their definition of illness, which is a complaint or disorder where the primary mode does not involve the transfer of kinetic energy (Bahr et al. 2020 ). Illnesses include disorders resulting from the loss of vital elements (e.g. dehydration) (Bahr et al. 2020 ), or from the external environment (e.g. heat stroke) during sport, among others.

Injuries often present as pain or other physical symptoms such as aches, soreness, stiffness, or deformities that affect normal physical function (Hainline et al. 2017 ; Clarsen et al. 2013 ; Bolling et al. 2019 ). As such, injuries are sometimes recorded as any patient-reported symptom or complaint of the musculoskeletal system due to physical activity (Bahr et al. 2020 ; Clarsen et al. 2020 ). However, the perception of pain can differ between individuals based on age, sex, and level of activity (Hainline et al. 2017 ; Tesarz et al. 2012 ; Racine et al. 2012 ). Further, injuries can occur without pain (e.g., microdamage), and pain can be present independent of tissue damage (Hainline et al. 2017 ). Therefore, the same underlying biological damage in different individuals may not be similarly defined as an injury.

Some athletes and researchers use a more restricted definition of injury compared to the consensus definition. Many athletes (particularly elite) consider pain and other physical symptoms as a normal part of sport participation (Bolling et al. 2019 ; Verhagen et al. 2021 ; Wiese-bjornstal et al. 1998 ). They perceive an injury as a condition that must preclude them from performing at their normal or optimal level, beyond the experience of pain, which results in altered or missed participation from sport (Bolling et al. 2019 ; Verhagen et al. 2021 ).

Common definitions of injury in epidemiologic studies are “any complaint”, encompassing all injuries whether the complaint is symptom- or performance-based; “medical attention injuries”, or injuries where medical attention was sought; and “time-loss injuries”, or injuries causing the athlete to be unable to complete a current or future activity session (Bahr et al. 2020 ). Medical attention injuries may or may not limit an athlete’s ability to participate in sport, and not all time-loss injuries may be reported, nor require or result in medical attention (Bahr et al. 2020 ). The optimal definition for a sport injury depends on the research question of interest. For example, researchers may be interested in tissue damage as evidenced by microscopy, a medical diagnosis, a patient-reported complaint, or time lost from sport participation.

Although diagnosing injuries is beyond the scope of this article, a positive diagnostic test for an injury could occur due to (1) an actual injury; (2) a false positive test (i.e. test illustrates abnormal morphology when it is normal); or (3) a misunderstanding of what is “abnormal” (i.e. test accurately illustrates abnormal morphology which is not an actual injury). For example, over 30% of individuals over 50 years who have not had back pain (i.e. uninjured) will have disc herniation on a magnetic resonance image (Deyo and Mirza 2016 ). Epidemiologists must avoid conflating abnormal tests with clinical diagnoses, especially in small, individual clinic-based studies.

Categorizations of musculoskeletal sport injuries

Musculoskeletal sport injuries are often categorized by characteristics such as their mode of onset, severity, and anatomical and tissue location in research and surveillance. These categorizations can be used to study the prevention or occurrence of specific groups of injuries. Common categorizations are summarized in Table  1 and discussed further below, and in “ The musculoskeletal system ” section.

Mode of onset

A common categorization of sport injuries is by their mode of onset or mechanism of injury. Injuries are often categorized as acute versus overuse, although these definitions are not always consistent. Acute injuries have a sudden onset related to a specific event (Bahr et al. 2020 ). Biologically, this occurs when an undamaged tissue is subjected to a sudden load beyond its load capacity (McBain et al. 2012 ). One such injury is a broken bone (i.e. fracture) occurring upon a fall. Overuse injuries, on the other hand, occur gradually due to repetitive loading and associated microdamage to a tissue, without a specific inciting event (Bahr et al. 2020 ). For example, an athlete might develop considerable pain in the calf muscle (i.e. strain) occurring gradually with repeated physical activity that impedes further activity. The term “overuse” implies that these injuries occur due to excessive activity beyond what the tissues are prepared for from previous activity and loading. Instead, we could consider that these injuries occur from previous “underuse” of the tissues, whereby the individual has not been sufficiently active previously and has not developed the load capacity within their tissues to handle this level of activity (Stovitz and Johnson 2006 ).

Despite these distinctions, there are grey zones and limitations to this categorization. Some injuries with specific inciting events may have been due to underlying “overuse” (i.e. repeated loading and microdamage without sufficient recovery) rather than a sudden load. An injury may occur if the damaged tissue is subjected to a load that would normally be tolerated if not for the microdamage and reduced load capacity (Fig.  1 C) (Bahr et al. 2020 ). While these injuries occur acutely, their underlying mechanism is consistent with that of overuse.

Another closely-related categorization that is sometimes used in the literature is traumatic (similar to and often used as a synonym for acute) or atraumatic (a.k.a. non-traumatic, similar to overuse) (Merkel and Molony 2012 ). Like acute injuries, traumatic injuries have a sudden onset related to a specific event. Atraumatic injuries occur without a specific inciting event. Injuries occurring alongside a specific event but that are due to underlying microdamage are considered traumatic under this categorization. For instance, a bone can be weakened due to repetitive activity and associated microdamage without any symptoms. The bone might fracture due to a sudden force, such as pushing off during a sprint race. This might be considered an acute injury because it occurred alongside a specific inciting event, but atraumatic because there was no direct trauma.

Injury severity

Injuries can also be described by their severity. Severity may have different meanings for different athletes (e.g., the need for medication or surgery) (Mountjoy et al. 2015 ), but is often defined as the duration of injury, particularly duration of time lost from participation (Bahr et al. 2020 ). Time loss is assessed from the date that an athlete begins to have altered or missed participation from sport (which may not be the date of injury onset from a biological standpoint), until the date that they are fully able to participate (Bahr et al. 2020 ). Time loss can underestimate severity if an athlete returns to sport before the injury is resolved, or overestimate severity if an athlete does not resume their normal participation upon healing (e.g., because their fitness and performance is not at a competitive level) (Bahr et al. 2020 ).

A similar measure to injury duration is the duration of symptoms. While defined as the amount of time that symptoms are present, it is sometimes dichotomized as acute versus chronic. Under this categorization, acute injuries refer to recent injuries, while chronic injuries refer to injuries where symptoms have been present for an extended period of time. For example, some categorize back pain as chronic when symptoms have been present longer than 12 weeks, and as acute when symptoms have been present for less than 12 weeks (Flint et al. 2014 ; Chou et al. 2014 ). Some authors use chronic as a synonym for overuse when describing mode of onset (Lavallee and Balam 2010 ); however, these are separate concepts (Bahr et al. 2020 ). Symptom duration may be longer than time loss duration if an athlete returns to full participation with lingering symptoms, or shorter if an athlete does not resume their normal participation at the time their symptoms cease.

Injury severity can also be described by the severity of symptoms, functional consequences, or a composite score (i.e. patient-reported outcome measure), as self-reported by athletes (Westerblad et al. 1993 ; Clarsen et al. 2013 ; Chou et al. 2014 ) or assessed by a clinician. Severity can also be assessed through functional tests. These measures are often assessed using scales or scoring systems (Bahr et al. 2020 ; Clarsen et al. 2020 ; Stevenson et al. 2001 ).

Finally, strain (muscle) and sprain (ligament) injuries are often graded as 1st, 2nd, or 3rd degree based on amount of tissue damage and/or clinical symptoms, with 1st degree being the least severe and 3rd degree being the most severe (Bartlett and Bussey 2013 ; Järvinen et al. 2000 ; Noonan and Garrett 1999 ). However, we note that certain injuries are graded based on factors other than severity of symptoms or tissue damage. These categorizations are reviewed in “ The musculoskeletal system ” section.

Injury anatomical location

Injuries are often grouped by their anatomical location for summary purposes. These groups may be broad (e.g., upper extremity, lower extremity, trunk) or specific (e.g., knee, lower leg, ankle) depending on the context (Habelt et al. 2011 ). While reporting specific anatomical locations is recommended for injury surveillance programs (Bahr et al. 2020 ), different injuries to the same location can have very different clinical presentations and outcomes. Although specific locations are important for clinicians who are recommending treatment, epidemiologists often focus on broader categories. These broader categories result in larger sample sizes with potentially greater generalizability, although they have limitations when considering specific injuries (Bahr et al. 2020 ; Mountjoy et al. 2015 ).

The musculoskeletal system and its injuries by tissue type

The musculoskeletal system is made up of distinct tissues that function together to provide shape, stability, and movement to the human body. It plays a central role in the ability for humans to do sport and physical activity (Wind et al. 2005 ).

The major connective tissues of the musculoskeletal system are bone, cartilage, skeletal muscle, tendons, and ligaments. Bones are the structural basis for the human skeleton (Klenerman 2015 ). The attachment points of adjacent bones are known as joints. Joints allow for movement of bones with respect to each other (Klenerman 2015 ). Joints contain cartilage, and are surrounded by ligaments, which are tissue bands that physically attach bones to other bones (Frank 2004 ). Muscles pass over joints, and are attached to two different bones by tissue bands called tendons (Klenerman 2015 ). Muscles can be thought of as elastics that shorten (i.e., when contracting) and lengthen (i.e., when relaxing). When a muscle shortens, their attachment sites to each bone come closer together. In general, this leads to movement at the joint (Klenerman 2015 ).

The following section expands on each location and tissue, and summarizes the injuries that can occur to them.

Bones make up the human skeleton, which provides structure and support for the body (Klenerman 2015 ). They provide a rigid attachment point for muscles, which allow bones to move (Klenerman 2015 ). Bones also protect the internal organs, produce blood cells, and store and release minerals and fats (Klenerman 2015 ; Florencio-Silva et al. 2015 ).

Bone composition

Bones are comprised of two types of bone tissue: cortical, and cancellous. Cortical bone is dense and forms a protective outer layer around cancellous bone, which is spongey and responsible for absorbing the load transmitted to bones (Bartlett and Bussey 2013 ; Radasch 1999 ; Klenerman 2015 ). The thickness of cortical bone and the relative distribution of cortical and cancellous bone tissue differs between different locations within a bone and between bones (Radasch 1999 ). Bones are surrounded by an outer fibrous sheath called the periosteum. The periosteum contains nerves and blood vessels, and plays a role in bone remodelling both regularly and after injury (Klenerman 2015 ; Maia Ferreira Alencar et al. 2020 ).

In the long bone of the limbs, the end of a bone is termed the epiphysis, and primarily consists of cancellous bone (Klenerman 2015 ). The main shaft of a bone is the diaphysis, and consists primarily of cortical bone (Klenerman 2015 ). The metaphysis lies between the diaphysis and epiphyses, and consists primarily of cancellous bone (Klenerman 2015 ). Between the metaphysis and the epiphyses is the epiphyseal plate (growth plate) in children, which is a region of bone growth in long bones such as the tibia (shin bone) or femur (thigh bone) (Jorge et al. 2021 ). While the epiphyseal plate is composed of cartilage during childhood and adolescence, it calcifies into bone tissue after growth has completed (Radasch 1999 ; Klenerman 2015 ). Damage to the epiphyseal plate can result in the slowing or stopping of growth for the affected bone, which can lead to angular deformities or asymmetry in lengths of the lower limbs (Caine et al. 2006 ; Jones et al. 2017 ).

Bone injuries

Bones respond to load by thickening in mass and strengthening (Klenerman 2015 ). However, forces beyond the load capacity may result in injury to the bone, called a fracture (Radasch 1999 ). Among athletes, fractures generally occur after a sudden traumatic event such as a fall, contact with another athlete, or contact with an object (e.g. ball travelling with speed) (Court-Brown et al. 2008 ).

Stress fractures are distinct from regular fractures in that they do not arise from a single traumatic event, but rather repeated exposure to load (i.e., overuse injury) (Bahr et al. 2012a ; Saunier and Chapurlat 2018 ). Bones are normally in a state of remodelling, where bone tissue breaks down (resorption) and is replaced by new tissue (Saunier and Chapurlat 2018 ; Hughes et al. 2017 ). Remodelling increases in response to load and associated microdamage (Saunier and Chapurlat 2018 ; Hughes et al. 2017 ). Without adequate recovery time, more bone tissue is resorbed than deposited, creating stress fractures that appear as small cracks (Saunier and Chapurlat 2018 ; Hughes et al. 2017 ). While most stress fractures will heal with adequate rest and rehabilitation (Milner 2008 ), some may require surgery (Shindle et al. 2012 ). Continued loading onto a stress fracture without adequate recovery may result in a complete fracture (Shindle et al. 2012 ).

Two special types of fractures that occur in children are greenstick and buckle fractures. Greenstick fractures only affect one side of the bone, creating a crack that does not extend through the entire bone and causing bending of the bone rather than a full break (Noonan and Price 1998 ). This may be due to increased cartilage content and compliance of young bones compared to adult bones, or because the periosteum sheath surrounding a bone is more elastic in children than in adults, decreasing the likelihood of complete fractures (Maia Ferreira Alencar et al. 2020 ). Greenstick fractures occur most commonly after a fall (Noonan and Price 1998 ). Buckle fractures, also referred to as torus fractures, occur due to compression of cortical bone that creates a bulge on one side of the bone, without a full break (Sharp and Edwards 2019 ). These fractures are common among children and are generally simple to treat through immobilization (Sharp and Edwards 2019 ).

Muscles are the core component of the human muscular system, which is responsible for generating force and movement (Frontera and Ochala 2015 ). While there are three types of muscle (smooth, cardiac, and skeletal), only skeletal muscles are responsible for movement of the human body (Frontera and Ochala 2015 ).

Muscle composition

Skeletal muscle tissue is comprised of individual muscle fibres (Klenerman 2015 ). The patterns in which the muscle fibres are organized help determine the strength and velocity of the muscular contraction (Bahr et al. 2012a ; Klenerman 2015 ). Muscle is the tissue that is the most capable of strengthening in response to load (Bahr et al. 2012a ). This occurs primarily through hypertrophy, in which muscle fibres increase their cross-sectional area (Bahr et al. 2012a ). Skeletal muscle tissue also contains structures that convey information about muscle tension and position (Macefield and Knellwolf 2018 ).

Muscle injuries

There are several types of muscle injuries. Muscle tissue will become damaged when the applied load exceeds its load capacity, either through a single event or repetitive microdamage. These injuries are referred to as strains (or pulled muscles by the general public) when caused by stretching or contraction forces (Bahr et al. 2012a ). Strains are the most common sport injury (Bahr et al. 2012a ; Järvinen et al. 2000 ). The amount of damage can range from tearing of a few muscle fibres with minimal loss of strength (1st degree) to visible partial tearing of the muscle tissue (2nd degree) to a complete tear/rupture of the muscle (3rd degree) (Järvinen et al. 2000 ; Noonan and Garrett 1999 ). Damage to muscle tissue may also damage the structures within the tissue responsible for sensing tension and position through the same mechanisms. For instance, a stretching of the muscle will also stretch these structures. As such, balance and position sense are often disrupted in muscle strains (Kröger 2018 ).

Muscles can also be injured by compressive forces that exceed the load capacity, typically through a direct blow to the muscle. These injuries are referred to as contusions (Bahr et al. 2012a ; Beiner and Jokl 2001 ). Contusions are often associated with rupture of blood vessels, causing internal bleeding that results in a bruise (Bahr et al. 2012a ; Beiner and Jokl 2001 ). Internal bleeding can lead to various clinical consequences, including some conditions that result in permanent disability (Bahr et al. 2012a ; Beiner and Jokl 2001 ).

One particular consequence of unaccustomed activity is delayed onset muscle soreness (DOMS), which is sometimes used as a proxy for injury (Drew and Finch 2016 ). DOMS presents as soreness, stiffness, or pain that follows 1–3 days after unaccustomed activity (Bahr et al. 2012a ; Cheung et al. 2003 ). Although frequently studied in animal models, its mechanism is not entirely understood. It is unclear whether findings from studies on DOMS are generalizable to muscle injuries (Lieber 2018 ).

Tendons represent connective tissue that physically connects muscles to bones (Klenerman 2015 ). They enable the transmission of force from muscle to bone, and help stabilize joints (Bahr et al. 2012a ; Gracey et al. 2020 ).

Tendon composition

Tendons are composed of dense collagen fibres aligned in the same direction as the muscle fibres (Bahr et al. 2012a ; Gracey et al. 2020 ). Tendons contain few elastic fibres (Bahr et al. 2012a ), which causes them to experience only a small change in length for a large amount of force compared to muscle (Gracey et al. 2020 ). The muscle–tendon interface contains receptors that sense and transmit information about forces within the tendon (Maas et al. 2022 ).

Tendon injuries

Tendons will stiffen when lengthened in response to increasing load but become damaged under excessive load (Gracey et al. 2020 ). Sudden injuries to the tendons are known as tears or ruptures. These can range in severity from partial tearing/rupture of the tissue to complete tearing/rupture (Bahr et al. 2012a ; Gracey et al. 2020 ).

Tendons can also experience overuse injuries (Bahr et al. 2012a ). Repetitive loading and microdamage to tendons result in chronic pain, known as tendinopathy (Bahr et al. 2012a ). Tendinopathies include tendinosis in cases with tissue degeneration, and tendinitis in cases with tendon inflammation (Bartlett and Bussey 2013 ; Bahr et al. 2012a ; Bass 2012 ).

Most musculoskeletal injuries occur at the junction where the muscle joins the tendon (Gimigliano et al. 2021 ). Although the tendon may be involved, these are generally considered muscle injuries. In general, tendon injuries that are not close to the muscle–tendon junction often have poor blood supply, resulting in longer recovery times for tendon-specific injuries compared to muscle injuries (Sharma and Maffulli 2006 ; Ahmed et al. 1998 ).

A joint is the point where two or more bones connect. Joints may provide stability or support movement depending on their type and composition (Klenerman 2015 ).

Joint composition

There are three types of joints that differ in composition and function: fibrous, cartilaginous, and synovial joints.

Fibrous joints are fixed, generally immobile joints comprised of dense collagen rather than cartilage, and are found in the skull among other locations (Miroshnichenko et al. 2018 ). Cartilaginous joints are joined by fibrocartilage or hyaline cartilage (see “ Cartilage ” section) and are slightly mobile or immobile (Juneja et al. 2022 ). The epiphyseal plate in long bones, which connects the diaphysis and epiphyses in childhood and adolescence, is considered a cartilaginous joint (Juneja et al. 2022 ). They are found where the right pelvis joins the left pelvis, among other locations (Becker et al. 2010 ).

Synovial joints are the most common joints in the human body and include the major joints of the limbs (e.g. knee, elbow, shoulder) (Klenerman 2015 ). They are mobile and are comprised of a joint cavity, consisting of the ends of the bones that are covered by articular cartilage (Pacifici et al. 2005 ; Hui et al. 2012 ). The joint cavity is surrounded by fibrous tissue known as a joint capsule, and lined with a synovial membrane that secretes fluid to keep the joint lubricated (Klenerman 2015 ; Hui et al. 2012 ; Ralphs and Benjamin 1994 ). The joint capsule seals the joint cavity, keeping synovial fluid inside, and provides stability by limiting joint movements and preventing bones from separating (Ralphs and Benjamin 1994 ). As synovial joints allow for movement, they are the most commonly implicated in sport injuries (Bartlett and Bussey 2013 ).

Joints can be categorized by the types of movement they allow. Joints can move in three planes: sagittal (longitudinal), frontal (coronal), and transverse (axial) (Milner 2008 ). Movement in the sagittal plane is seen from the side of the body (e.g. knee flexion and extension), movement in the frontal plane is seen from the front of the body (e.g. hip abduction and adduction), and movement in the transverse plane is seen from above (e.g. hip rotation) (Milner 2008 ). Hinge (uniaxial) joints are a type of synovial joint where most movement occurs in a single plane, and are found in the elbow and knee (Fornalski et al. 2003 ; Hefzy and Grood 1988 ). Biaxial joints often experience movement in two planes, and include the metacarpophalangeal (finger knuckle) joints (Batmanabane and Malathi 1985 ). Multiaxial joints often experience movement in three planes, and include the hip and shoulder joints (Bowman et al. 2010 ; Peat 1986 ).

Joint injuries

Joint dislocations are a common injury that occur when the bones that connect at the joint are displaced, resulting in immediate pain and limited range of motion (Skelley et al. 2014 ; Hodge and Safran 2002 ). Dislocations typically occur through a sudden traumatic force such as a collision or fall, and occur commonly in the shoulder and elbow among athletes (Skelley et al. 2014 ; Hodge and Safran 2002 ). Dislocations can cause damage to the ligaments, cartilage, and bones (Skelley et al. 2014 ). Dislocations are treated by physical manipulation of the joint back into its normal location, followed by a recovery period often involving immobilization to heal tissue damage (Skelley et al. 2014 ; Hodge and Safran 2002 ). Recurrent dislocations in some joints (e.g. shoulder) are common among athletes (Skelley et al. 2014 ).

Subluxations are partial joint dislocations where the connecting bones do not completely separate. Unlike full dislocations, subluxations sometimes spontaneously “relocate” to their original position without physical manipulation (Aronen 1986 ). Joint injuries also include specific injuries to cartilage and ligaments. These injuries are covered in their respective sections.

Cartilage is an important connective tissue mainly present in joints (Klenerman 2015 ). Cartilage is weaker and more flexible than bone. However, it is still weight-bearing and resilient (Klenerman 2015 ). There are three types of cartilage: hyaline, fibrocartilage, and elastic (Klenerman 2015 ). Elastic cartilage, which is present in the ear and larynx, is not considered a component of the musculoskeletal system (Pollard et al. 2017 ).

Hyaline cartilage composition

Hyaline cartilage is the most common cartilage in the human body (Krishnan and Grodzinsky 2018 ). It does not contain any nerves or blood vessels, and is limited in its ability to repair itself following damage (Sophia Fox et al. 2009 ). It is found inside joints covering the ends of adjacent bones, where it is referred to as articular cartilage (Krishnan and Grodzinsky 2018 ).

Articular cartilage is a highly specialized tissue that reduces friction and provides a smooth surface for movement at the joints (Sophia Fox et al. 2009 ). It redistributes pressure across bones to minimize high pressure point loads that could cause bone swelling and injury (Sophia Fox et al. 2009 ). Its nutrition comes from molecules dissolved in the normal joint fluid (synovial fluid). As the joint moves, synovial fluid circulates and distributes nutrients (Sophia Fox et al. 2009 ). When joint movement is restricted, for instance due to injury or casting, cartilage nutrition is impaired (Sophia Fox et al. 2009 ; Wang et al. 2013 ). Immobilizing a joint is one method of creating osteoarthritis in animals (Palmoski and Brandt 1981 ).

Hyaline cartilage injury

Similar to bone, hyaline cartilage can be injured through a single traumatic event (Buckwalter 2003 ). Furthermore, extensive damage to articular cartilage with insufficient repair leads to unequal redistribution of forces within the joint. One possible consequence of articular damage is post-traumatic osteoarthritis, a condition characterized by joint pain, dysfunction, and malalignment (Buckwalter 2003 ). Although articular cartilage itself is not visible on X-rays, insufficient repair may lead to a decreased cartilage thickness, causing the two bones of a joint to appear much closer together than normal on an X-ray. This is called joint space narrowing, and is an important sign for clinically meaningful osteoarthritis (Chan et al. 2008 ).

Fibrocartilage composition

Fibrocartilage is a stronger and denser type of cartilage than hyaline cartilage (Klenerman 2015 ). Unlike hyaline cartilage, it contains nerves and blood vessels, but only at its periphery (Bahr et al. 2012a ). It is typically found in larger joints, and functions to absorb and distribute forces more evenly across bones (Bahr et al. 2012a ). Fibrocartilage tissue is present in the meniscus of the knee, and the labrum in the hip and shoulder (Bahr et al. 2012a ; Klenerman 2015 ). It also forms part of the intervertebral discs that lie between the bones of the lumbar spine (Klenerman 2015 ).

Fibrocartilage injury

Damage to fibrocartilage occurs in meniscal and labral injuries (tears). Acute meniscal tears occur due to trauma to the knee, and can occur in isolation, or alongside injury to ligaments (Lohmander et al. 2007 ). Degenerative meniscal tears are more common with increasing age and increased loading (e.g. weight-bearing activities) (Adams et al. 2021 ; Howell et al. 2014 ). As we age, the meniscus becomes weaker and more susceptible to tears with low loads. Labral tears in the hip and shoulder can also occur from trauma or degeneration from repetitive loading (Keener and Brophy 2009 ; O’Reilly et al. 2022 ; Banerjee and Mclean 2011 ). Hip labral tears are associated with certain types of abnormal hip morphologies (Banerjee and Mclean 2011 ).

Ligaments represent connective tissue that physically connects bones, spanning a joint (Frank 2004 ). They are often just local prominent thickenings of the joint capsule tissue that run from one bone to the other, with a different tissue composition (Ralphs and Benjamin 1994 ). However, some ligaments exist inside or outside of the joint capsule (Ralphs and Benjamin 1994 ). Their primary function is to stabilize joints and prevent excessive movement at the joint (Frank 2004 ; Milner 2008 ). However, ligaments also play an important role in proprioception as they contain nerve endings that convey information about joint position and movement that are necessary to coordinate contractions by different muscles during movement (Bahr et al. 2012a ). Ligaments stretch out in response to low amounts of load, but will resist movement when pulled tight in response to further load, thus preventing further movement of the joint (Frank 2004 ; Milner 2008 ). However, ligaments will tear if stretched too far, causing injury (Frank 2004 ; Gracey et al. 2020 ).

Ligament composition

Ligaments are primarily composed of dense collagen fibres, with only small amounts of elastic fibres (Frank 2004 ; Milner 2008 ). They are generally strong and stiff, with limited elasticity (Milner 2008 ). The degree of stiffness differs by the relative composition of collagen versus elastic fibres and other components (Solomonow 2004 ). The stiffness of a ligament also increases as the ligament is stretched (Frank 2004 ).

Ligament injuries

While ligaments have a relatively high load tolerance, excessive load will cause damage and injury to the ligaments and other joint structures (Frank 2004 ; Gracey et al. 2020 ). Injuries to ligaments are referred to as sprains (Bartlett and Bussey 2013 ). These can range in severity from some tissue damage with minimal symptoms (1st degree), partial tear of the ligament (2nd degree), to complete tearing/rupture and separation of the damaged ends of the ligament (3rd degree). Grade definitions and terminology can differ by injury type. For instance, there are three ligaments that are implicated in a lateral ankle sprain. Some categorizations use 3rd degree ankle sprain to refer to the complete tearing of each ligament, while some consider a 3rd degree ankle sprain to mean all three ligaments are damaged (Wolfe et al. 2001 ; Lacerda et al. 2023 ). Surgery may be recommended for some complete ligament tears but not others (Beynnon et al. 2005 ; Miyamoto et al. 2009 ; Stone et al. 2021 ). Because damage to a ligament may damage local nerves, proprioception, balance and position sense are often disrupted in sprains (Bahr et al. 2012a ).

Nearly all ligament injuries occur due to a single event (Bartlett and Bussey 2013 ). However, ligaments and other joint structures may also experience microdamage when subjected to repetitive loading (Frank 2004 ; Solomonow 2004 ). This microdamage is usually asymptomatic, but may affect joint stability and predispose individuals to other injuries (Bahr et al. 2012a ).

Healing and rehabilitation of injury

Biological and clinical perspective.

Injury healing occurs in three overlapping phases: (1) inflammation; (2) repair or regeneration; and (3) remodelling (Pieters et al. 2021 ). Briefly, inflammation causes damaged cells and tissues to degenerate (Pieters et al. 2021 ; Bigham-Sadegh and Oryan 2015 ). Tissue repair or regeneration replaces damaged cells with new cells. Finally, the repaired or regenerated tissue is remodelled to regain optimal strength and function in a process that can take months to years (Pieters et al. 2021 ; Bigham-Sadegh and Oryan 2015 ). The specific healing process of an injury depends on the type of tissue that is damaged and the degree of damage. Bone heals by a regenerative process, whereby the healed tissue is the same as the original bone tissue (Bahr et al. 2012a ). Other musculoskeletal tissues heal by a regenerative process when there is less severe tissue damage or microdamage. However, severe tissue damage heals by a repair process, where the healed tissue is a scar tissue rather than the original tissue (Bahr et al. 2012a ; Liu et al. 2018 ; Leong et al. 2020 ). While scar tissue is initially weak, its strength increases during the repair and remodelling process until it is close to that of the original uninjured tissue (Järvinen et al. 2000 ). Extensive scar tissue formation due to tissue bleeding and inflammation may result in decreased tissue strength and increased risk of recurrent injuries (discussed below) (Bahr et al. 2012a ).

Tissue strength can be increased during the repair and remodelling process by applying progressive loads that cause microdamage and subsequent adaptation, but remain below the load capacity for injury (Bahr et al. 2012b ; Magnusson and Kjaer 2019 ; Kiviranta et al. 1988 ; Tipton et al. 1975 ; Liphardt et al. 2024 ; Sun 2010 ; Killian et al. 2012 ). Overloading may lead to further tissue damage and disrupt the healing process. For instance, low loads during a recovery period may still be large enough to cause further tissue damage because of the reduced load capacity (Bahr et al. 2012b ). Just as overloading may cause tissue damage, underloading or extensive immobilization may prevent tissues from strengthening and cause tissue atrophy (Magnusson and Kjaer 2019 ; Liphardt et al. 2024 ; Sun 2010 ; Fournier 2015 ).

Clinicians usually prescribe rehabilitation therapy for injuries. Rehabilitation refers to restoring the tissue to its preinjury state, and involves many components that are not always well described in studies (Wade et al. 2010 ). Most injury rehabilitation programs start with reducing pain and preventing excess bleeding and inflammation, reducing scar tissue formation. Additionally, injuries often result in decreased range of motion, strength and proprioception. Therefore, exercises including those for balance, strengthening, and stretching are often prescribed to specifically address these limitations (Fournier 2015 ). Other components of rehabilitation may include electronic modalities (e.g. ultrasound), manual therapy (e.g. massage, mobilizations, manipulations), and prevention education. Finally, patients are recommended to gradually return to participation in sport once they are largely symptom-free and have regained adequate strength to minimize injury recurrence (Fournier 2015 ).

Epidemiological perspective

Injury healing can be assessed under different definitions, which will affect calculated injury durations. Ideally, an injury is considered healed when the athlete is able to return to their previous amount of activity without pain. When detailed data are available, healing date can be determined from clinical records or self-report of symptoms. Alternatively, researchers might use the date of last treatment for injury, under the assumption that treatment is only provided while an injury remains unhealed (Hamilton et al. 2011 ). However, the decision to stop treatment has subjectivity (Hamilton et al. 2011 ). As such, clinical data may not always minimize error or bias in the measurement of healing.

Unfortunately, detailed data are often unavailable and researchers often operationalize the healing date of an injury as the date of full return to participation or play in sport (Bahr et al. 2020 ; Hamilton et al. 2011 ). However, many athletes return to activity while they are still symptomatic, which could result in measurement error if utilizing return to play or even medical clearance date as the healing date. Epidemiologists should recognize that return to participation decisions can vary between athletes, coaches, and clinicians, and may not necessarily reflect biological healing (Bahr et al. 2020 ; Hamilton et al. 2011 ).

Subsequent and recurrent injuries

Initial injuries may predispose to subsequent injuries due to muscle imbalances, deficits in strength and proprioception, or changes in biomechanics (Fulton et al. 2014 ). Subsequent injuries to the same location account for a considerable proportion (10–25%) of all injuries (Hamilton et al. 2011 ).

Researchers must consider how to define and account for subsequent injuries, particularly in the longitudinal follow-up of athletes. Subsequent injuries generally refer to injuries that occur after an initial index injury. Subsequent injuries to a different body part are considered “subsequent new injuries” (Hamilton et al. 2011 ). Subsequent injuries to the same body part but a different tissue type are called “local injuries”. Finally, subsequent injuries to the same body part and tissue type are called “recurrent injuries”.

Recurrent injuries can be exacerbations or re-injuries. An “exacerbation” is a worsening of an index injury that was not fully healed or recovered (Bahr et al. 2020 ; Hamilton et al. 2011 ). A “re-injury” is a recurrent injury that occurs to the same location and tissue as an index injury that was fully healed or recovered (Bahr et al. 2020 ). Recurrences are sometimes further categorized by the time that they occurred following healing of the index injury (early: within 2 months; late: within 2–12 months; or delayed: more than 12 months) (Hamilton et al. 2011 ).

These categorizations have limitations. For instance, it can be unclear whether a subsequent injury is related or not related to the initial injury (Shrier and Steele 2014 ). Further, the definition of healing within an epidemiologic study will affect whether an injury is considered a re-injury or an exacerbation. This will in turn affect the overall injury count, risk, rate, and length of time loss. Consider an individual who suffers an index injury on January 1st and returns to participation on January 10th but continues to experience symptoms and receive medical treatment. They then experience a worsening of symptoms on January 30th. If healing is defined as return to participation, the individual will be considered as having had two injuries (an index injury and a subsequent/recurrent injury), each with a separate length of time loss with the sum being the total time lost. If healing is defined as date of last treatment or by cessation of symptoms, the individual will be considered as having one injury (an index injury and an exacerbation), with a longer time loss (equal to the total time lost). While there is no consensus for the optimal definition of healing, researchers would benefit from clearly defining their outcomes. Further, when synthesizing and interpreting findings from multiple studies, researchers should ensure that aggregated results use similar definitions.

Sport injuries are a concern for anyone participating in physical activity. Applying epidemiologic methods can greatly contribute to determining how to best prevent and treat sport injuries and their related morbidities. Understanding what constitutes a sport injury from a biological and epidemiologic perspective is important for researchers in these fields, who must determine how to best define and assess their research questions and measures.

Availability of data and materials

Not applicable.

Adams BG, Houston MN, Cameron KL. The epidemiology of meniscus injury. Sports Med Arthrosc Rev. 2021;29(3): e24.

Article   PubMed   Google Scholar  

Ahmed IM, Lagopoulos M, McConnell P, Soames RW, Sefton GK. Blood supply of the achilles tendon. J Orthop Res. 1998;16(5):591–6.

Article   CAS   PubMed   Google Scholar  

Aronen JG. Anterior shoulder dislocations in sports. Sports Med. 1986;3(3):224–34.

Bahr R, Alfredson H, Järvinen M, Järvinen T, Khan K, Kjær M, et al. Types and causes of injuries. In: Bahr R, editor., et al., The IOC manual of sports injuries. Hoboken: Wiley; 2012a. p. 1–24. https://doi.org/10.1002/9781118467947.ch1 .

Chapter   Google Scholar  

Bahr R, Cook J, Langberg H, MacAuley D, Matheson G, Mæhlum S. Treating sports injuries. In: Bahr R, editor. The IOC manual of sports injuries. Wiley: Hoboken; 2012b. p. 25–39. https://doi.org/10.1002/9781118467947.ch2 .

Bahr R, Clarsen B, Derman W, Dvorak J, Emery CA, Finch CF, et al. International Olympic Committee consensus statement: methods for recording and reporting of epidemiological data on injury and illness in sport 2020 (including STROBE Extension for Sport Injury and Illness Surveillance (STROBE-SIIS)). Br J Sports Med. 2020. https://doi.org/10.1177/2325967120902908 .

Banerjee P, Mclean CR. Femoroacetabular impingement: a review of diagnosis and management. Curr Rev Musculoskelet Med. 2011;4(1):23–32.

Article   PubMed   PubMed Central   Google Scholar  

Baoge L, Van Den Steen E, Rimbaut S, Philips N, Witvrouw E, Almqvist KF, et al. Treatment of skeletal muscle injury: a review. ISRN Orthop. 2012;2012: 689012.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Bartlett R, Bussey M. Sports biomechanics: reducing injury risk and improving sports performance. Florence: Routledge; 2013.

Google Scholar  

Bass E. Tendinopathy: why the difference between tendinitis and tendinosis matters. Int J Ther Massage Bodyw. 2012;5(1):14–7.

Batmanabane M, Malathi S. Movements at the carpometacarpal and metacarpophalangeal joints of the hand and their effect on the dimensions of the articular ends of the metacarpal bones. Anat Rec. 1985;213(1):102–10.

Becker I, Woodley SJ, Stringer MD. The adult human pubic symphysis: a systematic review. J Anat. 2010;217(5):475–87.

Beiner JM, Jokl P. Muscle contusion injuries: current treatment options. JAAOS J Am Acad Orthop Surg. 2001;9(4):227.

Beynnon BD, Johnson RJ, Abate JA, Fleming BC, Nichols CE. Treatment of anterior cruciate ligament injuries, Part I. Am J Sports Med. 2005;33(10):1579–602.

Bigham-Sadegh A, Oryan A. Basic concepts regarding fracture healing and the current options and future directions in managing bone fractures. Int Wound J. 2015;12(3):238–47.

Bolling C, Barboza SD, van Mechelen W, Pasman HR. How elite athletes, coaches, and physiotherapists perceive a sports injury. Transl Sports Med. 2019;2(1):17–23.

Article   Google Scholar  

Bowman KF, Fox J, Sekiya JK. A clinically relevant review of hip biomechanics. Arthrosc J Arthrosc Relat Surg. 2010;26(8):1118–29.

Buckwalter JA. Sports, joint injury, and posttraumatic osteoarthritis. J Orthop Sports Phys Ther. 2003;33(10):578–88.

Butterfield TA. Eccentric exercise in vivo: strain-induced muscle damage and adaptation in a stable system. Exerc Sport Sci Rev. 2010;38(2):51.

Caine D, DiFiori J, Maffulli N. Physeal injuries in children’s and youth sports: Reasons for concern? Br J Sports Med. 2006;40(9):749–60.

Chan WP, Huang GS, Hsu SM, Chang YC, Ho WP. Radiographic joint space narrowing in osteoarthritis of the knee: relationship to meniscal tears and duration of pain. Skeletal Radiol. 2008;37(10):917–22.

Cheung K, Hume PA, Maxwell L. Delayed onset muscle soreness. Sports Med. 2003;33(2):145–64.

Chou R, Cotton D, Cotton D, Rao JK, Taichman D, Williams S. Low back pain. Ann Intern Med. 2014;160(11):1.

Clarsen B, Myklebust G, Bahr R. Development and validation of a new method for the registration of overuse injuries in sports injury epidemiology: the Oslo Sports Trauma Research Centre (OSTRC) Overuse Injury Questionnaire. Br J Sports Med. 2013;47(8):495–502.

Clarsen B, Bahr R, Myklebust G, Andersson SH, Docking SI, Drew M, et al. Improved reporting of overuse injuries and health problems in sport: an update of the Oslo Sport Trauma Research Center questionnaires. Br J Sports Med. 2020;54(7):390–6.

Court-Brown CM, Wood AM, Aitken S. The epidemiology of acute sports-related fractures in adults. Injury. 2008;39(12):1365–72.

Deyo RA, Mirza SK. Herniated lumbar intervertebral disk. N Engl J Med. 2016;374(18):1763–72.

Drew MK, Finch CF. The relationship between training load and injury, illness and soreness: a systematic and literature review. Sports Med. 2016;46(6):861–83.

Emery CA, Pasanen K. Current trends in sport injury prevention. Best Pract Res Clin Rheumatol. 2019;33(1):3–15.

Enoka RM, Duchateau J. Muscle fatigue: what, why and how it influences muscle function. J Physiol. 2008;586(Pt 1):11–23.

Flint JH, Wade AM, Giuliani J, Rue JP. Defining the terms acute and chronic in orthopaedic sports injuries: a systematic review. Am J Sports Med. 2014;42(1):235–41.

Florencio-Silva R, Sasso GRS, Sasso-Cerri E, Simões MJ, Cerri PS. Biology of bone tissue: structure, function, and factors that influence bone cells. BioMed Res Int. 2015;2015: e421746.

Fornalski S, Gupta R, Lee TQ. Anatomy and biomechanics of the elbow joint. Sports Med Arthrosc Rev. 2003;11(1):1.

Fournier M. Principles of rehabilitation and return to sports following injury. Clin Podiatr Med Surg. 2015;32(2):261–8.

Frank CB. Ligament structure, physiology and function. J Musculoskelet Neuronal Interact. 2004;4(2):199–201.

CAS   PubMed   Google Scholar  

Frontera WR, Ochala J. Skeletal muscle: a brief review of structure and function. Calcif Tissue Int. 2015;96(3):183–95.

Fulton J, Wright K, Kelly M, Zebrosky B, Zanis M, Drvol C, et al. Injury risk is altered by previous injury: a systematic review of the literature and presentation of causative neuromuscular factors. Int J Sports Phys Ther. 2014;9(5):583–95.

PubMed   PubMed Central   Google Scholar  

Gefen A, Weihs D. Cytoskeleton and plasma-membrane damage resulting from exposure to sustained deformations: a review of the mechanobiology of chronic wounds. Med Eng Phys. 2016;38(9):828–33.

Gimigliano F, Resmini G, Moretti A, Aulicino M, Gargiulo F, Gimigliano A, et al. Epidemiology of musculoskeletal injuries in adult athletes: a scoping review. Medicina. 2021;57(10):1118.

Gómez-González M, Latorre E, Arroyo M, Trepat X. Measuring mechanical stress in living tissues. Nat Rev Phys. 2020;2(6):300–17.

Gracey E, Burssens A, Cambré I, Schett G, Lories R, McInnes IB, et al. Tendon and ligament mechanical loading in the pathogenesis of inflammatory arthritis. Nat Rev Rheumatol. 2020;16(4):193–207.

Habelt S, Hasler CC, Steinbrück K, Majewski M. Sport injuries in adolescents. Orthop Rev. 2011;3(2): e18.

Hainline B, Turner JA, Caneiro JP, Stewart M, Moseley GL. Pain in elite athletes—neurophysiological, biomechanical and psychosocial considerations: a narrative review. Br J Sports Med. 2017;51(17):1259–64.

Hamilton G, Meeuwisse W, Emery C, Shrier I. Subsequent injury definition, classification, and consequence. Clin J Sport Med. 2011;21(6):508–14.

Hefzy MS, Grood ES. Review of knee models. Appl Mech Rev. 1988;41(1):1–13.

Hodge DK, Safran MR. Sideline management of common dislocations. Curr Sports Med Rep. 2002;1(3):149.

Howell R, Kumar NS, Patel N, Tom J. Degenerative meniscus: pathogenesis, diagnosis, and treatment options. World J Orthop. 2014;5(5):597–602.

Hughes JM, Popp KL, Yanovich R, Bouxsein ML, Matheny RW. The role of adaptive bone formation in the etiology of stress fracture. Exp Biol Med. 2017;242(9):897–906.

Article   CAS   Google Scholar  

Hui AY, McCarty WJ, Masuda K, Firestein GS, Sah RL. A systems biology approach to synovial joint lubrication in health, injury, and disease. Wires Syst Biol Med. 2012;4(1):15–37.

Järvinen TAH, Kääriäinen M, Järvinen M, Kalimo H. Muscle strain injuries. Curr Opin Rheumatol. 2000;12(2):155.

Jones C, Wolf M, Herman M. Acute and chronic growth plate injuries. Pediatr Rev. 2017;38(3):129–38.

Jorge AAL, Grimberg A, Dattani MT, Baron J. 11—Disorders of childhood growth. In: Sperling MA, editor. Sperling pediatric endocrinology. 5th ed. Philadelphia: Elsevier; 2021. p. 299–356.

Juneja P, Munjal A, Hubbard JB. Anatomy, Joints. In: StatPearls. Treasure Island (FL): StatPearls Publishing; 2022. Available from: http://www.ncbi.nlm.nih.gov/books/NBK507893/

Keener JD, Brophy RH. Superior labral tears of the shoulder: pathogenesis, evaluation, and treatment. JAAOS J Am Acad Orthop Surg. 2009;17(10):627.

Killian ML, Cavinatto L, Galatz LM, Thomopoulos S. The role of mechanobiology in tendon healing. J Shoulder Elbow Surg. 2012;21(2):228–37.

Kiviranta I, Tammi M, Jurvelin J, Säämänen AM, Helminen HJ. Moderate running exercise augments glycosaminoglycans and thickness of articular cartilage in the knee joint of young beagle dogs. J Orthop Res. 1988;6(2):188–95.

Klenerman L. The skeleton and its attachments. In: Klenerman L, editor. Human anatomy: a very short introduction. Oxford: Oxford University Press; 2015. https://doi.org/10.1093/actrade/9780198707370.003.0002 .

Krishnan Y, Grodzinsky AJ. Cartilage diseases. Matrix Biol. 2018;71–72:51–69.

Kröger S. Proprioception 2.0: novel functions for muscle spindles. Curr Opin Neurol. 2018;31(5):592.

Lacerda D, Pacheco D, Rocha AT, Diniz P, Pedro I, Pinto FG. Current concept review: state of acute lateral ankle injury classification systems. J Foot Ankle Surg. 2023;62(1):197–203.

Lavallee ME, Balam T. An overview of strength training injuries: acute and chronic. Curr Sports Med Rep. 2010;9(5):307.

Leong NL, Kator JL, Clemens TL, James A, Enamoto-Iwamoto M, Jiang J. Tendon and ligament healing and current approaches to tendon and ligament regeneration. J Orthop Res. 2020;38(1):7–12.

Lieber RL. Biomechanical response of skeletal muscle to eccentric contractions. J Sport Health Sci. 2018;7(3):294–309.

Liphardt AM, Godonou ET, Dreiner M, Mündermann A, Tascilar K, Djalal N, et al. Immobilization by 21 days of bed rest results in type II collagen degradation in healthy individuals. Osteoarthr Cartil. 2024;32(2):177–86.

Liu J, Saul D, Böker KO, Ernst J, Lehman W, Schilling AF. Current methods for skeletal muscle tissue repair and regeneration. BioMed Res Int. 2018;2018: e1984879.

Lohmander LS, Englund PM, Dahl LL, Roos EM. The long-term consequence of anterior cruciate ligament and meniscus injuries: osteoarthritis. Am J Sports Med. 2007;35(10):1756–69.

Maas H, Noort W, Smilde HA, Vincent JA, Nardelli P, Cope TC. Detection of epimuscular myofascial forces by Golgi tendon organs. Exp Brain Res. 2022;240(1):147–58.

Macefield VG, Knellwolf TP. Functional properties of human muscle spindles. J Neurophysiol. 2018;120(2):452–67.

Magnusson SP, Kjaer M. The impact of loading, unloading, ageing and injury on the human tendon. J Physiol. 2019;597(5):1283–98.

Maia Ferreira Alencar CH, Sampaio Silveira CR, Cavalcante MM, Maia Vieira CG, Diógenes Teixeira MJ, Neto FA, et al. Periosteum: an imaging review. Eur J Radiol Open. 2020;7: 100249.

Maron BJ, Pelliccia A. The heart of trained athletes. Circulation. 2006;114(15):1633–44.

McBain K, Shrier I, Shultz R, Meeuwisse WH, Klügl M, Garza D, et al. Prevention of sports injury I: a systematic review of applied biomechanics and physiology outcomes research. Br J Sports Med. 2012;46(3):169–73.

Merkel DL, Molony JT. Recognition and management of traumatic sports injuries in the skeletally immature athlete. Int J Sports Phys Ther. 2012;7(6):691–704.

Milner C. Functional anatomy for sport and exercise: quick reference. London: Routledge; 2008. p. 160.

Book   Google Scholar  

Miroshnichenko K, Liu L, Tsukrov I, Li Y. Mechanical model of suture joints with fibrous connective layer. J Mech Phys Solids. 2018;111:490–502.

Miyamoto RG, Bosco JA, Sherman OH. Treatment of medial collateral ligament injuries. JAAOS J Am Acad Orthop Surg. 2009;17(3):152.

Mountjoy M, Junge A, Alonso J, Clarsen B, Pluim B, Shrier I, et al. Consensus statement on the methodology of injury and illness surveillance in FINA (aquatic sports): Table 1. Br J Sports Med. 2015;50:590–6.

Noonan TJ, Garrett WEJ. Muscle strain injury: diagnosis and treatment. JAAOS J Am Acad Orthop Surg. 1999;7(4):262.

Noonan KJ, Price CT. Forearm and distal radius fractures in children. JAAOS J Am Acad Orthop Surg. 1998;6(3):146.

O’Reilly OC, Behrens MB, Gulbrandsen TR, Wolf BR. Posterior shoulder instability and labral pathology. In: Werner BC, editor. MRI-arthroscopy correlations: a case-based atlas of the knee, shoulder, elbow, hip and ankle. Cham: Springer; 2022. p. 183–99. https://doi.org/10.1007/978-3-030-94789-7_13 .

Pacifici M, Koyama E, Iwamoto M. Mechanisms of synovial joint and articular cartilage formation: recent advances, but many lingering mysteries. Birth Defects Res Part C Embryo Today Rev. 2005;75(3):237–48.

Palmoski MJ, Brandt KD. Running inhibits the reversal of atrophic changes in canine knee cartilage after removal of a leg cast. Arthritis Rheum. 1981;24(11):1329–37.

Park E, Baker AJ. The pathophysiology of concussion. In: Schweizer TA, Baker AJ, editors. Tackling the concussion epidemic: a bench to bedside approach. Cham: Springer; 2022. p. 25–41. https://doi.org/10.1007/978-3-030-93813-0_2 .

Patel DR, Baker RJ. Musculoskeletal injuries in sports. Prim Care Clin off Pract. 2006;33(2):545–79.

Peat M. Functional anatomy of the shoulder complex. Phys Ther. 1986;66(12):1855–65.

Pieters D, Wezenbeek E, Schuermans J, Witvrouw E. Return to play after a hamstring strain injury: it is time to consider natural healing. Sports Med. 2021;51(10):2067–77.

Pollard TD, Earnshaw WC, Lippincott-Schwartz J, Johnson GT, editors. Chapter 32 - Connective Tissues. In: Cell Biology (Third Edition) [Internet]. Elsevier; 2017, p. 555–70. Available from: https://www.sciencedirect.com/science/article/pii/B9780323341264000323

Racine M, Tousignant-Laflamme Y, Kloda LA, Dion D, Dupuis G, Choinière M. A systematic literature review of 10 years of research on sex/gender and pain perception—part 2: Do biopsychosocial factors alter pain sensitivity differently in women and men? Pain. 2012;153(3):619–35.

Radasch RM. Biomechanics of bone and fractures. Vet Clin N Am Small Anim Pract. 1999;29(5):1045–82.

Ralphs JR, Benjamin M. The joint capsule: structure, composition, ageing and disease. J Anat. 1994;184(Pt 3):503–9.

CAS   PubMed   PubMed Central   Google Scholar  

Saunier J, Chapurlat R. Stress fracture in athletes. Joint Bone Spine. 2018;85(3):307–10.

Sharma P, Maffulli N. Biology of tendon injury: healing, modeling and remodeling. J Musculoskelet Neuronal Interact. 2006;6(2):181–90.

Sharp JW, Edwards RM. Core curriculum illustration: pediatric buckle fracture of the distal radius. Emerg Radiol. 2019;26(4):483–4.

Shindle MK, Endo Y, Warren RF, Lane JM, Helfet DL, Schwartz EN, et al. Stress fractures about the tibia, foot, and ankle. JAAOS J Am Acad Orthop Surg. 2012;20(3):167.

Shrier I, Steele RJ. Classification systems for reinjuries: a continuing challenge. Br J Sports Med. 2014;48(18):1338–9.

Signoretti S, Lazzarino G, Tavazzi B, Vagnozzi R. The pathophysiology of concussion. PM&R. 2011;3(10S2):S359–68.

Skelley NW, McCormick JJ, Smith MV. In-game management of common joint dislocations. Sports Health. 2014;6(3):246–55.

Solomonow M. Ligaments: a source of work-related musculoskeletal disorders. J Electromyogr Kinesiol. 2004;14(1):49–60.

Sophia Fox AJ, Bedi A, Rodeo SA. The basic science of articular cartilage. Sports Health. 2009;1(6):461–8.

Stauber T, Blache U, Snedeker JG. Tendon tissue microdamage and the limits of intrinsic repair. Matrix Biol. 2020;85–86:68–79.

Stevenson M, Segui-Gomez M, Lescohier I, Scala CD, McDonald-Smith G. An overview of the injury severity score and the new injury severity score. Inj Prev. 2001;7(1):10–3.

Stone AV, Marx S, Conley CW. Management of partial tears of the anterior cruciate ligament: a review of the anatomy, diagnosis, and treatment. JAAOS J Am Acad Orthop Surg. 2021;29(2):60.

Stovitz SD, Johnson RJ. “Underuse” as a cause for musculoskeletal injuries: Is it time that we started reframing our message? Br J Sports Med. 2006;40(9):738–9.

Sun HB. Mechanical loading, cartilage degradation, and arthritis. Ann N Y Acad Sci. 2010;1211:37–50.

Tesarz J, Schuster AK, Hartmann M, Gerhardt A, Eich W. Pain perception in athletes compared to normally active controls: a systematic review with meta-analysis. Pain. 2012;153(6):1253–62.

Tipton CM, Matthes RD, Maynard JA, Carey RA. The influence of physical activity on ligaments and tendons. Med Sci Sports. 1975;7(3):165–75.

Toth C, McNeil S, Feasby T. Central nervous system injuries in sport and recreation. Sports Med. 2005;35(8):685–715.

Verhagen E, Warsen M, Bolling CS. ‘I just want to run’: how recreational runners perceive and deal with injuries. BMJ Open Sport Exerc Med. 2021;7(3): e001117.

Wade DT, Smeets RJEM, Verbunt JA. Research in rehabilitation medicine: methodological challenges. J Clin Epidemiol. 2010;63(7):699–704.

Wang Y, Wei L, Zeng L, He D, Wei X. Nutrition and degeneration of articular cartilage. Knee Surg Sports Traumatol Arthrosc. 2013;21(8):1751–62.

Warburton DER, Nicol CW, Bredin SSD. Health benefits of physical activity: the evidence. CMAJ. 2006;174(6):801–9.

Westerblad H, Lee JA, Lamb AG, Bolsover SR, Allen DG. Spatial gradients of intracellular calcium in skeletal muscle during fatigue. Pflugers Arch. 1990;415(6):734–40.

Westerblad H, Duty S, Allen DG. Intracellular calcium concentration during low-frequency fatigue in isolated single fibers of mouse skeletal muscle. J Appl Physiol. 1993;75(1):382–8.

Wiese-bjornstal DM, Smith AM, Shaffer SM, Morrey MA. An integrated model of response to sport injury: psychological and sociological dynamics. J Appl Sport Psychol. 1998;10(1):46–69.

Wiesinger HP, Kösters A, Müller E, Seynnes OR. Effects of increased loading on in vivo tendon properties: a systematic review. Med Sci Sports Exerc. 2015;47(9):1885–95.

Wind H, Gouttebarge V, Kuijer PPFM, Frings-Dresen MHW. Assessment of functional capacity of the musculoskeletal system in the context of work, daily living, and sport: a systematic review. J Occup Rehabil. 2005;15(2):253–72.

Wolfe MW, Uhl TL, Mattacola CG, Mccluskey LC. Management of ankle sprains. Am Fam Physician. 2001;63(1):93–105.

Download references

Acknowledgements

This research did not receive any specific funding.

Author information

Authors and affiliations.

Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, Canada

Chinchin Wang & Jay S. Kaufman

Centre for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, 3755 Côte Ste-Catherine Road, Montreal, QC, H3T 1E2, Canada

Chinchin Wang & Ian Shrier

Department of Family Medicine and Community Health, University of Minnesota, Minneapolis, USA

Steven D. Stovitz

Department of Mathematics and Statistics, McGill University, Montreal, Canada

Russell J. Steele

Department of Family Medicine, McGill University, Montreal, Canada

You can also search for this author in PubMed   Google Scholar

Contributions

CW and IS conceived the manuscript. CW drafted the initial manuscript. IS, JK, RS, and SS critically reviewed and edited the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Ian Shrier .

Ethics declarations

Ethics approval and consent to participate, consent for publication, competing interests.

The authors declare that they have no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Wang, C., Stovitz, S.D., Kaufman, J.S. et al. Principles of musculoskeletal sport injuries for epidemiologists: a review. Inj. Epidemiol. 11 , 21 (2024). https://doi.org/10.1186/s40621-024-00507-3

Download citation

Received : 01 February 2024

Accepted : 14 May 2024

Published : 27 May 2024

DOI : https://doi.org/10.1186/s40621-024-00507-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Sport injuries
• Musculoskeletal system
• Epidemiology
• Strains and sprains

Injury Epidemiology

ISSN: 2197-1714

• Submission enquiries: [email protected]
• General enquiries: [email protected]

Youth sport injury research: a narrative review and the potential of interdisciplinarity

Affiliations.

• 1 Department of Food and Nutrition, and Sport Science, University of Gothenburg, Gothenburg, Sweden.
• 2 Faculty of Education, Monash University, Melbourne, Victoria, Australia.
• 3 School of Health Sciences, Örebro University, Örebro, Sweden.
• PMID: 33489308
• PMCID: PMC7805357
• DOI: 10.1136/bmjsem-2020-000933

To prevent sports injuries, researchers have aimed to understand injury aetiology from both the natural and social sciences and through applying different methodologies. This research has produced strong disciplinary knowledge and a number of injury prevention programmes. Yet, the injury rate continues to be high, especially in youth sport and youth football. A key reason for the continued high injury rate is the development of injury prevention programmes based on monodisciplinary knowledge that does not account for the complex nature of sport injury aetiology. The purpose of this paper is to consider and outline an interdisciplinary research process to research the complex nature of sport injury aetiology. To support our proposition, we first present a narrative review of existing youth football and youth sport injury research demonstrating an absence of paradigmatic integration across the research areas' main disciplines of biomedicine, psychology and sociology. We then demonstrate how interdisciplinary research can address the complexity of youth sport injury aetiology. Finally, we introduce the interdisciplinary process we have recently followed in a youth football injury research project. While further research is necessary, particularly regarding the integration of qualitative and quantitative sport injury data, we propose that the pragmatic interdisciplinary research process can be useful for researchers who aim to work across disciplines and paradigms and aim to employ methodological pluralism in their research.

Keywords: athlete; football; injury; methodological.

© Author(s) (or their employer(s)) 2021. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

Publication types

Log in using your username and password

• Search More Search for this keyword Advanced search
• Latest Content
• For authors
• BMJ Journals More You are viewing from: Google Indexer

You are here

• Volume 7, Issue 1
• Youth sport injury research: a narrative review and the potential of interdisciplinarity
• Article Text
• Article info
• Citation Tools
• Rapid Responses
• Article metrics

This article has a correction. Please see:

• Correction: Youth sport injury research: a narrative review and the potential of interdisciplinarity - May 01, 2022

• http://orcid.org/0000-0002-3762-2976 Solveig Elisabeth Hausken-Sutter 1 ,
• http://orcid.org/0000-0002-1487-8987 Richard Pringle 2 ,
• http://orcid.org/0000-0001-8436-7149 Astrid Schubring 1 ,
• http://orcid.org/0000-0002-9742-527X Stefan Grau 1 ,
• http://orcid.org/0000-0002-3918-7904 Natalie Barker-Ruchti 1 , 3
• 1 Department of Food and Nutrition, and Sport Science , University of Gothenburg , Gothenburg , Sweden
• 2 Faculty of Education , Monash University , Melbourne , Victoria , Australia
• 3 School of Health Sciences , Örebro University , Örebro , Sweden
• Correspondence to Solveig Elisabeth Hausken-Sutter; solveig.sand.hausken{at}gu.se

To prevent sports injuries, researchers have aimed to understand injury aetiology from both the natural and social sciences and through applying different methodologies. This research has produced strong disciplinary knowledge and a number of injury prevention programmes. Yet, the injury rate continues to be high, especially in youth sport and youth football. A key reason for the continued high injury rate is the development of injury prevention programmes based on monodisciplinary knowledge that does not account for the complex nature of sport injury aetiology. The purpose of this paper is to consider and outline an interdisciplinary research process to research the complex nature of sport injury aetiology. To support our proposition, we first present a narrative review of existing youth football and youth sport injury research demonstrating an absence of paradigmatic integration across the research areas’ main disciplines of biomedicine, psychology and sociology. We then demonstrate how interdisciplinary research can address the complexity of youth sport injury aetiology. Finally, we introduce the interdisciplinary process we have recently followed in a youth football injury research project. While further research is necessary, particularly regarding the integration of qualitative and quantitative sport injury data, we propose that the pragmatic interdisciplinary research process can be useful for researchers who aim to work across disciplines and paradigms and aim to employ methodological pluralism in their research.

• methodological

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:  http://creativecommons.org/licenses/by-nc/4.0/ .

https://doi.org/10.1136/bmjsem-2020-000933

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known?

Youth athletes sustain injuries.

Youth sport injuries have been researched from multiple scientific disciplines.

There is growing support for the application of a complexity approach to sport injury research.

What are the new findings?

An overview of youth football and youth sport injury research from the scientific disciplines of biomedicine, psychology and sociology.

Paradigmatic distinctions in youth football and youth sport injury research create scientific differences between bodies of injury research.

Youth football and youth sport injury research lacks integration across disciplines and paradigms.

Our proposed interdisciplinary research process consists of five phases that demonstrates the working process of a project researching youth football injuries.

The five-phase process can be considered a response to the call for interdisciplinarity in sport injury research as well as a practical guide.

Introduction

Youth sport injury research has during the past years produced important knowledge concerning injury aetiology and injury prevention. Most of this scholarship has emerged from research conducted in specific scientific disciplines, primarily biomechanics, sport medicine, exercise physiology, sport psychology and sport sociology. 1 Broadly speaking, researchers of these disciplines follow distinctive assumptions of what an injury is and what research questions, ethical stances, research methods and interpretations and explanations of results are most appropriate to study this phenomenon. 2 3 These constellations of beliefs, values and methodologies are referred to as scientific paradigms . Existing youth sport injury research can be categorised by three paradigms: positivism, postpositivism and interpretivism. The positivist paradigm is closely related to reductionism, and reality and truth are understood as singular and identifiable. Positivist sport injury research has specifically focused on identifying and separating risk factors. 4 Methodologically, objectivity is paramount, which requires researchers to detach themselves from the study object. In analysis, the identified risk factors are used to generalise causality. 3 5 In the postpositivist paradigm, reality and truth are understood alike positivism. However, researchers adopt diverse research methods to establish causality, including through qualitative methodology. 3 In the interpretivist paradigm, reality and truth are understood as multiple and relative. 2 3 Researcher objectivity is not considered possible as it is assumed that truth is constructed between researchers and the researched. The overall aim is to gain a deeper understanding of what an injury means, and how it is experienced and made sense of.

Scientific paradigms guide researchers and research, 2 3 5 and they have shaped research on youth sport injury aetiology. The existing body of knowledge is comprehensive, however, seldom brought together to understand sport injury and sport injury aetiology from the perspectives presently adopted. Thus, an interdisciplinary understanding that is interparadigmatic is missing. Recently, a number of sport injury researchers have critiqued the reductionist methodology that characterises most research on sport injury aetiology and suggested a turn to complexity to account for the multidimensional nature of sport injury aetiology. 4 6–9 At present, however, relatively few researchers adopt a complexity approach and scholarship is only at the conceptual level. Moreover, current discussions have not included ways to integrate the influence socio-cultural context has.

The purpose of this paper is to advance existing youth sport injury aetiology scholarship by considering and outlining an interdisciplinary research process to research the complex nature of youth sport injury aetiology. In our view, an interdisciplinary approach that is interparadigmatic has potential to generate data that can address the complexity of an injury. To support our proposition for interdisciplinarity, the paper aims to: (1) present the results of a narrative review that examined multidisciplinary literature on youth sport injury aetiology; (2) discuss interdisciplinary research to consider how such an approach can address the complexity of youth sport injury aetiology and (3) introduce an interdisciplinary research process that we have adopted in a research project on youth football injury aetiology.

To achieve the above aims, we draw on the research context of the ongoing project ‘Injury free children and adolescents: towards best practice in Swedish football’ (FIT project). 10 The purpose of the FIT project is to provide evidence-based interdisciplinary injury prevention strategies. To achieve this, the project aims to produce a comprehensive picture of injury aetiology in a sample of male and female Swedish football players aged 10–19 years through integrating natural and social science and producing quantitative and qualitative data. The research team includes scholars from biomechanics, sport medicine, sociology and sport coaching. The project started in January 2017 and involved a prospective questionnaire to record incidence and prevalence of injuries over a 5-month period, one-time biomedical testing (clinical examination, isometric strength measurements, running analysis and knee stability), a 5-month analysis of training protocols, observation of training sessions of players and their coaches and interviews with players and coaches. At present, the project is working towards the integration of collected quantitative and qualitative data.

Narrative review of youth football and sport injury research

We conducted a narrative review of literature relevant to youth football injury aetiology to provide a critical synthesis of existing literature and to specifically identify paradigmatic and methodological assumptions, research approaches, research methods and data analysis procedures. The focus on existing literature’s paradigmatic distinctions is an important first step in interdisciplinary work. 3 According to Phoenix et al 3  (p. 220), a ‘step back’ to understand underlying assumptions can advance knowledge and awareness of the respective research field as a way to ‘move beyond debates about right or wrong ways of approaching research’.

The first author has from January 2017 to May 2020 searched for literature on youth football injuries in the databases PubMed, Scopus, PsycINFO and Web of Science. Search terms included various combinations of ‘youth’, ‘soccer’, ‘football’, ‘injury’, ‘aetiology‘ and ‘risk’. Inclusion criteria were set to youth football players aged approximately 10–19 years and injury aetiology, injury prevention and/or injury risk factors. The search revealed a paucity of sociological youth football injury research. Thus, the search was broadened to sociological injury research in different sports and age groups, which identified research in biathlon, figure skating, rhythmic gymnastics, rugby and softball. To identify assumptions underlying paradigmatic distinctions, we predefined three analytic areas: body and injury perspective (state of the art regarding how an injury is explained and defined); paradigmatic assumptions (reality/truth/knowledge) and research approaches (methodology). Based on how these categories were approached in the literature included in the review, as well as discussions in the FIT project team and a presentation at a scientific conference, 11 we placed the literature in the three paradigms dominant for youth sport injury research: positivism, postpositivism and interpretivism. Furthermore, five dominant disciplines of youth sport injury research were identified based on the reviewed literature; biomedicine (biomechanics, sport medicine and exercise physiology), sport psychology and sport sociology. Table 1 summarises key findings of the narrative review.

• View inline

Distinctions of youth football and sport injury aetiology research

Biomedical research

Biomedical research on injuries in youth football has mostly focused on individual injury risk factors relating to kinematics, kinetics and spatiotemporal variables, physical development and amount of training and competitions. Key findings from our literature search show that age, growth and biological maturation contribute to an increased risk of injuries, especially during the year of peak height velocity. 12–14 Possible risk factor for injuries in youth football players during this period include muscle strength, 15 familial disposition, 16 previous injury, 17 physical stress (ie, training and match duration and perceived exertion) 18 and match playing. 12 16 Furthermore, external factors such as playing turf and type of shoes are associated with injuries among youth football players. 19 Findings from the biomedical discipline have resulted in the development of injury prevention interventions applied to youth through programmes such as the Nordic hamstring strengthening exercise, 20 the FIFA 11+ exercise programme 21 and a neuromuscular warm-up programme called ‘Knee control’. 22 These exercise-based programmes include multiple training components, such as agility, balance, mobility, running and strength activities.

In addition to identifying risk factors, biomedical research has contributed with a number of sport-general models on injury causation, including the Sequence of Prevention model, 23 and the Multifactorial Model of Athletic Injury Aetiology, 24 which was later revised and updated to a more dynamic model. 25 McIntosh 26 also proposed a multifactorial model of injury causation but with a biomechanical focus on tissue properties and injury. Furthermore, sport injury research frameworks such as the Translating Research into Injury Prevention Practice framework, 27 and the extension of the Reach; Effectiveness; Adoption; Implementation; Maintenance framework 28 have been developed in order to close the gap between research and implementation of interventions. These models and frameworks have, in general, been used by youth football injury aetiology researchers.

Biomedical youth football injury research reviewed in this paper is situated in the positivist paradigm. Given its assumptions of a singular and identifiable reality and truth, an injury is objectively defined as a specified damage to the physical body. Injury aetiology is often linear and related to identifiable individual physical factors. Methodologically, researchers are required to stay objective and abstain from interacting with research participants. A hypothetic-deductive reasoning leads the research from broad hypotheses to testing using quantitative methods (eg, physical testing, questionnaires), often in isolation and through manipulation of typical risk factors, often in a laboratory setting.

Psychological research

Psychological research on injuries in youth football players has mainly focused on psychosocial dimensions. Main risk factors identified in this research are personality traits (eg, high level of trait anxiety; low level of mistrust; ineffective coping), 29 history of stressors (eg, negative life event, daily hassles), 30 mental and physical fatigue 31 and team climate (eg, lack of support from coach and teammates). 31 32

Findings from the psychological discipline have resulted in prevention and research focusing on stress management techniques and goal setting skills. 33–35 Furthermore, the Model of Stress and Athletic Injury, 36 37 which demonstrates how the magnitude of stress and athletes’ appraisal of the situation may be influenced by the interplay between various psychosocial factors (eg, personality, history of stressors, coping resources), has become one of the most cited research models. In extending this model through the Biopsychosocial Model of Stress Athletic Injury and Health, the authors suggest that behavioural mechanisms associated with stress response (eg, impaired self-care; poor sleep quality) should also be addressed. 38 Moreover, emotions and environmental factors have been included in the Biopsychosocial Sport Injury Risk Profile, 39 as important risk factors related to sport injury.

Psychological youth football injury research reviewed in this paper is mainly situated in the positivist paradigm, with some research fitting into the postpositivist paradigm. Like the positivist paradigm, injury and injury aetiology definition is specific and related to the physical body, but adds the mind. Methodologically, researchers keep interaction with athletes to a minimum, and similar to positivist researchers, tend to follow a hypothetic-deductive reasoning as a method to gain knowledge. Assuming that establishing truth requires diverse sets of data, some researchers test the hypotheses by using qualitative methods, such as questionnaires and in-depth interviews.

Sociological research

Sociological research on youth sport injury has examined injuries from the perspective of athletes and by analysing the sporting culture athletes are immersed in. Key findings from our search show that sociocultural values (eg, sporting success; discipline and striving for perfection), social norms (eg, femininity; masculinity; power; being on time; respecting the coach), 40–43 pressure to play through injury and pain (especially socialisation to accept training/competing with pain; silence around pain and injury), 41 44 45 medical support (eg, lack of medical support; information being withheld) 44 46 and coach-athlete relationships, 42 47 can influence injury aetiology among athletes.

Sociological injury researchers do not use models to conceptualise injury aetiology, but employ theoretical frameworks such as sport ethic (a set of ideas and beliefs that together comprise norms of traditional athleticism), 42 narratives of self, 47 culture of risk, 48 social control (eg, individuals become inscribed and normalised by particular dominant standards) and masculinity theories, 49 in order to understand and explain sport injury aetiology.

Sociological youth sport injury research reviewed in this paper tends to be situated in the interpretivist paradigm. An injury and its aetiology are interpreted by athletes and researchers and can be explained or understood by examining athletes’ sociocultural context. To examine injury aetiology, researchers interact with athletes and the context. Methodologically, sociological researchers sometimes use deductive reasoning as described above, but they also use inductive reasoning, where the process of creating insight develops from empirical data towards a theory (ie, patterns, themes and categories of analysis emerge out of the data). Researchers in the interpretivist paradigm apply qualitative methods such as observation and interviews, and interpretive analysis techniques such as thematic, content, discourse and narrative analysis.

Reflections on distinctions of youth football and youth sport injury research

The narratively reviewed literature on youth football and youth sport injury research shows two paradigmatic distinctions. First, the literature is monoparadigmatic and tends to focus on one subdiscipline of sport science. Most of the literature is biomedical, sport psychological and sport sociological, and often based on either quantitative or qualitative research techniques. The French philosopher and sociologist Edgar Morin 50 argues that by approaching research from single disciplines, complexity becomes invisible. The disintegration of complexity through monodisciplinarity is not due to the discipline in itself, but as Morin 50  (p. 2) emphasises, because ‘of the discipline as it is conceived, non-communicating with the other disciplines, closed to itself’. Monodisciplinarity in research reduces the phenomenon into separate parts, thus making the whole as well as interactions between parts and between parts and the whole, invisible. 7 50 Second, and related to monodisciplinarity, there is no integration of research across the three paradigms.

These limitations can be clearly seen in our narratively reviewed literature included in table 1 . Thus, current youth sport injury research can be seen to shield off—or in other words leave out—possibly important aspects, creating a picture that is not actually representative of the complex phenomenon under study (eg, injury aetiology). 51 For example, biomedical and psychological disciplines, by mainly focusing on athletes’ physical bodies and minds, leave out their interaction with and in the larger sociocultural context the athletes participate in. Similarly, sociological research, by focusing on sociocultural context, often ignores how the physical body and mind interact with and influence athletes.

Researchers operating in these paradigms and disciplines can and do apply different research approaches and methods, thus, paradigms do not confine. Indeed, researchers have adopted approaches to be more inclusive, such as an ‘ecological’, ‘integrated’ or more ‘real-world’ approach to sport injuries. 52–54 However, these approaches either miss to account for multiple and complex interactions between different ‘parts of the whole’, 7 or do not attempt to integrate disciplines across paradigms or use multiple quantitative and qualitative research methods.

The complexity approach to sport injury research

The recognition that sport injury aetiology involves numerous interactions between various variables across multiple dimensions has led some researchers to explore the potential of a complex systems approach for understanding and researching sport injury aetiology. 4 6–9 Although researchers understand and apply complexity to sport injury research differently, three elements characterise current discussions. First, a complexity approach embraces an understanding of context as open systems consisting of components that are actively connected through mostly non-linear relationships. 4 7 8 A non-linear relationship between components in the system implies that A plus B does not necessarily equal C. For example, a weak muscle plus psychological stress does not necessarily result in an injury. Second, these non-linear relationships between the components can induce dramatic new effects giving rise to unexpected structures and events, also known as emergence. 55 56 Emergence can, for example, be certain tendencies, powers or complex phenomena such as a sport injury. Emergence, or the injury and its aetiology is constantly evolving, and not strictly predictable. 55 In other words, emergence is self-organising, meaning a complex system can transform over time, either growing or shrinking. 55 Emergence should be accounted for when researching sport injury aetiology, for example, through longitudinal studies. Finally, and what Newell 55 argues is typically overlooked in the understanding of complex systems, is the importance of ‘local knowledge’, or knowledge of specific parts of the system. An athlete can, for example, be stressed over an upcoming exam at school and is injured right before the exam. The athlete might believe exam stress caused the injury. A closer examination may, however, show that it was the way the coach had changed behaviour towards the athlete that affected the injury aetiology. When applying a complex systems approach to injury research, the focus moves from identifying single risk factors for injuries to recognising patterns of interaction among multilevel components, acknowledging that these components interact in unpredictable ways and may be moderated by a number of individual and contextual factors. 4 6 8

Researchers are currently discussing how the implementation of complex systems thinking can advance sport injury research, specifically through statistical mathematical models and computer-based simulations guided by equations, rules and laws. Complex models 4 and statistical procedures such as the Agent-Based Modelling (ABM) and Systems Dynamics (SD) modelling 9 are recent examples. According to Bekker 6  (p. 80), however, this development represents an “overemphasis on the epistemological question of how multifactorialism is accounted for in research and a corresponding underemphasis on the ontological considerations and assumptions we make about the world”, which reflects a dissonance in how complexity is understood and applied to sport injury research. The current argument for complexity assumes a reductionist positivist/postpositivist view where complexity is understood as emerging from the rule-based interactions of simple agents/elements and explored through agent-based modelling. 56 Consequently, it tends to ignore that there is more to emergence than the product of interactions of agents; athletes are themselves complex systems, and more complex than agents in agent-based simulations. 56 Thus, the current argument for complexity is limited in recognising the importance of understanding athletes’ interactions with and in a certain context; hence, it is ignoring the significance of incorporating sociocultural conditions that researchers from the interpretivist paradigm have identified as relevant. Bekker further argues that injury researchers’ focus should not be on methodological preferences, but rather on ‘understanding system goal behaviour using methodological pluralism’ 6 (p. 81). We agree. A complexity approach requires alternative research methodologies. One such approach, which we have found to have potential in the FIT project to account for the multiple dimensions of injury aetiology, is an interdisciplinary research process that includes scientific disciplines from different paradigms and integrates qualitative and quantitative research methods.

The potential of interdisciplinary sport injury research

Drawing on research by Klein and Newell, 57 we understand interdisciplinarity as a research process aiming to address a topic that is too broad or complex to be dealt with adequately by a single discipline. Furthermore, an interdisciplinary research process integrates disciplinary perspectives through construction of a more comprehensive perspective. 57 Both Newell 55 and sports researchers such as Buekers et al 58 argue that complex systems and phenomena are a necessary condition for interdisciplinary studies. The rationale for this argument lies in the way complex phenomena can be understood and studied. If an injury is to be understood as complex, then this would mean, as argued earlier, that the phenomenon is multidimensional. Seeing it from one single angle, such as from biomedicine or sociology, only, the phenomenon appears different than from another angle. Since the overall pattern of behaviour is non-linear and dynamic, an effective method for modelling such a phenomenon must offer insight into the separate parts as well as the complex pattern produced by their overall interactions. 55 An interdisciplinary research process, which draws insights from relevant disciplines and integrates these insights into a more comprehensive understanding, is thus proposed to have potential to study sport injury aetiology. 1

The benefits of interdisciplinarity have been pointed out in a recent Nature editorial. 59 The editor specifically endorsed the incorporation of social sciences, cautioning that ‘[i]f social, economic and/or cultural factors are not included in the framing of the questions’ then ‘a great deal of creativity can be wasted’. Recently, sports sociologists Pringle and Falcous 60 have also advocated a collaboration between the social and natural sciences. They argue that a collaboration between biomedical researchers and sport sociologists, which they refer to as methodological border crossing, could improve the political impact of research from the sociology of sport. Regarding sport injury research, Burwitz et al 1 argued for the benefits of interdisciplinarity over two decades ago. Despite their call, however, very little empirical work has been conducted to this end. Reasons provided for this reservation have been related to various obstacles that prevent high quality, truly integrated interdisciplinary work, as well as practical difficulties relating to gaining funding and publishing. 1 58 Philosophical obstacles relating to paradigmatic differences may also result in a lack of shared understandings, research approaches and methods, 61–63 preventing researchers from adopting such an approach. Nevertheless, our research in the FIT project has demonstrated that interdisciplinarity has potential for sport injury research.

The interdisciplinary research process of the FIT project

The interdisciplinary research process of the FIT project consisted of five phases depicted in figure 1 . The five-phase research process was developed based on key interdisciplinary research steps outlined by interdisciplinary researchers such as Alan Repko, Rick Szostak and William H. Newell. 55 64 While presented sequentially, the five-phase research process is an iterative and non-linear process. At times it was necessary to move back to an earlier phase or forward towards later phases, which is common and reflects the complexity of the interdisciplinary research process. 55 64

• Download figure
• Open in new tab
• Download powerpoint

The interdisciplinary research process of the FIT project. The figure illustrates the five-phase interdisciplinary research process adopted for the study of youth football injury aetiology. 10 The blue boxes present the main theme of each of the five phases of the process. The green boxes entail research actions. The interdisciplinary research process was not linear, but involved the researchers moving back to an earlier phase or forward towards later phases during the process.

In the first phase of the process, the FIT project research group came together from multiple scientific disciplines and agreed on researching injury aetiology in youth football using an interdisciplinary approach. In the second phase, the project team members decided which scientific disciplines to draw on to study the problem, a process that took place through meetings, discussions, reviewing literature and presentations of research within and outside of the project group. In phase III, we conducted research applying the following research methods to study youth football injury aetiology: questionnaire; measures of weight and height; injury diagnostics; clinical examinations; strength testing; running analyses; knee laxity measurements; examination of training protocols; field observations and interviews. Research methods were chosen based on knowledge gained from project team members’ disciplinary expertise, existing literature, internal meetings and discussions with experts outside of the research group. Additionally, phase III entailed recruitment of research participants.

In the final two phases, data were analysed through measurement-specific analyses (eg, three-dimensional kinematics to analyse movement-specific characteristics; isometric strength to analyse muscle weakness and/or muscle imbalance; content analysis of field notes and interview transcripts) and an analytic procedure was prepared that makes integrated analysis of measurement-specific analysis possible.

During the FIT project’s five-phase interdisciplinary research process, the research group encountered paradigmatic challenges typical of working interdisciplinarily. 61 One example is sample size and recruitment. Where positivist paradigmatic distinctions require a statistically powerful sample size, the interpretivist paradigm demands possibilities for in-depth and detailed examination of one or a few cases. A further challenge was the lack of an integrated data analysis procedure that would allow us to integrate the measurement-specific analysis referred to above. As we could not locate a suitable analytical procedure, we needed to spend time and effort to develop and test such a procedure. Finally, we have found that presenting the FIT project research in multiple cases resulted in misunderstanding and scientific criticism. As a result, we had to place extra effort in considering the communication of the FIT project’s scientific rationale, methodology and results. Indeed, while we have not struggled to respect our different paradigmatic worldviews and demands, it has been challenging to find and communicate what interdisciplinary researchers like Repko and Szostak 64 and Welch 65 call the ‘space in between’ scientific paradigms and disciplines we have been trying to create. The key measure that helped us to move forward in the interdisciplinary research process was a negotiation process that entailed regular meetings.

The purpose of this paper was to consider and outline an interdisciplinary research process to research the complex nature of youth sport injury aetiology. To facilitate this proposal, our narrative review of existing biomedical, sport psychological and sociological literature summarised youth football and youth sport injury research and demonstrated paradigmatic distinctions and methodological assumptions, research approaches, research methods and data analysis procedures. The narrative review has shown a paucity on youth football injury research in the sociological discipline as well as a dominance of monodisciplinary research and a lack of integration across research disciplines and paradigms.

Our paper has further shown that to advance youth sport injury research, specifically considering complexity, an interdisciplinary research process, such as proposed in figure 1 , has potential to integrate disciplinary knowledge and measurement-specific data across research paradigms. The integrated potential is particularly promising for research aiming to examine the interactions of components proposed vital to understand the complexity of injury aetiology. We recognise, however, that to advance this potential, additional research is necessary. We particularly see a need to develop and trial analytical procedures that integrate qualitative and quantitative injury aetiology data to produce complex injury aetiology findings, and to explore interdisciplinary research teams’ pragmatic negotiation of the challenges of melding seemingly opposing paradigms in the interests of better understanding the complexities of sport injury processes.

Ethics statements

Patient consent for publication.

Not required.

• Burwitz L ,
• Wilkinson DM
• Markula P ,
• Phoenix C ,
• Osborne NJ ,
• Redshaw C , et al
• Bittencourt NFN ,
• Meeuwisse WH ,
• Mendonça LD , et al
• Thompson J ,
• Nielsen RO , et al
• Hausken S ,
• Barker-Ruchti N ,
• Schubring A , et al
• Hausken SES ,
• Schubring A
• Pfirrmann D ,
• Ingelfinger P , et al
• Oliver JL ,
• De Ste Croix MBA , et al
• van der Sluis A ,
• Elferink-Gemser MT ,
• Coelho-e-Silva MJ , et al
• De Ridder R ,
• Witvrouw E ,
• Dolphens M , et al
• Hägglund M ,
• Steffen K ,
• Myklebust G ,
• Andersen TE , et al
• Visscher C ,
• Arends S , et al
• OʼKane JW ,
• Levy MR , et al
• Al Attar WSA ,
• Sinclair PJ , et al
• Bizzini M ,
• Atroshi I ,
• Magnusson H , et al
• van Mechelen W ,
• Meeuwisse WH
• Tyreman H ,
• Hagel B , et al
• McIntosh AS
• Donaldson A
• Johnson U ,
• Ivarsson A ,
• Lindwall M , et al
• Pensgaard AM ,
• Gledhill A ,
• Forsdyke D ,
• Andersen MB , et al
• Ekengren J ,
• Andersen MB
• Andersen MB ,
• Williams JM
• Williams JM ,
• Appaneal RN ,
• Wiese-Bjornstal DM
• Cavallerio F ,
• Wagstaff CRD
• Fenton LT ,
• Roderick M ,
• Waddington I ,
• Schubring A ,
• Barker-Ruchti N
• Hägerstrand T
• Bolling C ,
• Pasman HR , et al
• Vardon P , et al
• Jacobsson J ,
• Timpka T , et al
• Callaghan G
• Buekers M ,
• Ibáñez-Gijón J ,
• Morice AHP , et al
• Pringle R ,
• Campbell LM
• Bracken LJ ,

Contributors NB-R and SG conceptualised the FIT project, which is the basis for this paper. NB-R acquired the funding. AS, NB-R, SG and SEH-S contributed with the preparation of tools and data production for the FIT project. SEH-S is the PhD student in the FIT project. NB-R and SEH-S were responsible for the initial preparation and writing of the manuscript. RP, AS and SG contributed to the accurate and critical revision of the manuscript as well as approval of the final version. All authors were involved in the preparation and editing of the manuscript.

Funding This study was funded by The Swedish Research Council for Sport Science.

Competing interests None declared.

Provenance and peer review Not commissioned; externally peer reviewed.

Read the full text or download the PDF:

Advertisement

Athletic Injury Research: Frameworks, Models and the Need for Causal Knowledge

• Review Article
• Open access
• Published: 20 March 2024
• Volume 54 , pages 1121–1137, ( 2024 )

Cite this article

You have full access to this open access article

• Judd T. Kalkhoven   ORCID: orcid.org/0000-0002-1720-6879 1 , 2  

3295 Accesses

14 Altmetric

Explore all metrics

Within applied sports science and medicine research, many challenges hinder the establishment and detailed understanding of athletic injury causality as well as the development and implementation of appropriate athletic injury prevention strategies. Applied research efforts are faced with a lack of variable control, while the capacity to compensate for this lack of control through the application of randomised controlled trials is often confronted by a number of obstacles relating to ethical or practical constraints. Such difficulties have led to a large reliance upon observational research to guide applied practice in this area. However, the reliance upon observational research, in conjunction with the general absence of supporting causal inference tools and structures, has hindered both the acquisition of causal knowledge in relation to athletic injury and the development of appropriate injury prevention strategies. Indeed, much of athletic injury research functions on a (causal) model-blind observational approach primarily driven by the existence and availability of various technologies and data, with little regard for how these technologies and their associated metrics can conceptually relate to athletic injury causality and mechanisms. In this article, a potential solution to these issues is proposed and a new model for investigating athletic injury aetiology and mechanisms, and for developing and evaluating injury prevention strategies, is presented. This solution is centred on the construction and utilisation of various causal diagrams, such as frameworks, models and causal directed acyclic graphs (DAGs), to help guide athletic injury research and prevention efforts. This approach will alleviate many of the challenges facing athletic injury research by facilitating the investigation of specific causal links, mechanisms and assumptions with appropriate scientific methods, aiding the translation of lab-based research into the applied sporting world, and guiding causal inferences from applied research efforts by establishing appropriate supporting causal structures. Further, this approach will also help guide the development and adoption of both relevant metrics (and technologies) and injury prevention strategies, as well as encourage the construction of appropriate theoretical and conceptual foundations prior to the commencement of applied injury research studies. This will help minimise the risk of resource wastage, data fishing, p-hacking and hypothesising after the results are known (HARK-ing) in athletic injury research.

Similar content being viewed by others

A Review of Elite Athlete Evidence-Based Knowledge and Preparation for Competing in the Heat

The mental health of elite athletes: a narrative systematic review.

Exploring the Role of Sprint Biomechanics in Hamstring Strain Injuries: A Current Opinion on Existing Concepts and Evidence

Avoid common mistakes on your manuscript.

1 Introduction

Injuries in sport remain a major concern to athletes, practitioners and sporting organisations as the negative impacts of athletic injuries are widespread, affecting athlete health, individual and team performances, as well as sporting clubs’ economies. Additionally, previous injury remains one of the leading risk factors for an increased risk of subsequent injury [ 1 , 2 ], although in some contexts the causal nature of this relationship has been questioned [ 3 ]. Large scale epidemiological studies continue to report the persistent nature of athletic injuries, with injury rates remaining relatively stable over a number of years across a multitude of sports [ 4 , 5 , 6 ]. Markedly, some types of injuries, such as hamstring injuries, are reportedly even increasing in certain contexts [ 7 , 8 ].

Although it is difficult to attribute any lack of progress relating to athletic injury prevention to a single factor, a potential major contributor is a general absence of understanding regarding athletic injury causality. Indeed, injury prevention strategies have been widely implemented in the absence of causal explanations, causal frameworks or more generally, coherent supporting theories and conceptual foundations [ 9 , 10 , 11 ]. Such strategies are contradictory to the popularised ‘sequence of prevention’ [ 12 ], whereby the establishment of injury causality and aetiology form the foundations on which appropriate injury prevention strategies should be developed [ 12 ].

Determining the causes of natural phenomena is a key goal of science [ 13 ], and is an a priori requirement to the manipulation of variables to produce a favourable outcome. However, in many scientific areas, developing causal knowledge is difficult. Within applied sports science and medicine research, numerous challenges hinder attempts to develop detailed understandings of injury causation. For example, the application of randomised controlled trials (RCT; Table  1 ), particularly in elite sporting populations, is often confronted by a number of obstacles relating to either ethical or practical constraints that restrict their implementation. Such difficulties have arguably resulted in poor research practices and an over-reliance on observational research (Table  1 ) (without appropriate supporting causal inference tools and structures) for guiding both the acquisition of causal knowledge and the development and implementation of athletic injury prevention strategies. In addition, many current approaches to athletic injury research and prevention, and the development of new metrics in this area, appear to be primarily driven by the mere existence and availability of various technologies and data. For example, many metrics utilising technologies, such as global positioning systems (GPS) or inertial measurement units (IMU), have been proposed and adopted across the research and sporting landscape to assess athletic injury risks. However, little attention has been devoted to how these metrics and their underlying technologies can conceptually be related to the causal (mechanical and physiological) processes and mechanisms governing athletic injury occurrence (of which many different types exist) [ 9 , 10 , 11 ]. Such approaches are not reflective of theory-driven research, which is an essential component of the scientific method [ 14 , 15 ]. Rather, these approaches more closely resemble a (causal) model-blind approach at high risk of bias, data fishing [ 16 ], p-hacking [ 16 , 17 ] and hypothesising after the results are known (HARK-ing) [ 16 , 18 , 19 ]. Considering the current reliance on observational studies in athletic injury research, there remains an uncomfortable scarcity of much needed supporting causal inference tools and structures within the existing literature to guide statistical analyses and causal inferences. It follows that, many current approaches to understanding athletic injury and developing injury prevention strategies in the applied sporting world lack causal justification and are overly speculative.

To assist with some of the concerns described above, causal diagrams, including frameworks [ 20 ], models [ 14 , 21 ], causal directed acyclic graphs (DAGs; Table  1 ) [ 22 , 23 , 24 , 25 , 26 ] and other causal diagrams [ 27 , 28 , 29 ], are relevant tools that provide substantial value, organising ideas and directing future research and causal inferences. Specifically, these tools help guide the research process by outlining key concepts relating to causality, such as relevant causal assumptions, pathways and mechanisms. In addition, these diagrams also have important implications for statistical analyses, with causal DAGs in particular outlining which variables to include and adjust for in a statistical analysis to identify causal and non-causal effects. Indeed, causal DAGs are instrumental in addressing bias, explicitly illustrating potential confounders (Table  1 ), while also highlighting when controlling for a variable inappropriately will introduce new bias into the analysis, such as collider-stratification bias (Table  1 ). Accordingly, the adoption of causal diagrams may help alleviate many of the challenges facing athletic injury research by facilitating the investigation of specific causal links, assumptions and mechanisms with appropriate scientific methods, aiding the translation of lab-based research into the applied sporting world, and guiding causal inferences from applied research efforts by establishing appropriate supporting causal structures. Further, the utilisation of causal diagrams will also help with the development and adoption of both relevant metrics (and technologies) and appropriate injury prevention strategies, and encourage the construction of coherent theoretical and conceptual foundations prior to the commencement and allocation of resources to applied injury research efforts. This will help minimise the risk of bias, resource wastage, data fishing [ 16 ], p-hacking [ 16 , 17 ] and HARK-ing [ 16 , 18 , 19 ] in athletic injury research. In light of these proposed benefits of causal diagrams, the aims of this review article are two-fold; (1) to highlight some of the major challenges and shortcomings of some currently adopted approaches to athletic injury research, and (2) to illustrate how using causal diagrams for athletic injury research can lead to more robust scientific findings and causal understandings. Further, within this article a new model proposing the integration and utilisation of frameworks and models for investigating injury aetiology and mechanisms, and for developing athletic injury prevention strategies, is also presented.

Finally, it is important to acknowledge that ‘causality’ is a complicated metaphysical concept that remains the subject of many philosophical debates. Further, ‘causal inference’ is a complex and technical scientific and mathematical task that relies on triangulating evidence from multiple sources and on the application of a variety of methodological approaches [ 24 , 25 , 26 ]. While a deep dive into the philosophy and history of causality, and the vast array of methodological tools available for causal inference across the sciences, would certainly be of value, such a task requires a large body of work that is outside the scope of this article. Rather, this article simply highlights the necessity for causal knowledge within the context of athletic injury prevention, some of the major challenges facing the pursuit of causal knowledge in athletic injury research and prevention, and the value of graphical tools to the research process. In the current literature, various attempts to define causation have been presented [ 30 ], but a clear and agreed upon definition of causation continues to be elusive [ 25 , 30 ], with some authors suggesting that the formation of such a definition may be too reductionist [ 25 ]. Accordingly, to avoid entering too deeply into the depths of the philosophy of causality, and for the operational purposes of this article, causality should tentatively be interpreted by its common understanding of cause and effect, i.e. the operation or relation of a cause and its effect. It should be noted that the terminology and language used in this article has been adopted from the (mutually compatible) causal inference [ 24 , 25 , 26 ] and sufficient component causal model [ 31 , 32 , 33 ] perspectives for understanding causality.

2 The Four Step ‘Sequence of Prevention’ and ‘TRIPP’: A Brief Overview

Injury prevention research has been described by van Mechelen et al. [ 12 ] as a four step sequence in the highly popularised ‘sequence of prevention’ (Fig.  1 a). The ‘sequence of prevention’ is as follows. First, the magnitude of the problem should be identified and described in terms of the incidence and severity of sports injuries. Secondly, the risk factors and injury mechanisms that play a part in the occurrence of sports injuries should then be identified. The third step is to introduce measures that are likely to reduce the future risk and/or severity of sports injuries. Such measures should be founded on knowledge regarding the aetiological factors and injury mechanisms identified in the second step. Finally, the effect of the measures should be evaluated by repeating the first step, which can be achieved by time trend analysis of injury patterns or by means of a RCT [ 34 ]. Although various modifications and adaptations to this model have been proposed in literature, with arguably the most notable being the ‘Translating Research into Injury Prevention Practice framework’, or ‘TRIPP’ [ 35 ] (Fig.  1 b), this article is primarily concerned with the establishment of injury aetiology and mechanisms and how to use this information to form and implement injury prevention strategies. Therefore, the current variations to this model and their added components will not be explored in detail here. However, their existence and contributions to the area are acknowledged. For further exploration on these models please see articles by van Mechelen et al. [ 12 ], Bahr and Krosshaug [ 36 ], and Finch [ 35 ].

Modified from van Mechelen et al. [ 12 ] and Finch [ 35 ] with permission

The ‘sequence of prevention’ ( a ) and the ‘Translating Research into Injury Prevention Practice (TRIPP)’ ( b ) models.

3 Establishing Injury Aetiology and Mechanisms: Current Challenges

3.1 variable control and randomised controlled trials in athletic injury research.

A key goal of science is to establish the causes of natural phenomena [ 13 ]. Causal knowledge provides the critical foundations on which informed and appropriate actions can take place to alter a given outcome, i.e. intervention. In the context of athletic injury, it is unsurprising then that a key feature of the van Mechelen ‘sequence of prevention’ [ 12 ] and Finch’s ‘TRIPP’ framework [ 35 ] is to introduce measures that are likely to reduce the future risk and/or severity of sports injuries on the basis of causal and mechanistic understandings. However, developing causal knowledge is a challenging endeavour. To develop causal knowledge, the controlled experiment is a mainstay of modern science, and for good reason. Although context dependent, it is generally maintained that, to estimate a causal effect, all major variables at baseline that can influence both the exposure and the outcome must be controlled for [ 24 , 25 , 26 ]. By holding constant all major variables influencing both the exposure and the outcome (with the exception of the independent variable of interest), one can determine if a particular variable is indeed responsible for a given outcome. However, achieving such conditions in real world settings can be extremely challenging. This is problematic. If a confounding (Table  1 ) variable remains uncontrolled for, this can lead to inaccurate conclusions about the causal relationship between variables (confounder bias). Concerns such as these gave rise to a particular type of controlled study, the RCT. In modern medicine, RCTs are commonly regarded as the ‘gold standard’ for assessing causality [ 37 ], as it is widely considered that the randomisation process applied to the samples within these types of studies serves to eliminate the effects of confounding variables [ 13 , 37 ]. However, while RCTs certainly have high internal validity (Table  1 ) [ 38 ], the ‘gold standard’ label commonly attributed to this type of study remains controversial, as under many circumstances ethical or practical constraints oppose their implementation [ 25 , 26 , 38 ]. For example, in the absence of clinical equipoise (Table  1 ), it would be unethical to conduct an RCT that actively exposes one group of athletes to an intervention that is harmful (in relation to injury, performance or wellbeing). Notably, in some cases intervention may be physically impossible. For example, if we are interested in investigating the effect of either age or previous injury on injury risk, we cannot randomise individuals to different age groups or to having a previous injury. Thus, a RCT is not a feasible option for directly studying the causal effect of either of these variables on injury risk. It follows that, while RCTs certainly have high internal validity and should be conducted when feasible, the best method for understanding causality ultimately depends on what methods are actually available for answering a specific causal question [ 25 , 26 , 38 ].

Where RCTs are technically feasible in athletic injury research, there are also many practical challenges to overcome for the application of this particular type of study. At the forefront is obtaining a sufficient number of participants and injuries. This is crucial for achieving the necessary statistical power to ensure exchangeability (Table  1 ) between groups and to maintain the integrity of the RCT. However, this can be difficult, especially when a relatively small, potentially highly specialised population, such as elite athletes, is being investigated or when a specific type of injury is being studied. Indeed, conducting RCTs on elite sporting populations can be a particularly challenging endeavour, as not only are these populations limited in size, but owing to their professional nature, elite athletes and sports organisations may prove unwilling to cooperate and participate in studies that could potentially interfere with their training and preparation processes. As a result, RCTs are often conducted on alternative population groups, e.g. sub-elite athletes. Approaches such as these raise a number of concerns relating to the external validity and transportability (Table  1 ) of results when the findings from these studies are applied to different contexts, such as elite athletes [ 39 ]. While some of the practical considerations mentioned are not strictly insurmountable, these obstacles still pose significant challenges that researchers need to overcome to conduct this type of study successfully. Encouragingly, however, it is simply untrue that RCTs are the only way of discovering causality and developing causal knowledge [ 13 , 25 , 26 ]. Certainly, young children did not require a robust RCT to learn that putting their hand in the fire causes a painful burning sensation [ 13 ]. While RCTs are still widely considered to be the ‘gold standard’ for assessing causality [ 13 ] and should be done when feasible, exploration of alternative research approaches that may assist with the development of causal knowledge in relation to athletic injury appears necessary.

3.2 Observational Research: Benefits and Pitfalls

Consistent struggles relating to ethical or practical constraints often restrict applied athletic injury research efforts and the appropriate application of RCTs. While RCTs are still widely considered to be the ‘gold standard’ for establishing causation, owing to their often-infeasible nature, especially at the elite sporting level, exploration of alternative methods and research practices that may assist with developing causal knowledge and guiding applied practice is warranted. In sports science, owing to their general simplicity and greater feasibility, descriptive research methodologies are frequently conducted to determine associations or correlations between variables, make predictions or describe injury risks and rates. These are useful studies that can provide the first exploratory step in the process of causality determination by identifying specific variables of interest. Across sports science, observational studies investigating the relationship between a selected variable of interest and athletic injury risk are particularly common. However, an important consideration when examining these types of studies and the potential relationships that may be present, is that association and correlation do not imply causation [ 13 , 40 ]. This is a frequently reiterated mantra within the scientific community, and for good reason. Associations can present for a number of reasons other than causality, such as confounding [ 13 , 24 , 25 , 26 , 40 ].

To better highlight the issue of confounding, a simple non-sporting example of confounding is presented in Fig.  2 . In this example, the causal relationships between three variables (age, shoe size and education level) in children are considered. Reasonably, it can be concluded that shoe size is not a cause of the level of education that a child receives, and the level of education a child receives is not a cause of a child’s shoe size, i.e. there is no causal link between these two variables. Despite this, shoe size and education level remain highly correlated in children. The reason for this correlation is that these two variables are statistically related by the effects of a third ‘confounding’ variable, age. In this example, age is considered to be a confounder as it is a common cause of both shoe size and education level (as children get older their feet grow and they go up school years). The consequence is that, owing to the effects of age on both shoe size and education level in children, shoe size and education level are statistically associated with one another despite there being no causal relation between them. Confounding is common in science, with observational studies being particularly susceptible to this form of bias. Accordingly, while observational studies may provide an estimate of the statistical association or correlation between variables, interpreting these relationships in a causal manner can be problematic and should only be undertaken after careful evaluation of the underlying assumptions.

A signed causal DAG [ 62 , 63 ] illustrating the basic structure of confounding: age is a common cause (confounder) of the relationship between shoe size and education level, producing a statistical relationship between shoe size and education level despite there being no causal relationship between these two variables. The (+) symbols represent a positive effect of age on shoe size and education level. DAG directed acyclic graph

3.3 Confounding in Athletic Injury Research

3.3.1 definitions of athletic injury.

Confounding is a causal concept [ 25 ], and accordingly, prior to providing some examples of confounding in applied athletic injury research, it is worthwhile to explore some of the causal complexities and fundamental mechanisms that underpin athletic injury occurrence. It has previously been noted that injury occurs as a result of a transfer of kinetic energy to a tissue [ 36 , 41 , 42 ]. Athletic injury has also been described as occurring when the stresses and strains (Table  1 ) experienced by a tissue result in damage that is deemed severe enough to be considered an injury, i.e. tissue loading exceeds tissue strength [ 42 , 43 ]. Note that these descriptions are, in fact, the same as one another, with the area under a stress–strain curve representing the energy absorbed during deformation, typically as a result of a transfer of kinetic energy. It follows that the International Olympic Committee (IOC) defines injury as “tissue damage or other derangement of normal physical function due to participation in sports, resulting from rapid or repetitive transfer of kinetic energy” [ 41 ]. While this is a useful definition, it is important to clarify some inaccuracies. Most notably, it is clear that damage or other derangement can exist without an injury necessarily being present. For example, muscle damage is common after sporting activity and may be an unavoidable part of the training process, while it can also be argued that fatigue constitutes a form of derangement of normal physical function. These should not be considered an injury. Rather, an injury more accurately occurs when the tissue damage experienced exceeds some critical damage threshold, whereby the damage sustained is not a normal part of the training process, but is rather chronically detrimental and severe enough to be considered an injury [ 43 ]. Future definitions of athletic injury should make such distinctions clearer.

3.3.2 The Ultimate Mechanism of Athletic Injury

Considering the definitions of athletic injury presented above, and as highlighted within a recently presented conceptual framework for athletic injury [ 42 ], the causes of an athletic injury can conceptually be separated into two fundamental components: the mechanical loading (force) experienced by a tissue and the mechanical strength (Table  1 ) of that tissue, one (or both) of which all causal variables must act through, i.e. these two components are reflective of the proximate (ultimate) mechanism of athletic injury. For illustrative purposes, two causal DAGs displaying this assumption for both acute and gradual onset injury in athletes are presented in Figs.  3 and 4 .

A signed causal DAG [ 62 , 63 ] depicting mechanical loading exceeding tissue strength as the ultimate mechanism of acute athletic injury. A generic wider causal structure is included for illustrative purposes with the ‘U’ variables representing unspecified variables and the ellipses alluding to the existence of a wider causal network beyond that presented in the DAG. The (+) symbol reflects that as the mechanical load increases, the probability of the load exceeding the tissue strength and causing injury also increases. The (−) symbol reflects that higher tissue strength reduces the probability of the mechanical load exceeding this strength, decreasing the chance of injury. DAG directed acyclic graph

A signed causal DAG [ 62 , 63 ] depicting repetitive mechanical loading as a cause of the deterioration of tissue strength (mechanical fatigue) and which eventually results in gradual onset injury occurrence. The subscript values represent the sequence of events, e.g. the tissue of interest has its original strength (tissue strength 1 ) prior to a mechanical load ( 1 ) being applied. This causes a reduction in tissue strength ( 2 ) owing to mechanical fatigue. A second subsequent mechanical load ( 2 ) is applied, which causes a further reduction in tissue strength ( 3 ). This process continues until a critical damage threshold is exceeded and injury occurs. The (−) symbols represent a negative effect of mechanical loading on tissue strength (i.e. a reduction in tissue strength). The (+) symbols reflect that a higher initial tissue strength (e.g. tissue strength 1 ) leads to a higher subsequent tissue strength (e.g. tissue strength 2 ) despite repetitive mechanical loading reducing tissue strength over time. The ellipses in the figure represent a time jump to the point where injury occurs. Note: Physiological processes, i.e. remodelling and repair, have been intentionally omitted from this DAG for simplicity. DAG directed acyclic graph

In Fig.  3 , a causal DAG is presented that is reflective of acute sudden onset athletic injury, whereby a single mechanical load exceeds the strength of a tissue. Some examples of this include anterior cruciate ligament rupture owing to sudden knee trauma, tibial break owing to a poorly timed tackle in soccer etc. In this DAG, the exceeding of tissue strength by a single mechanical load serves as the ultimate mechanism of injury. The ‘U’ variables presented in the diagram represent unspecified parent nodes (distal variables) of the ultimate mechanism and are included to represent a wider causal structure for athletic injury. This wider causal structure may be extensive and will vary between each individual injury type owing to unique, differing etiological pathways. In Fig.  4 , a simple (wider variables excluded) causal DAG for gradual onset injury is presented, assuming no common causes of mechanical load, tissue strength and injury. Under this same assumption, an example of this type of injury may include a tibial stress fracture owing to repetitive loads from running. In this figure, repetitive mechanical loads fatigue a tissue until a critical damage threshold is exceeded and an injury occurs. In this type of injury physiological processes can also affect tissue strength, e.g. through remodelling and repair processes, but for simplicity’s sake this has intentionally been excluded from the DAG. In Fig.  4 , the mechanical strength of the tissue is considered a time-varying variable as it deteriorates over time owing to cyclic loading. It is for this reason that tissue strength after each separate applied mechanical load is considered a different variable to tissue strength prior to that load; after each mechanical load is applied, tissue strength is represented by a new separate node.

3.3.3 Confounding in Athletic Injury Research

Within sports science and medicine research, many observational studies that assess relationships with and describe injury risks have been conducted across the sporting landscape, with a number of injury prevention recommendations and strategies commonly (but often erroneously) arising from such studies. As these studies are observational in nature, they are particularly susceptible to the previously highlighted issue of confounding. Accordingly, in the absence of appropriate supporting causal inference tools and structures, approaches such as these do not offer a reliable means for acquiring the necessary causal knowledge needed to intervene and alter injury outcomes.

To illustrate some of the problems that arise when using observational studies for estimating causal effects in relation to athletic injury, a contextually relevant example of confounding may prove useful. For instance, consider a theoretical scenario in which a researcher adopts an observational approach to investigate the relationship (association) between some hamstring tissue characteristic, labelled ‘T’, and the risk of hamstring injury in soccer players. For the purpose of this example, let us stipulate that (1) the data utilised to assess this relationship are derived from a large pool of soccer players of varying ages, including both elite and junior players; (2) T is some developmental characteristic, i.e. it increases with maturation into adulthood, and accordingly, elite level soccer players exhibit greater levels of T compared with junior level soccer players; and (3) greater levels of T also have a protective causal effect on hamstrings by increasing the mechanical strength of the hamstring muscles, but have no causal effect on athletic performance. Accordingly, in this scenario age has a beneficial causal effect on injury risk by increasing levels of T. However, age can also influence injury risk through other causal pathways. For example, elite adult soccer players sprint faster than junior soccer players, which subsequently increases the stresses placed on the hamstrings [ 44 , 45 ]. Reasonably, this may increase hamstring injury risk. The nature of these proposed relationships is depicted in Fig.  5 .

A causal DAG illustrating that age has a causal effect on hamstring injury through two identified causal pathways: (1) T, which has a protective effect on hamstring injury risk by increasing hamstring strength, and (2) increased hamstring stresses owing to faster sprint speeds (elite adult soccer players sprint faster than junior soccer players), which have a negative effect on hamstring injury risk. The (+/−) symbols represent a positive (increasing) and negative (decreasing) effect, e.g. the (−) symbol illustrates a reduction in hamstring injury risk owing to increased hamstring strength. Note: This is a simple DAG that is presented for illustrative purposes and should not interpreted as a complete example. DAG directed acyclic graph

As shown in Fig.  5 , age has an effect on hamstring injury through two causal pathways: (1) T, which has a protective (decreasing) effect on hamstring injury risk by increasing hamstring strength, and (2) increased hamstring stresses owing to faster sprint speeds (elite adult soccer players sprint faster than junior soccer players), which has an increasing (harmful) effect on hamstring injury risk. In this scenario, age is a common cause (confounder) of both T and the mechanical stresses placed on the hamstring. Therefore, the association that presents between T, our exposure of interest, and the risk of hamstring injury, our outcome of interest, will not be reflective of the causal relationship between these two variables, as this relationship will be impacted by confounding. This is important, since if higher levels of T are primarily associated with elite adult soccer players, and elite adult soccer players exhibit a heightened risk of hamstring injury (which indeed appears to be the case [ 46 ]), this may result in increased levels of T being statistically associated with an increased risk of hamstring injury despite actually having a causally protective effect. In scenarios such as this, causal interpretations of these associations may have dire consequences. Protective variables, such as increased levels of T, may be interpreted as being harmful to an athlete’s injury outcomes, while harmful variables, such as decreased levels of T, may be interpreted as being beneficial, i.e. causality is misinterpreted. Concerningly, in some contexts misinterpretations of causality will likely facilitate the implementation of inappropriate strategies that actively detrain athletes or damage athletic pursuits in an attempt to improve injury risk [ 9 ], and in some circumstances, may potentially even increase injury risk, e.g. implementing strategies that reduce levels of T as a result of T being associated with an increased injury risk when increased levels of T actually have a protective causal effect.

Alone, observational studies exploring relationships with injury risk cannot be expected to reliably bestow the necessary causal knowledge required to develop strategies to manipulate an athlete’s injury risk. Indeed, such is the potential impact of confounding on research outcomes, confounding can result in no association when there is a causal effect, an association when there is no causal effect and in some circumstances, associations that are in the opposite direction of the actual causal effect, such as in the example provided in Fig.  5 . While the utilisation of observational studies for causal inferences is problematic owing to bias (‘association or correlation does not imply causation’), it is important to note that ‘some correlations do imply causation’, but appropriate causal inference tools and structures are needed [ 25 ]. Indeed, various tools exist that can assist with providing proposed causal structures, identifying and controlling for confounding and other forms of bias, as well as obtaining causal effects and knowledge from observational research, with this article being particularly concerned with introducing the reader to the importance of causal frameworks, models and DAGs in particular.

4 Frameworks, Models and DAGs: Organising Ideas for Understanding Causation

Owing to the many challenges in athletic injury research and the potential pitfalls accompanying attempts to acquire causal knowledge from observational studies, there is a need to explore potential strategies that can assist with developing causal understandings of athletic injury. Notably, the formation and utilisation of frameworks, models and causal DAGs provides an approach that may alleviate some of these current obstacles and provide value for research pursuits.

4.1 Frameworks and Models

In science, a framework serves as an intellectual structure (often graphically represented) that is used to make conceptual distinctions and organise ideas, deconstructing complex phenomena into relevant theories, assumptions, causal links and concepts of interest. In this respect, frameworks break down a system into relevant components, and align these components within a proposed theoretical, conceptual and/or causal architecture. Through this, frameworks help orientate research, with the adopted ideas and concepts assisting with the development and investigation of various research problems/questions, hypotheses, assumptions and causal links, while also serving as guides for collecting and analysing data and for interpretations of causality. Considering these benefits, it is unsurprising that some researchers have argued that, in the absence of a framework, research and scholarship are considerably ‘weakened’ [ 47 , 48 ]. A variety of framework types are recognised across science, such as conceptual, theoretical and practical (Table  1 ), each distinguished by its unique functions and objectives [ 20 ].

While a scientific theory (Table  1 ) provides a substantiated explanation of a phenomenon in the natural world, and is used to make predictions that are testable by experiments or observations [ 47 ], a framework has the unique capacity to weave together multiple theories by providing a context within which the theories apply. Theories that might seem disparate at first can be interlinked within a framework that highlights their relevance to one another, and to a larger body of knowledge. Notably, when researchers contemplate how the various elements within a framework interrelate, they engage in theorising, i.e. form conjectures about the relationships between the concepts identified for a particular phenomenon. This process involves crafting hypotheses about the interplay among the concepts delineated for a given phenomenon, enhancing our understanding of the phenomenon and potentially leading to new scientific insights.

In contrast to theories and frameworks, models can be conceptualised as instantiations of theories or frameworks, and are therefore typically narrower in scope. Models may focus on a particular component of a given theory or framework, highlighting key elements, properties, and relationships within that component [ 14 , 21 ]. Accordingly, models typically provide a more local description or understanding of a phenomenon, and commonly serve as intermediaries between theories and the real world [ 14 ]. However, owing to their more focused nature, models may also ignore a large part of reality. This is considered a feature [ 14 , 21 ]. Much in the same way that a map of Rome is useful because it ignores much of reality to help us navigate Rome, a model may be useful because it ignores the rest of a particular theory or framework, instead focusing on specific components of interest [ 14 , 21 ].

The utilisation and investigation of frameworks and models, and their proposed research links and assumptions, is akin to identifying, researching and confirming pieces of a puzzle with the intention of, one day, explaining the puzzle in its entirety. By deconstructing the problem into components, alternative scientific approaches and practices may be considered, and various elements and mechanisms may be targeted for investigation, potentially alleviating many of the challenges facing injury outcome-based research. Indeed, such an approach may be particularly useful for the translation of lab-based research into the applied sporting world, whereby the investigation of various causal links and assumptions within the controlled laboratory setting can better illuminate the specific roles of relevant variables and mechanisms within a given system. For example, as opposed to investigating the relationship between various tissue characteristics and athletic injury within the uncontrolled environment of the sporting arena, an alternative (and supplementary) approach that might offer important scientific insights could be to assess the relationship between these same tissue characteristics and tissue strength (under the assumption that tissue strength is a relevant mechanism through which various tissue characteristics can influence injury outcomes) within the controlled environment of the laboratory. Similarly, as opposed to researching the relationship between certain training load metrics of interest and athletic injury outcomes in applied sports settings, researchers can take an alternative approach. They could instead research how these training load metrics relate to tissue forces and cumulative tissue damage under the assumption that tissue forces and cumulative tissue damage are relevant mechanisms through which training load can cause certain types of athletic injury [ 9 , 43 , 49 ]. The knowledge gained from investigating specific research links in more controlled, and potentially more viable (depending on the context) environments can then inform the construction or revision of relevant frameworks. Researchers can then use the underlying theory and causal knowledge presented within these frameworks to develop useful multi-faceted models constructed with appropriate scientific reasoning for application in the applied sporting world, and to guide injury prevention strategies and decision-making. The effectiveness of these strategies can then be assessed by time trend analysis of injury patterns, observational studies using appropriate causal inference methodologies, or when possible, RCTs. Additionally, the comprehensive utilisation of frameworks and models can help guide the development of new, relevant technological advancements and metrics, while also ensuring that a coherent theoretical and conceptual underpinning is presented prior to the allocation of resources for resource intensive applied injury research studies (such as large-scale multicentre studies and RCTs). Accordingly, their adoption will assist with the development of appropriate conceptual foundations prior to the commencement of research studies, reducing resource wastage and minimising the risk of data fishing [ 16 ], p-hacking [ 16 , 17 ] and HARK-ing [ 16 , 18 , 19 ]. Importantly, a framework can include a number of propositions, and accordingly, the various assumptions and links put forward within a framework should be challenged and either confirmed or dismissed through empirical evidence. Upon the emergence of data supporting new ideas and the subsequent dismissal of certain assumptions and links, a framework must be either revised or replaced, which is a normal step in the scientific process [ 20 ]. A proposal on how best to utilise frameworks and models for the investigation of injury causation and the development of injury intervention strategies is presented in Fig.  6 .

A proposed model for facilitating the integration of frameworks and models into the athletic injury research and prevention process

4.2 Causal Directed Acyclic Graphs (DAGs)

Causal DAGs, numerous examples of which have been presented throughout this article, are powerful tools in the science of causal inference. DAGs are a specific type of acyclic graph, i.e. there are no cycles, which provides a simple way of graphically representing, communicating and understanding key concepts of relevance regarding causality [ 23 ]. Specifically, DAGs provide a proposed causal model of reality, constructed by an investigator on the basis of their beliefs regarding the topology of the causal processes at work [ 25 ]. In this respect, DAGs require a subjective commitment by the investigator and can serve as a form of causal framework that explicitly lays out any underlying causal assumptions. This is an important feature that provides much needed transparency, as too often the causal assumptions of researchers are undeclared, and therefore unknown and unverifiable [ 26 ]. However, it is important to emphasise that DAGs are not valuable simply because they are explicit, nor are they solely conceptual in nature. DAGs are also complex mathematical tools, with each arrow between variables representing a quantifiable causal effect. It follows that, the development of DAGs has provided a critical step in the ‘mathematization’ of causal inference [ 25 ], and their value and importance in the science of causal inference should not be underestimated [ 24 , 25 , 26 ]. Indeed, these diagrams have opened up new avenues for understanding causality from observational research [ 24 , 25 , 26 ], as not only do DAGs provide a practical means of graphically representing specific causal pathways, mechanisms and assumptions, they also have important implications for guiding statistical analyses for the acquisition of causal effects from observational data. DAGs serve as potent instruments for addressing bias, explicitly showcasing critical confounders [and deconfounders (Table  1 )], as well as highlighting when inappropriately adjusting for a variable will introduce new bias into the analysis, e.g. collider-stratification bias. As testament to their potency, a number of notable research successes can be attributed to DAGs, with these diagrams providing explanations for a series of apparent paradoxes, including the so-called Berkson’s [ 50 ], birth weight [ 51 ], obesity [ 52 ] and Simpson’s [ 53 ] paradoxes.

4.2.1 Causal Paths and Mediation (Mechanisms)

A major benefit of causal DAGs is the explicit nature in which these diagrams outline specific causal pathways. In any complex system, numerous causal pathways may exist. In DAGs, a causal path is represented by a sequence of arrows (edges) that connect a set of variables (nodes), indicating the direction of causal influence from one variable to another. These paths represent hypotheses about how changes in one variable might propagate through a system to affect another variable. The identification of a causal path in a DAG can be quite straightforward. For example, in Fig.  5 , it is quite clear that there are two hypothesised causal pathways through which age can impact hamstring injury risk. The identification of causal pathways is important for a number of reasons; it provides clarity in understanding direct and indirect effects (Table  1 ) within a causal relationship [a key feature of causal mediation analysis (Table  1 )], enhances the accuracy of statistical models in predicting outcomes, and guides the development of more effective interventions and policies. Furthermore, it plays a critical role in identifying instances where an effect is mediated through another variable [a mediator (Table  1 )]. In science, mediators are considered to be the mechanism that transmits the effect of one variable on another [ 25 ], and are visually depicted in a DAG as a variable that lies on the causal path between two variables. A simple example of this can be observed in Fig.  7 , whereby calcium intake is an identified mechanism through which diet impacts bone strength.

A causal DAG presenting the basic structure of mediation (Table  1 ). In this DAG, the effect of diet on bone strength is mediated by calcium intake, i.e. calcium intake is a mediator. DAG directed acyclic graph

While effective treatments have been produced in the absence of identified mechanisms, the identification of mechanisms is invaluable to science, as understanding mechanisms is critical for guiding interventions under changing conditions [ 25 ]. For example, in scenarios where bone strength is a concern and dietary consumption of calcium from food sources is lacking, understanding calcium intake as a key mechanism through which diet can impact bone strength can be vital for guiding the development and implementation of appropriate intervention strategies, such as additional calcium supplementation. Of course, many mechanisms can exist within a causal path. For example, in Fig.  7 , one might identify calcium absorption as an important mediator of the effect of calcium intake on bone strength, and this could be added to the DAG. Nutrition is an area of science that offers many examples highlighting the importance of understanding mechanisms. In the late 1800s and early 1900s, had it been known that vitamin C was the mechanism through which citrus fruit worked to prevent scurvy, many sailors would not have lost their lives. Indeed, oranges and lemons would never have been replaced with cheaper alternatives (such as West Indian limes) that contained a fraction of the vitamin C [ 25 ]. Further, sailors would not have taken to boiling these fruits, degrading the vitamin C within them and disabling the mechanism through which fruits prevented scurvy. Subsequently, had it been known that vitamin C was the mechanism through which certain fruits prevented scurvy, this also would have prevented the role of citrus fruits in preventing scurvy being brought into disrepute [ 25 ].

4.2.2 Back-Door Criterion

The back-door criterion is a fundamental concept in causal inference, providing a systematic method to identify and address confounding. This is achieved by identifying a set of variables that, when controlled for, block all ‘back-door paths’ (Table  1 ) from a treatment or exposure variable to the outcome variable of interest. A back-door path refers to any path from the treatment (or exposure) variable to the outcome variable that goes through a confounder. Accordingly, back-door paths are indicative of confounding and can create spurious (non-causal) associations between variables. The explicit nature of DAGs makes the identification of back-door paths, and by extent confounding, relatively straightforward. To exemplify this, a DAG displaying two back-door paths is presented in Fig.  8 .

Back-door paths. Note: This is a simplistic DAG that is presented for illustrative purposes and should not be interpreted as a genuine proposal for risk of bone fracture from falling (RBFFF). In this scenario, bone strength is our exposure variable, and RBFFF is our outcome variable of interest. The red lines and numbered circles in b represent back-door paths (paths 1 and 2) between bone strength and RBFFF, which reveals confounding, i.e. which variables to control for in the statistical analysis to reveal the true causal effect of bone strength on RBFFF. Note that DAGs are typically presented as in a . It is abnormal to explicitly illustrate back-door paths in DAGs. The red lines are included in b for illustrative purposes. DAG directed acyclic graph

In the example presented in Fig.  8 , bone strength is our exposure variable of interest, and risk of bone fracture from falling (RBFFF) is our outcome variable. In this example, we can see that there are two back-door paths (presented in red) between bone strength and RBFFF, both of which pass through age: path 1, which includes age and U (an unknown variable), and path 2, which includes age and balance. By identifying these paths, we now know which variables to control for in our analysis to reveal the true causal effect of bone strength on RBFFF (assuming that the DAG presented provides a valid representation of the world). As both back-door paths pass through age, the simplest solution to this problem would be to control for age, as doing so would close both back-door paths, i.e. by controlling for age we have controlled for all relevant bias in the DAG, and we can now determine the true causal effect of bone strength on RBFFF. However, let us assume that age is unavailable to us, and therefore cannot be controlled for. An alternative solution to this problem would be to control for the other variables along these back-door paths, as long as doing so does not introduce new bias into the analysis. By controlling for U, back-door path 1 is closed, but back-door path 2 remains open. Accordingly, to control for all of the bias, balance must also be controlled for to close back-door path 2. Once both back-door paths are closed, all of the bias is controlled for. Unfortunately, as U is an unknown variable in this example, we might not have a measure for it, which is a problem if we want to control for it. While there are certainly scenarios where controlling for all relevant confounding and bias using the back-door criterion is simply not possible, where the back-door adjustment fails, other methods not explored in this article are available and may offer alternative solutions, e.g. front-door adjustment, instrumental variables, Do-calculus, G-methods etc. [ 24 , 25 , 26 ].

4.2.3 Collider-Stratification Bias

Confounding has featured quite prominently throughout this article, with its most basic structure being presented in Fig.  2 . While confounding is a relatively well-known problem in research, a more commonly overlooked problem is that of ‘collider bias’, a form of overcontrol. Collider bias, in many ways, presents the opposite problem to confounding. In confounding, owing to the effects of a third confounding variable, two variables can start out statistically associated with one another despite actually being causally independent, i.e. there is a statistical relationship between two variables despite there being no causal relationship between these variables. In collider bias, two causally independent variables can start out statistically independent of one another, but by incorrectly controlling for a third (collider) variable, a spurious (non-causal) association is produced. To better illustrate this issue, a simple example of ‘collider bias’ is presented in Fig.  9 . For this example, the shoe size and school children example presented in Fig.  2 is revisited, except in this case the causal relationships between age, shoe size and a new variable, biological sex, are considered. On the basis of our causal understandings of the relationships between these variables, it can reasonably be concluded that a child’s age is not a cause of a child’s biological sex, and biological sex is not a cause of a child’s age, i.e. there is no causal link between these two variables. However, both age and biological sex are causes of a child’s shoe size, i.e. as a child gets older their shoe size increases, and biological males tend to have larger feet than biological females as they mature [ 54 ]. This proposed causal structure is visually depicted in the DAG presented in Fig.  9 .

A causal DAG presenting the basic structure of a collider. In this DAG, age and biological sex are independent causes of a child’s shoe size, and shoe size is a collider. DAG directed acyclic graph

In Fig.  9 , it is clear that there is no causal relationship between age and biological sex, as indicated by the absence of an arrow between these variables. However, both age and biological sex have a causal effect on shoe size. In this example, shoe size is a collider (Table  1 ; illustrated by the two arrows ‘colliding’ at this variable). In causal inference, conditioning on a collider is disastrous if one’s aim is to find the causal effect between two variables, as doing so distorts the estimates of associations (collider bias). For example, if we were to control for shoe size (the collider) when analysing the relationship between age and biological sex, a new bias will be introduced into our analysis (collider bias), and age and biological sex will become statistically associated despite there being no causal relationship between these two variables. In certain scenarios, collider-related bias can become even more pronounced, such as when the analysis is stratified on the basis of different levels of a collider (collider-stratification bias). For example, if we were to stratify our analysis by different shoe size (the collider) categories when analysing the relationship between age and biological sex, we would introduce collider-stratification bias into our analysis. While it is unlikely that one would stratify their analysis by shoe size in a study investigating the relationship between age and biological sex, there are a number of well-known examples where collider-stratification bias has not been so obvious. Indeed, the identification of collider-stratification bias has led to important developments in improving understandings of various phenomena. For example, this form of bias largely explains the association reported between postmenopausal hormone treatment and coronary heart disease [ 55 ], the birth weight paradox [ 51 ] and the obesity paradox [ 56 ].

4.2.4 A Message of Caution for the Implementation of Causal DAGs

While it should be noted that some previous calls to include DAGs in sports science and medicine research have been presented in literature [ 57 ], these have been bereft of information and a more detailed outlining appeared necessary to foster greater engagement with these particularly powerful research tools. However, it is also important to emphasise that the information provided in this article should not be considered comprehensive or sufficient for the practical implementation of causal DAGs. Despite its apparent simplicity, the use of causal DAGs in research includes nuances, and a deeper understanding can help avoid some common pitfalls [ 58 ]. In addition, DAGs have many functions and purposes beyond those that have been explored in this article. The aim of presenting this information was primarily to act as an intermediary between the reader and causal DAGs to encourage further engagement with, and in-depth study of, these diagrams and the science of causal inference. As such, individuals interested in applying these concepts are strongly advised to seek additional resources, engage with more detailed academic literature, and, if possible, consult with experts in the field to ensure accurate and responsible application of these methods in their research endeavours. For readers seeking a more comprehensive understanding of the science of causal inference and the utilisation of DAGs for causal research, the following seminal texts on this topic are recommended [ 24 , 25 , 26 ].

5 A Call for Injury-Specific Causal Diagrams (Where Necessary)

Considering the value of frameworks and models to the research process, it is unsurprising that a series of conceptual frameworks and models for athletic injury have been presented within the literature [ 36 , 42 , 43 , 59 , 60 , 61 ], with some notable examples including the comprehensive model for injury causation presented by Bahr and Krosshaug [ 36 ], the Edwards framework for modelling overuse injury as a mechanical fatigue phenomenon [ 43 ], the Kalkhoven framework for stress-related, strain-related and overuse athletic injury [ 42 ] and the Bolling model for contextual factors [ 60 ]. While these models and frameworks are certainly useful for the organisation of ideas surrounding athletic injury, and for communicating key concepts of relevance regarding athletic injury causality, these frameworks and models are generic in nature, addressing all injuries within a singular framework or model. Accordingly, many of these diagrams do not address the unique mechanisms and circumstances that contribute to different types of injury. Given the multifaceted and intricate nature of specific injury types, a more tailored approach to athletic injury research and prevention appears increasingly necessary.

To address this gap, and where researchers consider it necessary, the development of causal frameworks and models, including causal DAGs, that are tailored to specific injuries and sporting contexts, is encouraged. This will facilitate a more nuanced analysis and understanding of the complex causal networks underpinning specific types of athletic injury, offering a foundation for the development of more targeted injury prevention strategies. Indeed, to reliably intervene across varying contexts, an understanding of the specific causal effects, and ideally the pathways and mechanisms, of relevant variables contributing to athletic injury occurrence is needed. This necessitates the development of a coherent causal model that explicitly outlines all causal assumptions, thereby allowing these assumptions to be subject to evaluation. The construction of such models may prove particularly valuable in the context of complex injury events, which may occur in many different ways across different sports, and potentially within the same sport, indicating multiple sufficient causal sets (Table  1 ) for a given injury type.

6 Conclusions

Causal knowledge of athletic injury provides the critical foundations on which appropriate injury prevention strategies should be developed. However, acquiring causal knowledge is challenging. To assist, it is recommended that athletic injury research and prevention efforts should shift their attention towards the formation, utilisation, investigation and, where appropriate, revision or replacement of theoretically sound and evidence-informed causal diagrams, including frameworks, models and causal DAGs. The adoption of these tools will assist the research process by shifting athletic injury research away from a predominantly model-blind observational approach at high risk of bias, and towards a more sophisticated analysis of complex causal networks. Specifically, the adoption of causal diagrams will help organise key ideas and concepts surrounding athletic injury causation within a clear causal structure, opening up new avenues for the investigation of specific causal links and assumptions with appropriate scientific methods; leading to a more accurate and comprehensive understanding of the mechanisms underpinning sports injury occurrence. The adoption of causal DAGs in particular will provide much needed transparency regarding the causal assumptions of investigators (which are too often undisclosed and potentially even unknown by the investigators themselves) and will assist with the acquisition of causal effects from observational research. By enhancing understandings of injury causality, causal diagrams will also better facilitate the formation of appropriate athletic injury prevention strategies for utilisation in the applied sporting world. Such strategies should ideally be grounded in causal and mechanistic reasoning. Finally, to advance our field it is recommended that athletic injury researchers, and all of sports science and medicine, should engage more closely with the growing science of causal inference, for which seminal texts exist [ 24 , 25 , 26 ].

Hägglund M, Waldén M, Ekstrand J. Previous injury as a risk factor for injury in elite football: a prospective study over two consecutive seasons. Br J Sports Med. 2006;40(9):767–72.

Article   PubMed Central   PubMed   Google Scholar  

Fulton J, Wright K, Kelly M, Zebrosky B, Zanis M, Drvol C, et al. Injury risk is altered by previous injury: a systematic review of the literature and presentation of causative neuromuscular factors. Int J Sports Phys Ther. 2014;9(5):583–95.

PubMed Central   PubMed   Google Scholar  

Hamilton GM, Meeuwisse WH, Emery CA, Steele RJ, Shrier I. Past injury as a risk factor: an illustrative example where appearances are deceiving. Am J Epidemiol. 2011;173(8):941–8.

Article   PubMed   Google Scholar  

Lu D, McCall A, Jones M, Kovalchik S, Steinweg J, Gelis L, et al. Injury epidemiology in Australian male professional soccer. J Sci Med Sport. 2020;23(6):574–9.

Aoki H, O’Hata N, Kohno T, Morikawa T, Seki J. A 15-year prospective epidemiological account of acute traumatic injuries during official professional soccer league matches in Japan. Am J Sports Med. 2012;40(5):1006–14.

Ekstrand J, Hagglund M, Walden M. Injury incidence and injury patterns in professional football: the UEFA injury study. Br J Sports Med. 2011;45(7):553–8.

Article   CAS   PubMed   Google Scholar  

Ekstrand J, Walden M, Hagglund M. Hamstring injuries have increased by 4% annually in men’s professional football, since 2001: a 13-year longitudinal analysis of the UEFA Elite Club Injury Study. Br J Sports Med. 2016;50(12):731–7.

Ekstrand J, Bengtsson H, Walden M, Davison M, Khan KM, Hagglund M. Hamstring injury rates have increased during recent seasons and now constitute 24% of all injuries in men’s professional football: the UEFA Elite Club Injury Study from 2001/02 to 2021/22. Br J Sports Med. 2022;57(5):292–8.

Kalkhoven JT, Watsford ML, Coutts AJ, Edwards WB, Impellizzeri FM. Training load and injury: causal pathways and future directions. Sports Med. 2021;51(6):1137–50.

Impellizzeri FM, Woodcock S, Coutts AJ, Fanchini M, McCall A, Vigotsky AD. What role do chronic workloads play in the acute to chronic workload ratio? Time to dismiss ACWR and its underlying theory. Sports Med. 2021;51(3):581–92.

Schmida EA, Wille CM, Stiffler-Joachim MR, Kliethermes SA, Heiderscheit BC. Vertical loading rate is not associated with running injury, regardless of calculation method. Med Sci Sports Exerc. 2022;54(8):1382–8.

van Mechelen W, Hlobil H, Kemper HC. Incidence, severity, aetiology and prevention of sports injuries. A review of concepts. Sports Med. 1992;14(2):82–99.

Okasha S. Philosophy of science: very short introduction. Oxford: Oxford University Press; 2016.

Book   Google Scholar  

Fried EI. Theories and models: what they are, what they are for, and what they are about. Psychol Inq. 2020;31(4):336–44.

Article   Google Scholar  

Suppe F. The structure of scientific theories. Urbana: University of Illinois Press; 1977.

Google Scholar  

Andrade C. HARKing, cherry-picking, P-hacking, fishing expeditions, and data dredging and mining as questionable research practices. J Clin Psychiatry. 2021;82(1): 20f13804.

Head ML, Holman L, Lanfear R, Kahn AT, Jennions MD. The extent and consequences of p-hacking in science. PLoS Biol. 2015;13(3): e1002106.

Kerr NL. HARKing: hypothesizing after the results are known. Pers Soc Psychol Rev. 1998;2(3):196–217.

Murphy KR, Aguinis H. HARKing: how badly can cherry-picking and question trolling produce bias in published results? Key topics in psychological methods. Berlin: Springer; 2022. p. 93–109.

Eisenhart M, Eisenhart MA. Conceptual frameworks for research circa 1991: ideas from a cultural anthropologist; implications for mathematics education researchers. In: Underhill RG, editor. Proceedings of the 13th annual meeting of the North American chapter of the psychology of mathematics education, Blacksburg; 1991. p. 202–219.

Smaldino PE. How to translate a verbal theory into a formal model. Soc Psychol. 2020;51:207–18.

Shrier I, Platt RW. Reducing bias through directed acyclic graphs. BMC Med Res Methodol. 2008;30(8):70.

Williams TC, Bach CC, Matthiesen NB, Henriksen TB, Gagliardi L. Directed acyclic graphs: a tool for causal studies in paediatrics. Pediatr Res. 2018;84(4):487–93.

Pearl J. Models, reasoning and inference. Cambridge: Cambridge University Press; 2000.

Pearl J, Mackenzie D. The book of why: the new science of cause and effect. New York: Basic Books; 2018.

Hernán MA, Robins JM. Causal inference: what if. Boca Raton: Chapman and Hall/CRC; 2020.

McLean S, Kerherve HA, Stevens N, Salmon PM. A systems analysis critique of sport-science research. Int J Sports Physiol Perform. 2021;27:1–8.

Richardson TS, Robins JM. Single world intervention graphs (SWIGs): a unification of the counterfactual and graphical approaches to causality. Center for the Statistics and the Social Sciences, University of Washington Series Working Paper. 2013;128(30):2013.

Robins JM, Richardson TS. Alternative graphical causal models and the identification of direct effects. Causal Psychopathol. 2010;84:103–58.

Parascandola M, Weed DL. Causation in epidemiology. J Epidemiol Community Health. 2001;55(12):905–12.

Article   CAS   PubMed Central   PubMed   Google Scholar  

Rothman KJ. Causes. Am J Epidemiol. 1976;104(6):587–92.

VanderWeele TJ. Invited commentary: the continuing need for the sufficient cause model today. Am J Epidemiol. 2017;185(11):1041–3.

VanderWeele TJ, Shrier I. Sufficient cause representation of the four-way decomposition for mediation and interaction. Epidemiology. 2016;27(5):e32–3.

Bahr R, Kannus P, Van Mechelen W. Epidemiology and prevention of sports injuries. Textbook of sports medicine: basic science and clinical aspects of sports injury and physical activity; 2003. p. 299–314.

Finch C. A new framework for research leading to sports injury prevention. J Sci Med Sport. 2006;9(1–2):3–9 ( discussion 10 ).

Bahr R, Krosshaug T. Understanding injury mechanisms: a key component of preventing injuries in sport. Br J Sports Med. 2005;39(6):324–9.

Bothwell LE, Greene JA, Podolsky SH, Jones DS. Assessing the gold standard—lessons from the history of RCTs. N Engl J Med. 2016;374(22):2175–81.

Cartwright N. Are RCTs the gold standard? BioSocieties. 2007;2(1):11–20.

Westreich D, Edwards JK, Lesko CR, Cole SR, Stuart EA. Target validity and the hierarchy of study designs. Am J Epidemiol. 2019;188(2):438–43.

Altman N, Krzywinski M. Association, correlation and causation. Nat Methods. 2015;12(10):899–900.

Bahr R, Clarsen B, Derman W, Dvorak J, Emery CA, Finch CF, et al. International Olympic Committee consensus statement: methods for recording and reporting of epidemiological data on injury and illness in sport 2020 (including STROBE Extension for Sport Injury and Illness Surveillance (STROBE-SIIS)). Br J Sports Med. 2020;54(7):372–89.

Kalkhoven JT, Watsford ML, Impellizzeri FM. A conceptual model and detailed framework for stress-related, strain-related, and overuse athletic injury. J Sci Med Sport. 2020;23(8):726–34.

Edwards WB. Modeling overuse injuries in sport as a mechanical fatigue phenomenon. Exerc Sport Sci Rev. 2018;46(4):224–31.

McNally T, Edwards S, Halaki M, O’Dwyer N, Pizzari T, Blyton S. Quantifying demands on the hamstrings during high-speed running: a systematic review and meta-analysis. Scand J Med Sci Sports. 2023;33(12):2423–43.

Chumanov ES, Heiderscheit BC, Thelen DG. The effect of speed and influence of individual muscles on hamstring mechanics during the swing phase of sprinting. J Biomech. 2007;40(16):3555–62.

Valle X, Malliaropoulos N, Parraga Botero JD, Bikos G, Pruna R, Monaco M, et al. Hamstring and other thigh injuries in children and young athletes. Scand J Med Sci Sports. 2018;28(12):2630–7.

McGregor SL. Understanding and evaluating research: a critical guide. Thousand Oaks: Sage Publications; 2017.

Casanave CP, Li Y. Novices’ struggles with conceptual and theoretical framing in writing dissertations and papers for publication. Publications. 2015;3(2):104–19.

Matijevich ES, Scott LR, Volgyesi P, Derry KH, Zelik KE. Combining wearable sensor signals, machine learning and biomechanics to estimate tibial bone force and damage during running. Hum Mov Sci. 2020;22(74): 102690.

Snoep JD, Morabia A, Hernandez-Diaz S, Hernan MA, Vandenbroucke JP. Commentary: A structural approach to Berkson’s fallacy and a guide to a history of opinions about it. Int J Epidemiol. 2014;43(2):515–21.

Hernandez-Diaz S, Schisterman EF, Hernan MA. The birth weight “paradox” uncovered? Am J Epidemiol. 2006;164(11):1115–20.

Banack HR, Kaufman JS. The obesity paradox: understanding the effect of obesity on mortality among individuals with cardiovascular disease. Prev Med. 2014;62:96–102.

Hernan MA, Clayton D, Keiding N. The Simpson’s paradox unraveled. Int J Epidemiol. 2011;40(3):780–5.

Delgado-Abellan L, Aguado X, Jimenez-Ormeno E, Mecerreyes L, Alegre LM. Foot morphology in Spanish school children according to sex and age. Ergonomics. 2014;57(5):787–97.

Hernan MA, Alonso A, Logan R, Grodstein F, Michels KB, Willett WC, et al. Observational studies analyzed like randomized experiments: an application to postmenopausal hormone therapy and coronary heart disease. Epidemiology. 2008;19(6):766–79.

Lajous M, Bijon A, Fagherazzi G, Boutron-Ruault MC, Balkau B, Clavel-Chapelon F, et al. Body mass index, diabetes, and mortality in French women: explaining away a “paradox.” Epidemiology. 2014;25(1):10–4.

Nielsen RO, Shrier I, Casals M, Nettel-Aguirre A, Moller M, Bolling C, et al. Statement on methods in sport injury research from the first METHODS MATTER meeting, Copenhagen, 2019. J Orthop Sports Phys Ther. 2020;50(5):226–33.

Suzuki E, Shinozaki T, Yamamoto E. Causal diagrams: pitfalls and tips. J Epidemiol. 2020;30(4):153–62.

Windt J, Gabbett TJ. How do training and competition workloads relate to injury? The workload-injury aetiology model. Br J Sports Med. 2017;51(5):428–35.

Bolling C, van Mechelen W, Pasman HR, Verhagen E. Context matters: revisiting the first step of the ‘sequence of prevention’ of sports injuries. Sports Med. 2018;48(10):2227–34.

Bertelsen ML, Hulme A, Petersen J, Brund RK, Sorensen H, Finch CF, et al. A framework for the etiology of running-related injuries. Scand J Med Sci Sports. 2017;27(11):1170–80.

VanderWeele TJ, Hernan MA. Results on differential and dependent measurement error of the exposure and the outcome using signed directed acyclic graphs. Am J Epidemiol. 2012;175(12):1303–10.

VanderWeele TJ, Hernan MA, Robins JM. Causal directed acyclic graphs and the direction of unmeasured confounding bias. Epidemiology. 2008;19(5):720–8.

Download references

Acknowledgements

The author would like to thank Franco Impellizzeri for his mentorship and guidance, without which this article would not exist. The author would also like to thank Brook Kalkhoven and Mathias Lukauskis-Carvajal for proofreading this article.

Open Access funding enabled and organized by CAUL and its Member Institutions.

Author information

Authors and affiliations.

School of Health Sciences, Western Sydney University, Campbelltown, NSW, Australia

Judd T. Kalkhoven

Human Performance Research Centre, Faculty of Health, University of Technology Sydney, Sydney, NSW, Australia

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Judd T. Kalkhoven .

Ethics declarations

No sources of funding were used to assist in the preparation of this article.

Conflict of interest

The author has no conflicts of interest relevant to the content of this article.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Kalkhoven, J.T. Athletic Injury Research: Frameworks, Models and the Need for Causal Knowledge. Sports Med 54 , 1121–1137 (2024). https://doi.org/10.1007/s40279-024-02008-1

Download citation

Accepted : 22 February 2024

Published : 20 March 2024

Issue Date : May 2024

DOI : https://doi.org/10.1007/s40279-024-02008-1

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Find a journal
• Publish with us
• Track your research

• Open access
• Published: 05 December 2018

Overuse injuries in sport: a comprehensive overview

• R. Aicale 1 ,
• D. Tarantino 1 &
• N. Maffulli 1 , 2  

Journal of Orthopaedic Surgery and Research volume  13 , Article number:  309 ( 2018 ) Cite this article

32k Accesses

145 Citations

389 Altmetric

Metrics details

The absence of a single, identifiable traumatic cause has been traditionally used as a definition for a causative factor of overuse injury. Excessive loading, insufficient recovery, and underpreparedness can increase injury risk by exposing athletes to relatively large changes in load. The musculoskeletal system, if subjected to excessive stress, can suffer from various types of overuse injuries which may affect the bone, muscles, tendons, and ligaments.

We performed a search (up to March 2018) in the PubMed and Scopus electronic databases to identify the available scientific articles about the pathophysiology and the incidence of overuse sport injuries. For the purposes of our review, we used several combinations of the following keywords: overuse, injury, tendon, tendinopathy, stress fracture, stress reaction, and juvenile osteochondritis dissecans.

Overuse tendinopathy induces in the tendon pain and swelling with associated decreased tolerance to exercise and various types of tendon degeneration. Poor training technique and a variety of risk factors may predispose athletes to stress reactions that may be interpreted as possible precursors of stress fractures. A frequent cause of pain in adolescents is juvenile osteochondritis dissecans (JOCD), which is characterized by delamination and localized necrosis of the subchondral bone, with or without the involvement of articular cartilage. The purpose of this compressive review is to give an overview of overuse injuries in sport by describing the theoretical foundations of these conditions that may predispose to the development of tendinopathy, stress fractures, stress reactions, and juvenile osteochondritis dissecans and the implication that these pathologies may have in their management.

Conclusions

Further research is required to improve our knowledge on tendon and bone healing, enabling specific treatment strategies to be developed for the management of overuse injuries.

The specific definition of overuse injury was most commonly based on the concept of an injury occurring in the absence of a single, identifiable traumatic cause [ 1 ]. Professional soccer players sustain on average 2.0 injuries per season, which cause them to miss 37 days in a 300-day season on average [ 2 ]. Following the updated injury etiology model, training and match load contribute, together with intrinsic and extrinsic risk factors, to the multifactorial and dynamic etiology of injury [ 3 ]. Not only excessive loading and insufficient recovery, but also underpreparedness may increase injury risk by exposing players to relatively large changes, or spikes, in load during periods with higher training and match loads [ 4 ].

The tendons transfer the force produced from muscular contraction to the bone. In most instances, sports-related tendinopathies present well-defined histopathological lesions, providing an explanation for the chronicity of symptoms which often occur in athletes with tendinopathies [ 5 , 6 , 7 , 8 ].

The aim of the present article is to investigate the physiopathology, clinical presentation, diagnostic tools, and management of the most common overuse sport injuries. In particular, we focus on tendinopathy, stress reaction, stress fracture, and juvenile osteochondritis dissecans, which are the most frequent lesion caused by overuse. Furthermore, in the first part of this study, to better understand the changes of the bone, muscle, and tendon structures, we mention different mechanisms present in an overuse situation.

Overuse tendinopathy induces in the affected tendon pain and swelling, and associated decreased load tolerance and function during exercise of the limb [ 9 , 10 ]. Various types of tendon degeneration have been described at electron microscopy, namely (a) hypoxic degeneration, (b) hyaline degeneration, (c) mucoid or myxoid degeneration, (d) fibrinoid degeneration, (e) lipoid degeneration, (f) calcification, and (g) fibrocartilaginous and bony metaplasia [ 11 ]. Healing of tendinopathic tendons relies on the intrinsic ability of tenocytes to respond to the stimulus induced by the injury to the surrounding tissue matrix [ 12 , 13 ] and consists of a cellular response including apoptosis (programmed cell death), chemotaxis, proliferation, and differentiation [ 14 ]. The mechanism underlying the precise sequence of these events, which balance the effectiveness of healing and any subsequent predisposition to repetitive damage, remains obscure.

The essence of tendinopathy is a “failed healing response.” This model suggests that, after an acute insult to the tendon, an early inflammatory response that would normally result in successful injury resolution veers toward an ineffective healing response [ 15 ], with degeneration and proliferation of tenocytes, disruption of collagen fibers, and subsequent increase in non-collagenous matrix [ 7 , 16 , 17 , 18 , 19 ] (Table  1 ).

Poor training technique and a variety of risk factors may predispose players to lower limb overuse injuries affecting the bone, including stress reactions to full-fledged stress fractures. The underlying principle of the bone response to stress is Wolff’s law, whereby changes in the stresses imposed on the bone lead to changes in its internal architecture [ 20 , 21 ]. Stress fractures, defined as microfractures of the cortical bone tissue, affect thousands of athletes per year [ 22 , 23 ]. Certain subpopulations, including runners, gymnasts, and female athletes, exhibit higher rates of stress fractures [ 24 , 25 ]. If left untreated, a stress fracture can progress to a complete fracture of a bone, which may require surgical fixation [ 26 ]. In addition, factors contributing to stress fractures increase the risk for osteoporosis, a substantial long-term health concern [ 27 ].

Stress reactions of the musculoskeletal system may be interpreted as possible precursors of stress fractures. Biological tissues, in contrast to artificial products, can react in numerous and complex ways. This can lead not only to a continual weakening of the tissue, but also to adaption phenomena in response to overuse. The causes of such stress reactions are still unclear.

Juvenile osteochondritis dissecans (JOCD) is a frequent cause of pain in adolescents, both athletes and non-athletes. JOCD is characterized by delamination and localized necrosis of the subchondral bone, with or without the involvement of the overlying articular cartilage [ 28 , 29 , 30 , 31 ]. The etiology remains unclear, but repetitive microtrauma, such as that typical of overuse injury, is considered the significant factor leading to JOCD [ 28 , 29 ].

The role of inflammation and molecular factors in overuse injuries

The effects of an altered inflammatory response.

In this model, the question arises as to why the healing response is successful in some individuals but fails in others. More importantly, can we identify factors which may increase the risk of this ineffective healing response? For example, the incidence of tendinopathy is increased in individuals with obesity and decreased insulin sensitivity, as seen in patients with type 1 and type 2 diabetes mellitus (T1/T2DM) [ 10 , 32 , 33 , 34 ]. Evidence for a chronic, low-grade inflammatory state in obesity is represented principally by marked increases in plasma levels of proinflammatory cytokines such as tumor necrosis factor (TNF)-a and interleukin (IL)-6, and proinflammatory chemokines such as monocyte chemoattractant protein (MCP)-1 [ 34 ].

Patients with type 1 and type 2 diabetes exhibit a less effective healing response [ 35 ]. Recently, it has been demonstrated the presence of an independent relationship between impaired insulin sensitivity and the development of chronic low-grade inflammation through a protein, the levels of which are normally physiologically inhibited by insulin, called FOXO1, a key upregulator of the proinflammatory cytokine IL-b [ 35 ].

Considering the influence that a prolonged state of low-grade systemic inflammation may have on the healing process after acute tendon injury, it must be appreciated that tendon healing is a delicate and prolonged process even under optimal physiological conditions [ 10 , 34 ]. Even minor disruptions to any of the noted healing stages could result in a much more prolonged and complicated resolution of injury. Similarly, if several minor disruptions to this process occur (in the form, for example, of microtraumas), complete healing and resolution of injury become progressively unlikely [ 10 ].

The acute inflammatory phase noted in the first few days after a tendon injury is marked by the migration of inflammatory cells such as macrophages and monocytes [ 36 ]. As the chronic inflammatory state in obesity is associated with a reduction in the numbers of circulating macrophages [ 33 ], such a decrease in the availability of circulating cells may result in the mounting of a less effective early healing response.

Such findings are consistent with a post-injury state of “failed healing,” in which evidence of matrix disorganization, increased amounts of extracellular ground substance, and a degree of separation between collagen fibers has been noted [ 37 , 38 ], with associated greater vulnerability to future mechanical strain [ 10 ].

This relationship may help to explain the influence that mechanical overuse plays in the development of tendinopathy. Examining the incidence of tendinopathy among patients with type 2 diabetes, unilateral or bilateral tendinopathy was found in 32% of the diabetic patients studied versus 10% of controls [ 39 ]. Also, when the incidence of unilateral tendinopathy among diabetic patients was examined more closely, 45% were found to occur in the right shoulder compared with just 27% in the left shoulder [ 33 ].

Molecular factors in overuse injury

A lack of exposure to adequate levels of physiological stress over a prolonged time period or “underloading” may paradoxically predispose to overload injury [ 34 ]. An underloaded tendon may become unable to cope with increased demands imposed on it. Thus, underuse of a tendon may result in an imbalance between matrix metalloproteinases and their inhibitors (tissue inhibitors of matrix metalloproteinases), with resultant tendon degradation [ 34 ].

Molecular agents may link the events of tendon degeneration and ineffective tendon healing with the production and persistence of reactive oxygen species within both the intra- and extra-cellular milieu of the tendon tissue [ 40 ]. The reactive oxygen production is strongly influenced by lifestyle factors, e.g., nutrition and the intensity and frequency of exercise.

The term reactive oxygen species (ROS) encompasses reactive species derived from oxygen. A free radical is any species capable of independent existence that contains one or more unpaired electrons [ 41 ]. Physiologically relevant ROS include the superoxide anion (O 2− ), hydrogen peroxide (H 2 O 2 ), hydroxyl radical (HO − ), singlet oxygen ( 1 O 2 ), and peroxyl radicals (RO 2− ). An inter-related group of radical and non-radical reactive species is the reactive nitrogen species (RNS) [ 42 , 43 , 44 ].

The principal site for ROS formation in non-stressed cells is the mitochondrial respiratory chain. This series of coupled redox reactions leads to the formation of ATP with molecular oxygen the ultimate electron acceptor and being reduced to water [ 45 , 46 ].

ROS may mediate processes of cell proliferation [ 47 ], differentiation [ 48 ], and adaptive responses [ 45 , 49 ]. At higher levels, ROS may initiate and/or execute the demise of the cell through programmed as well as necrotic cell death mechanisms [ 50 , 51 , 52 ].

Traditionally, ROS are viewed as imposing cellular and tissue damage via lipid peroxidation, DNA damage, and protein modification [ 45 , 53 ]. However, in themselves, O 2− and H 2 O 2 are not potent biological oxidizers [ 45 ], although certain proteins may be prone to direct modification by these species. ROS production is involved in various cancers (e.g., the lung, colon), coronary heart disease, autoimmune diseases, etc. [ 41 ]. Furthermore, ROS are implicated in overuse exercise-related damage in muscle [ 49 ] and may impair fracture healing in the bone [ 54 ].

Cells and tissues contain many antioxidant molecules, but many antioxidants are also capable of acting as pro-oxidants [ 55 ]. For example, ascorbic acid in the presence of iron/copper generates HO − , and with flavonoids may generate O 2− [ 55 ].

ROS may be involved in tendinopathies or other stress reactions: indeed, synthesis, structure, and integrity of connective tissues are influenced by them. Reactive species may be produced within the intra- and extra-tendinous environment. Evidence suggest that ROS constitute a stress factor during not-hard exercise [ 49 ]. Excessive exercise induces elevated ROS production, primarily from mitochondria [ 56 ]. Exercise also stimulates the immune response [ 56 , 57 , 58 ], with increased leucocyte numbers, in particular granulocytes. Exhaustive exercise in cross-country skiers produced neutrophil mobilization and increased ROS generation on subsequent stimulation. Enhanced phagocytic O 2− generation accurs approximately 24 h after exhaustive exercise [ 59 ].

Increased phagocyte activity probably does not contribute to elevated ROS production during short-term exercise, but may act as a secondary source of ROS during recovery from heavy exercise [ 59 ].

During cyclical tendon loading, the period of maximum tensile load is associated with ischemia, and relaxation with reperfusion. This restoration of normal tissue oxygenation may lead to enhanced ROS production [ 59 , 60 ]. There is potential for re-oxygenation resulting in a cycle of enhanced ROS production, most probably at sublethal levels within the non-degenerate tendon [ 40 ].

Hyperthermia is a feature of tendon use inducing ROS production. During exercise, the central core temperature of the muscles can exceed 47 °C [ 49 ], a temperature resulting in increased ROS production in mitochondria [ 49 ]. Similarly, in exercising the tendon, core temperatures may reach 45 °C, contributing to their damage [ 60 ].

During fibrogenesis, ROS, primarily derived from specialized phagocytes and products arising from lipid peroxidation, induce overexpression of fibrogenic cytokines and increase the synthesis of collagen [ 61 ]. Endogenous and exogenous ROS may also exert effects on tenocyte proliferation, development, and viability, with implications on both tendinopathy and post-rupture healing [ 62 ]. Tenocytes are motile and highly proliferative and rapidly increase in number following injury [ 62 ].

Tenocyte numbers are altered in degenerated tendons, and the selective deletion of tenocytes from damaged tendon may be a factor in degeneration, but also a prerequisite to healing. Heightened levels of ROS production may not only induce cell death, but also determine the mechanistic form of that death, in particular, the ratio of programed cell death (PCD): necrosis [ 51 ].

High concentrations of H 2 O 2 can prevent apoptosis. Conversely, “bursts” of ROS [ 51 , 63 ] and reductions in antioxidant enzyme activity [ 64 ] frequently accompany the induction of apoptosis, and oxidative stress is a common feature of the late phase of apoptosis [ 51 ]. For example, the pro-apoptotic transcription factor p53 demonstrates impaired DNA binding following exposure to ROS. However, it can induce apoptosis by induction of enhanced mitochondrial ROS generation [ 65 ].

Overuse sport injuries

Tendinopathy: consideration and management.

Tendinopathy has been hypothesized to result from inflammatory changes in the tendon, and secondary to its frequent or excessive use, assigning the label of “tendinitis” or “tendonitis” to such a presentation [ 9 , 10 , 66 ]. However, anti-inflammatory agents are largely unsuccessful in the treatment of the condition [ 15 , 66 ], and with the increase in histopathological data showing degenerative changes but little inflammation, the inflammatory hypothesis in overuse tendon injury became decreasingly popular [ 10 , 15 , 36 , 67 ]. The term “tendonitis” became increasingly replaced by “tendinosis” [ 36 ], but a definitive diagnosis of either should only be made following histopathological confirmation [ 15 , 36 , 67 ].

However, it became evident that tendon biopsies from operated patients were likely to represent the end stage of a pathological continuum [ 10 ], probably demonstrating a different histopathological picture to that which would be seen in the initial stages of injury [ 36 , 67 ]. This was supported by evidence from human and animal biopsies that showed that both peritendinitis and a failed healing response, wrongly labeled “tendinosis,” could be present concurrently [ 36 ].

In tendinopathic lesions, the parallel orientation of collagen fibers is lost, with a decrease in collagen fiber diameter and in the overall density of collagen. Collagen microtears may also occur and may be surrounded by erythrocytes, fibrin, and fibronectin deposits. Normally, collagen fibers in tendons are tightly bundled in a parallel fashion. In tendinopathic samples, there is unequal and irregular crimping, loosening, and increased waviness of collagen fibers, with an increase in type III (reparative) collagen [ 17 , 19 , 68 , 69 , 70 , 71 ]. Vascularity is typically increased, and blood vessels are randomly oriented, sometimes perpendicular to collagen fibers [ 72 , 73 ]. Inflammatory lesions [ 72 , 73 , 74 , 75 ] and granulation tissue [ 74 , 75 ] are infrequent and, when found, are associated with partial rupture: therefore, tenocytes are abnormally plentiful in some areas [ 76 , 77 ].

Tendinopathies are common in elite and recreational athletes and are traditionally considered overuse injuries, involving excessive tensile loading and subsequent breakdown of the loaded tendon [ 78 , 79 ]. Although acute traumatic conditions such as ligament and muscle tears receive much attention in the lay press, tendinopathies account for much of the lost time in practice and competition [ 80 , 81 ].

Biopsy studies have shown that classic inflammatory changes are not frequently seen in chronic tendon conditions and that histopathology features in tendinopathic tendons are clearly different from normal tendons [ 82 , 83 ].

All tendons can develop tendinopathy [ 5 , 84 ]. The supraspinatus, common wrist extensor, quadriceps, patellar, posterior tibialis, and Achilles tendons are probably the most commonly affected tendons. Insertional tendinopathy is one of the most common forms of tendinopathy, and, in particular, the supraspinatus, common wrist extensor, quadriceps, and patellar tendons are most affected by it [ 84 ]. The Achilles tendon, on the other hand, can present tendinopathy of the main body of the tendon, paratendinopathy and insertional tendinopathy, each with different clinical features and management implications [ 84 ].

Achilles tendinopathy (AT) is a common overuse injury among athletes, with an increasing incidence over the past 30 years [ 85 , 86 ]. AT is particularly prevalent in athletes whose sport involve running and jumping activities [ 87 , 88 ] and is thus common in sports such as soccer. In the four principal soccer leagues in England, there are an average of 3.5 Achilles tendon-related injuries per week in the preseason and an average of one injury per week in the competitive season [ 89 , 90 ].

Tendinopathies may result from excessive loading of the tendon and subsequent mechanical breakdown of the loaded tendon [ 91 ]. Theoretically, repeated microinjuries may occur, and the tendon may be able to heal a certain level of microinjury. However, as training and heavy loading of the tendon continues, this healing process may be overwhelmed, and a further injury ensues.

Other factors in addition to training errors may lead to increased loading of the tendon, such as poor technique [ 92 , 93 ] or inadequate athletic equipment [ 94 ]. Also, intrinsic factors, such as the status of the muscles, ligaments, and bones surrounding the tendon, may alter the level of the load on the tendon [ 19 ]. Recent biomechanical studies about failure modes of the muscle-tendon units have shown that failure occurs within the muscle near the muscle-tendon junction [ 95 , 96 ].

Relatively, little is known about the role of neuronal regulation in tendinopathy, and the source of pain has not been clarified yet [ 97 ]. The presence of pain in tendinopathy requires not only mechanical changes, but also alterations in the way the local cells and the peripheral nerves react to this change. A recent systematic review showed that the peripheral neuronal phenotype is altered in tendinopathy [ 97 ] and that the peripheral and central pain processing pathways are important factors in the pathogenesis of painful human tendinopathy. Changes in the peripheral neuronal phenotype may be the primary source of pain [ 97 ] .

Clinical history and examination are essential for diagnosis. Clinically, tendinopathy is characterized by pain, swelling (diffuse or localized), and impaired performance [ 6 ].

Pain is the cardinal symptom, and it occurs at the beginning and a short while after the end of a training session. As the pathological process progresses, pain may occur during the entire exercise session, and, in severe cases, it may interfere with the activities of daily living. Clinical examination is the best diagnostic tool. In tendinopathy of the main body of the Achilles tendon, the location of pain is 2–6 cm above the insertion into the calcaneum, and pain on palpation is a reliable and accurate test for diagnosis [ 98 ].

In addition to the swelling on the posteromedial aspect of the tendon and palpation pain, some clinical tests have been described for non-insertional AT diagnosis. They can be divided into palpation tests (tendon thickening, crepitus, pain on palpation, the Royal London Hospital (RLH) test, the painful arc sign) and tendon loading tests (pain on passive dorsiflexion, pain on single heel raise, and pain on hopping).

Plain radiography can be used to diagnose associated or incidental bony abnormalities [ 99 ].

Ultrasound is an effective imaging method since it correlates well with the histopathologic findings despite being an operator-dependent [ 100 ]. MRI studies should be performed only if the ultrasound scan remains unclear. The ultrasound (US) signs of hypoechoic areas, spindle-shaped thickening, neovascularization, and paratenon blurring [ 101 ] are associated with AT [ 102 ] and may be potential predictors of future tendinopathy [ 87 ] when present in asymptomatic individuals.

The first line of management for AT is conservative, and different treatments such as nonsteroidal anti-inflammatory drugs, physical therapy, taping, cryotherapy, shock wave therapy, hyperthermia, and various peritendinous injections have been used with varying success [ 103 ]. The management of AT lacks strong evidence-based support, because few treatment modalities have been investigated in randomized controlled trials [ 103 ], and approximately 25% of patients do not respond to conservative management [ 104 ]. Good results have been reported with eccentric exercises [ 105 , 106 ], but these alone may not work in all patients [ 107 ], and their mechanism of action is not completely understood [ 106 ]. These are the most effective conservative treatment for non-insertional AT. The most commonly used protocol is the Alfredson’s protocol: the exercises are performed in three sets of 15 repetitions, twice a day for 12 weeks [ 108 ].

ESWT, when compared with eccentric strengthening in a RCT, showed comparable outcomes, with 60% of the patients at least significantly improved in both of the treatment groups, and significantly better than those in the “wait and see” control group [ 109 ]. Where available, ESWT should probably be a second-line treatment.

Various injection therapies have been proposed [ 110 ]. In a recent systematic review [ 111 ], only ultrasound-guided sclerosing polidocanol injections seemed to yield promising results, but these results do not appear to have been duplicated outside Scandinavia [ 112 ]: indeed, Ebbesen et al. [ 113 ], in a RTC, concludes that polidocanol injections are a safe treatment, but in the mid-term, the effects are the same of a placebo treatment for chronic Achilles tendinopathy. The use of platelet-rich plasma (PRP) is growing exponentially, especially among sports medicine physicians, but the only well-designed RCT published on PRP in AT showed no significant difference in pain or activity level between PRP and saline injection at 6, 12, or 24 weeks when combined with an eccentric stretching program [ 114 ]. High-volume image-guided injections (HVIGI) significantly reduce pain and improve function in patients with resistant AT [ 115 ]. A recent study found relevant clinical results with the contemporaneous administration of platelet-rich plasma and high-volume image-guided injections of saline treatments, which influence tendon repair by different mechanisms and grants a greater improvement for patellar tendinopathy [ 116 ].

Conservative treatment fails in between one quarter and one third of patients, and surgical intervention is required [ 117 ]. Minimally invasive therapies which strip the paratenon from the tendon, either directly [ 118 ] or indirectly with high-volume fluid injection [ 115 ], have shown good initial results in relieving the symptoms of non-insertional AT [ 103 , 119 ].

Another technique consists in multiple percutaneous longitudinal tenotomies, which can be performed under ultrasound guidance [ 120 , 121 ]. Minimally invasive surgical treatment would appear to be a useful intermediate step between failed conservative treatment and formal open surgery [ 103 ].

The high recurrence rate (27%) for AT when managed conservatively reflects the chronic and recurrent character of this condition. The frequent relapse of symptoms when players return to football after a short rehabilitation period could be explained if the pain is only the tip of the iceberg. Therefore, it could be suggested that a longer rehabilitation period at the first signs of AT could be beneficial to avoid recurrences [ 122 ].

Stress reaction and stress fracture

Stress reactions may be interpreted as precursors of stress fractures [ 123 ]. The causes of such stress are still unclear. For example, it is unknown to what extent a predisposition to these stress symptoms by mechanical stress alone or whether other factors such as physical condition, nutrition, or even hormone balance come in to play. Early diagnosis considerably reduces the impairment of the healing process. The treatment of a stress reaction should be the same as for a diagnosed stress fracture [ 123 ]. Much of our epidemiological knowledge about stress fractures originates from research on military recruits [ 124 ] and high school athletes [ 25 ].

The response of the bone to repetitive stress is increased osteoclastic activity over osteoblastic new bone formation, which results in temporary weakening of the bone [ 125 ]. The eventual adaptive response is periosteal new bone formation to provide reinforcement [ 126 ]. However, if physical stress continues, an osteoclastic activity may predominate, resulting initially in microfractures (commonly seen as bone marrow edema on MRI, consistent with a stress reaction), and eventually, a true cortical break (stress fracture) may result [ 1 ]. If strain becomes excessive or adequate rest is not implemented, stress reaction and eventually a stress fracture can results [ 126 , 127 ].

There is a difference between stress fractures from fatigue and insufficiency type. Fatigue fractures are the typical overuse stress fractures observed in athletes and military recruits with normal bone density. They result from an imbalance in the ability of the bone to keep up with skeletal repair from an excessive bone strain with progressive accumulation of microdamage [ 126 ]. An insufficiency fracture is seen in those with low bone mineral density (BMD), such as runners with the female athlete triad; metabolic bone disease; or osteoporosis. Insufficiency fractures result from poor bone remodeling (increased resorption and depressed formation) in response to normal strain [ 126 ].

Rizzone et al. [ 128 ] investigated the epidemiology of stress fractures in 671 collegiate student-athletes for the academic years 2004–2005 through 2013–2014. The rate of stress fracture was highest among endurance athletes and higher in women than in men. Higher rates among female athletes were found not only in cross-country athletes, indoor track, and outdoor track athletes, but also in basketball and soccer athletes. Twenty-two percent of stress fractures were recurrent, and 20% resulted in season-ending injuries [ 128 ].

The number of reports in the literature of lower extremity stress fractures in female soccer athletes is small [ 129 ]. Of the 18 million Americans who play soccer, 78% are younger than 18 years and more than 40% are female [ 130 ]. Women collegiate soccer increased from 1855 athletes on 80 teams during the 1981–1982 seasons to 22,682 athletes on 956 teams during the 2007–2008 seasons, making women’s soccer the NCAA sport with the greatest number of athletes [ 131 ]. A study of 2016 [ 132 ] showed that elite female soccer athletes are susceptible to stress fractures and menstrual dysfunction and experience delayed onset of menarche despite normal BMI and appropriate body perception and attitudes toward eating. Education about the detrimental effects of menstrual dysfunction and the importance of adequate energy balance and nutritional requirements should be encouraged to minimize the risk for poor bone health, manifesting as a stress fracture in the short term and osteoporosis over the long term in these athletes [ 132 , 133 ].

The typical history of a stress fracture is localized pain of insidious onset which is initially not present at the start but occurs toward the end of a run. A sign of a more advanced fracture is pain progressing to occur during non-running-related activities, affecting day-to-day walking [ 126 ].

Defining the causative risk factors for stress fractures is difficult because there are many interrelated variables which make risk assessment problematic to study independently. Extrinsic and intrinsic factors may lead to stress fractures [ 126 , 133 , 134 ]. An increase in frequency, duration, or intensity of training load is often cited as a primary risk factor [ 135 ]. Hard training surfaces are also factors associated with lower-limb overuse injuries [ 135 ]. Training in shoes older than 6 months is a risk factor for stress fractures, likely related to the decrement in shock absorption as shoes age [ 135 ].

Regarding the intrinsic factors, Bennell et al. [ 22 ] demonstrated that smaller calf girth and less muscle mass in the lower limb of female runners was associated with a higher incidence of stress fractures. Kinematic and kinetic biomechanical variables have also been recently studied as potential risk factors for stress fractures; for example, in runners, an excessive hip adduction and rear-foot eversion are predictors of tibial stress fractures [ 136 ].

The hallmark physical examination finding is focal bony tenderness. Overlying swelling, erythema, or warmth are other potential examination findings. Less sensitive tests for fractures of long bones include the fulcrum test and hop test [ 137 ]. A functional kinetic chain assessment is useful to elucidate biomechanical factors that may predispose the runner to injury. Evaluating muscle imbalances, leg-length discrepancies, foot mechanics, genu varum, and femoral anteversion is appropriate because all have been associated with stress fractures [ 137 ].

Radiographic imaging should be used to supplement the clinical history and physical examination if uncertainty persists. Imaging can also be used to grade the severity of an injury and can thus be helpful in guiding treatment. CT is best used to differentiate lesions seen on a bone scan that may mimic stress fracture, including osteoid osteoma, osteomyelitis, and malignancy [ 138 ]. MRI is becoming the imaging study of choice, with many considering MRI the gold standard for the evaluation of bony stress injuries [ 139 ].

It is not only important to understand the significance of protection and rest, but also to understand the predisposing factors to the injury. Treatment is the time to explore and treat the contributing risk factors. For example, if low bone density is found, appropriate treatment is mandatory; if biomechanical issues are identified, and inappropriate shoes and training are determined, and specific rehabilitation is required [ 126 ].

Therapeutic ultrasound and electrical stimulation are purported modalities for enhancing the healing rate of fractures. Therapeutic ultrasound has been demonstrated to decrease healing time in acute tibial shaft, in distal radius fractures, and in navicular stress injuries [ 140 , 141 ]. Electrical stimulation for bone growth has some support in delayed unions and non-unions, but only in uncontrolled trials for stress fractures [ 142 ].

Juvenile osteochondritis dissecans (JOCD)

JOCD is a frequent cause of knee pain in adolescent athletes and non-athletes, with an incidence higher in boys than in girls [ 143 , 144 ] and with delamination and localized necrosis of the subchondral bone. The etiology remains unclear [ 28 , 29 , 31 ]. Repetitive microtrauma, such as that of overuse injury, is considered the significant factor leading to JOCD [ 28 , 29 , 145 ].

The most common site of JOCD is the medial femoral condyle, accounting for 85% of the cases [ 28 ]. The term “osteochondritis” suggests an inflammatory etiology: however, histology shows damage of the bone and cartilage with no inflammation [ 146 ]. Local bone vascular insufficiency is also postulated to contribute to JOCD [ 29 ].

Highly active athletes present with a history of aching and gradual onset of knee pain of several days to weeks duration, typically located over the anterior portion of the knee, worse during activity. There may be a history of intermittent knee effusion following a practice or game session [ 29 ].

The examination may reveal mild effusion or limitation of motion of the knee. Findings may also vary depending on the stage of the disease [ 147 ]. In the early stages, with the articular cartilage over the femoral condyle still intact, the signs are non-specific. In the later stages, when the articular cartilage is eroded, the fragment may separate and become an intra-articular loose body. This can cause pain, effusion, and locking. Typically, in lesions, the medial femoral condyle when the athlete flexes and internally rotates the leg, from full extension to about 30°, pain is elicited and is relieved upon external rotation [ 28 , 29 ].

Radiographic examination, with comparison with the other knee, is indicated when JOCD is suspected. In addition to the anteroposterior (AP) and lateral views, a tunnel view is useful to better identify the lesion, which appears as a well-demarcated radiolucent area [ 29 , 31 ]. In those who demonstrate significant edema, a hemarthrosis or discomfort, and inability to bear weight without pain, an MRI is often obtained. An MRI can be helpful in identifying unstable lesions [ 31 ].

Suzue et al. [ 148 ] investigated the prevalence of JOCD in children and adolescent soccer players using a questionnaire, distributed to 1162 players. Of these, 547 patients experienced pain in the legs or lumbar spine. Radiographic or ultrasonographic examination was performed in 106 players, and 80 (75.5%) were diagnosed JOCD. In conclusion, the majority of players who had experienced pain and were found to have osteochondritis had severe injuries such as JOCD or lumbar spondylolysis [ 148 ].

Early diagnosis followed by restriction of activities and symptomatic treatment of pain generally allows for healing of lesions over a period of 8–12 weeks [ 30 , 31 , 149 ]. Spontaneous healing of the lesion is the usual outcome in children and adolescents with open distal femoral physis. Prognosis is excellent in younger patients [ 149 ].

Treatment is based on the stability of the lesion and the status of the overlying cartilage. The lesion may be unstable or loose, and these cases as well as in those athletes with large effusions or with marked symptoms which do not improve with conservative care may go to surgery for drilling, reattachment, or excision of the osteochondral lesion [ 31 ].

Overuse injuries can affect the muscle, tendon, and bone. Tendon injuries give rise to substantial morbidity, and current understanding of the mechanisms involved in tendon injury and repair is limited. Tendon physiology and structure may include ROS involvement in various aspects of the predisposition to and participation in the degenerative process and subsequent response to injury. Bone can be damaged by repeated microtrauma and overuse. Stress reaction and stress fractures are very common in athletes, and the treatment consists in the treatment of the risk factors. Further research is required to improve our knowledge of tendon and bone healing. This will enable specific treatment strategies to be developed.

Abbreviations

Anteroposterior

Achilles tendinopathy

Bone mineral density

Food and Drug Administration

High-volume image-guided injections

Interleukin

Juvenile osteochondritis dissecans

Low-intensity pulsed ultrasound

Monocyte chemoattractant protein

Programed cell death

Pulsed electromagnetic fields

Platelet-rich plasma

Royal London Hospital

Reactive nitrogen species

Reactive oxygen species

Tumor necrosis factor

Chéron C, Le Scanff C, Leboeuf-Yde C. Association between sports type and overuse injuries of extremities in adults: a systematic review. Chiropr Man Ther. 2017;25 Available from: http://chiromt.biomedcentral.com/articles/10.1186/s12998-017-0135-1 . [cited 2017 Nov 8].

Ekstrand J, Hagglund M, Walden M. Injury incidence and injury patterns in professional football: the UEFA injury study. Br J Sports Med. 2011;45:553–8.

Article   CAS   PubMed   Google Scholar  

Windt J, Gabbett TJ. How do training and competition workloads relate to injury? The workload—injury aetiology model. Br J Sports Med. 2017;51:428–35.

Article   PubMed   Google Scholar  

Gabbett TJ, Kennelly S, Sheehan J, Hawkins R, Milsom J, King E, et al. If overuse injury is a ‘training load error’, should undertraining be viewed the same way? Br J Sports Med. 2016;50:1017–8.

Khan KM, Maffulli N. Tendinopathy: an Achilles’ heel for athletes and clinicians. Clin J Sport Med. 1998;8:151–4.

Maffulli N, Khan KM, Puddu G. Overuse tendon conditions: time to change a confusing terminology. Arthrosc J Arthrosc Relat Surg Off Publ Arthrosc Assoc N Am Int Arthrosc Assoc. 1998;14:840–3.

Article   CAS   Google Scholar  

Magra M, Maffulli N. Matrix metalloproteases: a role in overuse tendinopathies. Br J Sports Med. 2005;39:789–91.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Sharma P, Maffulli N. Biology of tendon injury: healing, modeling and remodeling. J Musculoskelet Neuronal Interact. 2006;6:181–90.

CAS   PubMed   Google Scholar  

Abate M, Silbernagel KG, Siljeholm C, Di Iorio A, De Amicis D, Salini V, et al. Pathogenesis of tendinopathies: inflammation or degeneration? Arthritis Res Ther. 2009;11:235.

Article   PubMed   PubMed Central   Google Scholar  

Rees JD, Maffulli N, Cook J. Management of tendinopathy. Am J Sports Med. 2009;37:1855–67.

Longo G, Ripalda P, Denaro V, Forriol F. Morphologic comparison of cervical, thoracic, lumbar intervertebral discs of cynomolgus monkey (Macaca fascicularis). Eur Spine J. 2006;15:1845–51.

Gigante A, Specchia N, Rapali S, Ventura A, de Palma L. Fibrillogenesis in tendon healing: an experimental study. Boll Della Soc Ital Biol Sper. 1996;72:203–10.

CAS   Google Scholar  

Sharma P, Maffulli N. Basic biology of tendon injury and healing. Surgeon. 2005;3:309–16.

Banes AJ, Tsuzaki M, Yamamoto J, Fischer T, Brigman B, Brown T, et al. Mechanoreception at the cellular level: the detection, interpretation, and diversity of responses to mechanical signals. Biochem Cell Biol Biochim Biol Cell. 1995;73:349–65.

Longo UG, Ronga M, Maffulli N. Achilles tendinopathy. Sports Med Arthrosc Rev. 2009;17:112–26.

Longo UG, Franceschi F, Ruzzini L, Rabitti C, Morini S, Maffulli N, et al. Light microscopic histology of supraspinatus tendon ruptures. Knee Surg Sports Traumatol Arthrosc. 2007;15:1390–4.

Maffulli N, Kenward MG, Testa V, Capasso G, Regine R, King JB. Clinical diagnosis of Achilles tendinopathy with tendinosis. Clin J Sport Med Off J Can Acad Sport Med. 2003;13:11–5.

Article   Google Scholar  

Kvist M, Józsa L, Järvinen M, Kvist H. Fine structural alterations in chronic Achilles paratenonitis in athletes. Pathol Res Pract. 1985;180:416–23.

Aicale R, Tarantino D, Maffulli N. Basic science of tendons. Bio-Orthop [Internet]. Springer, Berlin, Heidelberg; 2017. p. 249–73. Available from: https://link.springer.com/chapter/10.1007/978-3-662-54181-4_21 . [cited 2017 Aug 3]

Chapter   Google Scholar  

Ackerman KE, Nazem T, Chapko D, Russell M, Mendes N, Taylor AP, et al. Bone microarchitecture is impaired in adolescent amenorrheic athletes compared with eumenorrheic athletes and nonathletic controls. J Clin Endocrinol Metab. 2011;96:3123–33.

Ackerman KE, Sokoloff NC, De NM, Clarke HM, Lee H, Misra M. Fractures in relation to menstrual status and bone parameters in young athletes. Med Sci Sports Exerc. 2015;47:1577–86.

Bennell KL, Malcolm SA, Thomas SA, Wark JD, Brukner PD. The incidence and distribution of stress fractures in competitive track and field athletes: a twelve-month prospective study. Am J Sports Med. 1996;24:211–7.

Difiori JP, Benjamin HJ, Brenner JS, Gregory A, Jayanthi N, Landry GL, et al. Overuse injuries and burnout in youth sports: a position statement from the American Medical Society for Sports Medicine. Br J Sports Med. 2014;48:287–8.

Tenforde AS, Sayres LC, McCurdy ML, Collado H, Sainani KL, Fredericson M. Overuse injuries in high school runners: lifetime prevalence and prevention strategies. PM R. 2011;3:125–31.

Changstrom BG, Brou L, Khodaee M, Braund C, Comstock RD. Epidemiology of stress fracture injuries among us high school athletes, 2005–2006 through 2012–2013. Am J Sports Med. 2015;43:26–33.

Goolsby MA, Barrack MT, Nattiv A. A displaced femoral neck stress fracture in an amenorrheic adolescent female runner. Sports Health. 2012;4:352–6.

Barrack MT, Gibbs JC, De S, Williams NI, Nichols JF, Rauh MJ, et al. Higher incidence of bone stress injuries with increasing female athlete triad-related risk factors: a prospective multisite study of exercising girls and women. Am J Sports Med. 2014;42:949–58.

Peters TA, McLean LD. Osteochondritis dissecans of the patellofemoral joint. Am J Sports Med. 2000;28:63–7.

Pascual-Garrido C, Moran CJ, Green DW, Cole BJ. Osteochondritis dissecans of the knee in children and adolescents. Curr Opin Pediatr. 2013;25:46–51.

Schulz JF, Chambers HG. Juvenile osteochondritis dissecans of the knee: current concepts in diagnosis and management. Instr Course Lect. 2013;62:455–67.

PubMed   Google Scholar  

Kocher MS, Tucker R, Ganley TJ, Flynn JM. Management of osteochondritis dissecans of the knee: current concepts review. Am J Sports Med. 2006;34:1181–91.

Ramchurn N, Mashamba C, Leitch E, Arutchelvam V, Narayanan K, Weaver J, et al. Upper limb musculoskeletal abnormalities and poor metabolic control in diabetes. Eur J Intern Med. 2009;20:718–21.

Hirai S, Takahashi N, Goto T, Lin S, Uemura T, Yu R, et al. Functional food targeting the regulation of obesity-induced inflammatory responses and pathologies. Mediat Inflamm. 2010;2010:8.

Gaida JE, Ashe MC, Bass SL, Cook JL. Is adiposity an under-recognized risk factor for tendinopathy? A systematic review. Arthritis Care Res. 2009;61:840–9.

Wenner M. Inflammatory clues. Sci Am. 2009;301:24–6.

Almekinders LC, Temple JD. Etiology, diagnosis, and treatment of tendonitis: an analysis of the literature. Med Sci Sports Exerc. 1998;30:1183–90.

Lewis JS, Sandford FM. Rotator cuff tendinopathy: is there a role for polyunsaturated fatty acids and antioxidants? J Hand Ther. 2009;22:49–56.

Karousou E, Ronga M, Vigetti D, Barcolli D, Passi A, Maffulli N. Molecular interactions in extracellular matrix of tendon. Front Biosci - Elite. 2010;2 E:1–12.

Google Scholar  

Mavrikakis ME, Drimis S, Kontoyannis DA, Rasidakis A, Moulopoulou ES, Kontoyannis S. Calcific shoulder periarthritis (tendinitis) in adult onset diabetes mellitus: a controlled study. Ann Rheum Dis. 1989;48:211–4.

Longo UG, Olivia F, Denaro V, Maffulli N. Oxygen species and overuse tendinopathy in athletes. Disabil Rehabil. 2008;30:1563–71.

Nespolo M. Free radicals in biology and medicine. Fifth Edition. by Barry Halliwell and John M. C. Gutteridge. Oxford University Press, 2015. Pp. xxxviii + 905. Price GBP 70.00 (paperback, ISBN 9780198717485), GBP 125.00 (hardback, ISBN 9780198717478). Acta Crystallogr Sect Struct Biol. 2017;73:384–5.

Paoloni JA, Appleyard RC, Nelson J, Murrell GAC. Topical glyceryl trinitrate application in the treatment of chronic supraspinatus tendinopathy: a randomized, double-blinded, placebo-controlled clinical trial. Am J Sports Med. 2005;33:806–13.

Harper M-E, Bevilacqua L, Hagopian K, Weindruch R, Ramsey JJ. Ageing, oxidative stress, and mitochondrial uncoupling. Acta Physiol Scand. 2004;182:321–31.

Szomor ZL, Appleyard RC, Murrell GAC. Overexpression of nitric oxide synthases in tendon overuse. J Orthop Res. 2006;24:80–6.

Flowers F, Zimmerman JJ. Reactive oxygen species in the cellular pathophysiology of shock. New Horiz Sci Pract Acute Med. 1998;6:169–80.

Dennery PA. Introduction to serial review on the role of oxidative stress in diabetes mellitus. Free Radic Biol Med. 2006;40:1–2.

Burdon RH, Gill V, Rice-Evans C. Cell proliferation and oxidative stress. Free Radic Res. 1989;7:149–59.

Chénais B, Andriollo M, Guiraud P, Belhoussine R, Jeannesson P. Oxidative stress involvement in chemically induced differentiation of K562 cells. Free Radic Biol Med. 2000;28:18–27.

Clanton TL, Zuo L, Klawitter P. Oxidants and skeletal muscle function: physiologic and pathophysiologic implications. Proc Soc Exp Biol Med. 1999;222:253–62.

Gabbita SP, Robinson KA, Stewart CA, Floyd RA, Hensley K. Redox regulatory mechanisms of cellular signal transduction. Arch Biochem Biophys. 2000;376:1–13.

Jabs T. Reactive oxygen intermediates as mediators of programmed cell death in plants and animals. Biochem Pharmacol. 1999;57:231–45.

Lander HM. An essential role for free radicals and derived species in signal transduction. FASEB J. 1997;11:118–24.

Szabó C. DNA strand breakage and activation of poly-ADP ribosyltransferase: a cytotoxic pathway triggered by peroxynitrite. Free Radic Biol Med. 1996;21:855–69.

Göktürk E, Turgut A, Baygu C, Gunal I, Seber S, Gulbas Z. Oxygen-free radicals impair fracture healing in rats. Acta Orthop. 1995;66:473–5.

Halliwell B. Oxidative stress, nutrition and health. Experimental strategies for optimization of nutritional antioxidant intake in humans. Free Radic Res. 1996;25:57–74.

Ji LL. Antioxidants and oxidative stress in exercise. Proc Soc Exp Biol Med. 1999;222:283–92.

Aruoma OI. Free radicals and antioxidant strategies in sports. J Nutr Biochem. 1994;5:370–81.

Ji LL, Leeuwenburgh C, Leichtweis S, Gore M, Fiebig R, Hollander J, et al. Oxidative stress and aging. Role of exercise and its influences on antioxidant systems. Ann N Y Acad Sci. 1998;854:102–17.

Hack V, Strobel G, Rau J-P, Weicker H. The effect of maximal exercise on the activity of neutrophil granulocytes in highly trained athletes in a moderate training period. Eur J Appl Physiol. 1992;65:520–4.

Goodship AE, Birch HL, Wilson AM. The pathobiology and repair of tendon and ligament injury. Vet Clin North Am Equine Pract. 1994;10:323–49.

Poli G, Parola M. Oxidative damage and fibrogenesis. Free Radic Biol Med. 1996;22:287–305.

Murrell GAC. Using nitric oxide to treat tendinopathy. Br J Sports Med. 2007;41:227–31.

Wang T-S, Kuo C-F, Jan K-Y, Huang H. Arsenite induces apoptosis in Chinese hamster ovary cells by generation of reactive oxygen species. J Cell Physiol. 1996;169:256–68.

Briehl MM, Baker AF. Modulation of the antioxidant defence as a factor in apoptosis. Cell Death Differ. 1996;3:63–70.

Morel Y, Barouki R. Repression of gene expression by oxidative stress. Biochem J. 1999;342:481–96.

Maffulli N, Longo UG, Loppini M, Denaro V. Current treatment options for tendinopathy. Expert Opin Pharmacother. 2010;11:2177–86.

Maffulli N, Longo UG, Denaro V. Novel approaches for the management of tendinopathy. J Bone Jt Surg - Ser A. 2010;92:2604–13.

Järvinen M, Józsa L, Kannus P, Järvinen TLN, Kvist M, Leadbetter W. Histopathological findings in chronic tendon disorders. Scand J Med Sci Sports. 1997;7:86–95.

Waterston SW. Subcutaneous rupture of the Achilles tendon: basic science and some aspects of clinical practice. Br J Sports Med. 1997;31:285–98.

Maffulli N. Re: etiologic factors associated with symptomatic Achilles tendinopathy. Foot Ankle Int Am Orthop Foot Ankle Soc Swiss Foot Ankle Soc. 2007;28:660 author reply 660-661.

Maffulli N, Kader D. Tendinopathy of tendo Achillis. J Bone Jt Surg - Ser B. 2002;84:1–8.

Clancy WG Jr, Neidhart D, Brand RL. Achilles tendonitis in runners: a report of five cases. Am J Sports Med. 1976;4:46–57.

Williams JG. Achilles tendon lesions in sport. Sports Med Auckl NZ. 1993;16:216–20.

Denstad TF, Roaas A. Surgical treatment of partial Achilles tendon rupture. Am J Sports Med. 1979;7:15–7.

Fox JM, Blazina ME, Jobe FW, Kerlan RK, Carter VS, Shields J, et al. Degeneration and rupture of the Achilles tendon. CLINORTHOP. 1975;107:221–4.

Magra M, Maffulli N. Genetic aspects of tendinopathy. J Sci Med Sport. 2008;11:243–7.

Sharma P, Maffulli N. Tendon injury and tendinopathy: healing and repair. J Bone Joint Surg Am. 2005;87:187–202.

Waldecker U, Hofmann G, Drewitz S. Epidemiologic investigation of 1394 feet: coincidence of hindfoot malalignment and Achilles tendon disorders. Foot Ankle Surg Off J Eur Soc Foot Ankle Surg. 2012;18:119–23.

Albers IS, Zwerver J, Diercks RL, Dekker JH, Van den Akker-Scheek I. Incidence and prevalence of lower extremity tendinopathy in a Dutch general practice population: a cross sectional study. BMC Musculoskelet Disord. 2016;17:16.

Renstrom P. Sports traumatology today: a review of common current sports injury problems. Ann Chir Gynaecol. 1991;80:81–93.

Kujala UM, Kvist M, Österman K. Knee injuries in athletes: review of exertion injuries and retrospective study of outpatient sports clinic material. Sports Med. 1986;3:447–60.

Astrom M, Rausing A. Chronic Achilles tendinopathy: a survey of surgical and histopathologic findings. Clin Orthop. 1995;316:151–64.

Chard MD, Cawston TE, Riley GP, Gresham GA, Hazleman BL. Rotator cuff degeneration and lateral epicondylitis: a comparative histological study. Ann Rheum Dis. 1994;53:30–4.

Khan KM, Cook JL, Bonar F, Harcourt P, Åstrom M. Histopathology of common tendinopathies: update and implications for clinical management. Sports Med. 1999;27:393–408.

Longo UG, Rittweger J, Garau G, Radonic B, Gutwasser C, Gilliver SF, et al. No influence of age, gender, weight, height, and impact profile in Achilles tendinopathy in masters track and field athletes. Am J Sports Med. 2009;37:1400–5.

Järvinen TAH, Kannus P, Maffulli N, Khan KM. Achilles tendon disorders: etiology and epidemiology. Foot Ankle Clin. 2005;10:255–66.

Fredberg U, Bolvig L, Andersen NT. Prophylactic training in asymptomatic soccer players with ultrasonographic abnormalities in Achilles and patellar tendons: the Danish super league study. Am J Sports Med. 2008;36:451–60.

Alfredson H, Ohberg L. Neovascularisation in chronic painful patellar tendinosis--promising results after sclerosing neovessels outside the tendon challenge the need for surgery. Knee Surg Sports Traumatol Arthrosc Off J ESSKA. 2005;13:74–80.

Woods C, Hawkins R, Hulse M, Hodson A. The Football Association Medical Research Programme: an audit of injuries in professional football-analysis of preseason injuries. Br J Sports Med. 2002;36:436–41 discussion 441.

Sobhani S, Dekker R, Postema K, Dijkstra PU. Epidemiology of ankle and foot overuse injuries in sports: a systematic review. Scand J Med Sci Sports. 2013;23:669–86.

Leadbetter WB. Cell-matrix response in tendon injury. Clin Sports Med. 1992;11:533–78.

Ilfeld FW. Can stroke modification relieve tennis elbow? Clin Orthop. 1992;276:182–6.

Kibler WB, Chandler TJ, Pace BK. Principles of rehabilitation after chronic tendon injuries. Clin Sports Med. 1992;11:661–7.

MG JF and C. Lateral and medial epicondylitis of the elbow. - PubMed - NCBI [Internet]. Available from: https://www.ncbi.nlm.nih.gov/pubmed/10708988 . [cited 2017 Nov 3]

Almekinders LC, Gilbert JA. Healing of experimental muscle strains and the effects of nonsteroidal antiinflammatory medication. Am J Sports Med. 1986;14:303–8.

Garrett WE Jr. Muscle strain injuries: clinical and basic aspects. Med Sci Sports Exerc. 1990;22:436–43.

Dean BJF, Franklin SL, Carr AJ. The peripheral neuronal phenotype is important in the pathogenesis of painful human tendinopathy: a systematic review. Clin Orthop. 2013;471:3036–46.

Hutchison A-M, Evans R, Bodger O, Pallister I, Topliss C, Williams P, et al. What is the best clinical test for Achilles tendinopathy? Foot Ankle Surg Off J Eur Soc Foot Ankle Surg. 2013;19:112–7.

Maffulli N, Testa V, Capasso G, Oliva F, Sullo A, Benazzo F, et al. Surgery for chronic Achilles tendinopathy yields worse results in nonathletic patients. Clin J Sport Med Off J Can Acad Sport Med. 2006;16:123–8.

Maffulli N, Wong J, Almekinders LC. Types and epidemiology of tendinopathy. Clin Sports Med. 2003;22:675–92.

Ohberg L, Lorentzon R, Alfredson H. Neovascularisation in Achilles tendons with painful tendinosis but not in normal tendons: an ultrasonographic investigation. Knee Surg Sports Traumatol Arthrosc Off J ESSKA. 2001;9:233–8.

Rolf C, Movin T. Etiology, histopathology, and outcome of surgery in achillodynia. Foot Ankle Int. 1997;18:565–9.

Maffulli N, Oliva F, Maffulli GD, Giai Via A, Gougoulias N. Minimally invasive Achilles tendon stripping for the management of tendinopathy of the main body of the Achilles tendon. J Foot Ankle Surg Off Publ Am Coll Foot Ankle Surg. 2017;56:938–42.

Silbernagel KG, Brorsson A, Lundberg M. The majority of patients with Achilles tendinopathy recover fully when treated with exercise alone: a 5-year follow-up. Am J Sports Med. 2011;39:607–13.

Weinreb JH, Sheth C, Apostolakos J, McCarthy M-B, Barden B, Cote MP, et al. Tendon structure, disease, and imaging. Muscles Ligaments Tendons J. 2014;4:66–73.

PubMed   PubMed Central   Google Scholar  

Maffulli N, Via AG, Oliva F. Chronic Achilles tendon disorders. Tendinopathy and Chronic Rupture Clin Sports Med. 2015;34:607–24.

Sayana MK, Maffulli N. Eccentric calf muscle training in non-athletic patients with Achilles tendinopathy. J Sci Med Sport. 2007;10:52–8.

Alfredson H, Pietilä T, Jonsson P, Lorentzon R. Heavy-load eccentric calf muscle training for the treatment of chronic Achilles tendinosis. Am J Sports Med. 1998;26:360–6.

Rompe JD, Nafe B, Furia JP, Maffulli N. Eccentric loading, shock-wave treatment, or a wait-and-see policy for tendinopathy of the main body of tendo Achillis: a randomized controlled trial. Am J Sports Med. 2007;35:374–83.

van Sterkenburg MN, van Dijk CN. Injection treatment for chronic midportion Achilles tendinopathy: do we need that many alternatives? Knee Surg Sports Traumatol Arthrosc Off J ESSKA. 2011;19:513–5.

Maffulli N, Papalia R, D’Adamio S, Diaz Balzani L, Denaro V. Pharmacological interventions for the treatment of Achilles tendinopathy: a systematic review of randomized controlled trials. Br Med Bull. 2015;113:101–15.

van Sterkenburg MN, de Jonge MC, Sierevelt IN, van Dijk CN. Less promising results with sclerosing ethoxysclerol injections for midportion Achilles tendinopathy: a retrospective study. Am J Sports Med. 2010;38:2226–32.

Ebbesen BH, Mølgaard CM, Olesen JL, Gregersen HE, Simonsen O. No beneficial effect of polidocanol treatment in Achilles tendinopathy: a randomised controlled trial. Knee Surg Sports Traumatol Arthrosc Off J ESSKA. 2018;26:2038–44.

de Vos RJ, Weir A, van Schie HTM, Bierma-Zeinstra SMA, Verhaar JAN, Weinans H, et al. Platelet-rich plasma injection for chronic Achilles tendinopathy: a randomized controlled trial. JAMA. 2010;303:144–9.

Chan O, O’Dowd D, Padhiar N, Morrissey D, King J, Jalan R, et al. High volume image guided injections in chronic Achilles tendinopathy. Disabil Rehabil. 2008;30:1697–708.

Abate M, Di Carlo L, Verna S, Di Gregorio P, Schiavone C, Salini V. Syner gistic activity of platelet rich plasma and high volume image guided injection for patellar tendinopathy. Knee Surg Sports Traumatol Arthrosc Off J ESSKA. 2018;26(12):3645–51.

Paavola M, Kannus P, Paakkala T, Pasanen M, Järvinen M. Long-term prognosis of patients with achilles tendinopathy. An observational 8-year follow-up study. Am J Sports Med. 2000;28:634–42.

Longo UG, Ramamurthy C, Denaro V, Maffulli N. Minimally invasive stripping for chronic Achilles tendinopathy. Disabil Rehabil. 2008;30:1709–13.

Alfredson H. Ultrasound and Doppler-guided mini-surgery to treat midportion Achilles tendinosis: results of a large material and a randomised study comparing two scraping techniques. Br J Sports Med. 2011;45:407–10.

Testa V, Capasso G, Benazzo F, Maffulli N. Management of Achilles tendinopathy by ultrasound-guided percutaneous tenotomy. Med Sci Sports Exerc. 2002;34:573–80.

Maffulli N, Testa V, Capasso G, Bifulco G, Binfield PM. Results of percutaneous longitudinal tenotomy for Achilles tendinopathy in middle- and long-distance runners. Am J Sports Med. 1997;25:835–40.

Gajhede-Knudsen M, Ekstrand J, Magnusson H, Maffulli N. Recurrence of Achilles tendon injuries in elite male football players is more common after early return to play: an 11-year follow-up of the UEFA Champions League injury study. Br J Sports Med. 2013;47:763–8.

Schultz W, Stinus H, Schleicher W, Hess T. Stress reactions--stress fracture of the upper femoral neck in endurance sports. Sportverletz Sportschaden Organ Ges Orthopadisch-Traumatol Sportmed. 1991;5:81–4.

Armstrong DW, Rue J-PH, Wilckens JH, Frassica FJ. Stress fracture injury in young military men and women. Bone. 2004;35:806–16.

Fullem BW. Overuse lower extremity injuries in sports. Clin Podiatr Med Surg. 2015;32:239–51.

Harrast MA, Colonno D. Stress fractures in runners. Clin Sports Med. 2010;29:399–416.

Wright AA, Hegedus EJ, Lenchik L, Kuhn KJ, Santiago L, Smoliga JM. Diagnostic accuracy of various imaging modalities for suspected lower extremity stress fractures: a systematic review with evidence-based recommendations for clinical practice. Am J Sports Med. 2016;44:255–63.

Rizzone KH, Ackerman KE, Roos KG, Dompier TP, Kerr ZY. Epidemiology of stress fractures in collegiate student-athletes, 2004–2005 through 2013–2014 academic years. J Athl Train. 2017; Available from: http://natajournals.org/doi/10.4085/1062-6050-52.8.01 . [cited 2017 Nov 5].

Giza E, Mithöfer K, Farrell L, Zarins B, Gill T. Injuries in women’s professional soccer. Br J Sports Med. 2005;39:212–6 discussion 212-216.

Associations NF of SHS. High School Athletics Participation Survey 2011–2012. National Federation of State High School Associations Indianapolis, IN; 2012.

NCAA Publications - 1981-82 - 2008-09 NCAA Sports Sponsorship and Participation Rates Report [Internet]. [cited 2018 Dec 1]. Available from: https://www.ncaapublications.com/p-4177-1981-82-2008-09-ncaa-sports-sponsorship-and-participation-rates-report.aspx

Prather H, Hunt D, McKeon K, Simpson S, Meyer EB, Yemm T, et al. Are elite female soccer athletes at risk for disordered eating attitudes, menstrual dysfunction, and stress fractures? PM R. 2016;8:208–13.

Wright AA, Taylor JB, Ford KR, Siska L, Smoliga JM. Risk factors associated with lower extremity stress fractures in runners: a systematic review with meta-analysis. Br J Sports Med. 2015;49:1517–23.

Bowerman EA, Whatman C, Harris N, Bradshaw E. A review of the risk factors for lower extremity overuse injuries in young elite female ballet dancers. J Dance Med Sci Off Publ Int Assoc Dance Med Sci. 2015;19:51–6.

Johanson MA. Contributing factors in microtrauma injuries of the lower extremity. J Back Musculoskelet Rehabil. 1992;2:12–25.

Pohl MB, Mullineaux DR, Milner CE, Hamill J, Davis IS. Biomechanical predictors of retrospective tibial stress fractures in runners. J Biomech. 2008;41:1160–5.

James SL, Bates BT, Osternig LR. Injuries to runners. Am J Sports Med. 1978;6:40–50.

Kiss ZS, Khan KM, Fuller PJ. Stress fractures of the tarsal navicular bone: CT findings in 55 cases. Am J Roentgenol. 1993;160:111–5.

Kiuru MJ, Pihlajamaki HK, Hietanen HJ, Ahovuo JA. MR imaging, bone scintigraphy, and radiography in bone stress injuries of the pelvis and the lower extremity. Acta Radiol. 2002;43:207–12.

Heckman JD, Ryaby JP, McCabe J, Frey JJ, Kilcoyne RF. Acceleration of tibial fracture-healing by non-invasive, low-intensity pulsed ultrasound. J Bone Jt Surg - Ser A. 1994;76:26–34.

Malliaropoulos N, Alaseirlis D, Konstantinidis G, Papalada A, Tsifountoudis I, Petras K, et al. Therapeutic ultrasound in navicular stress injuries in elite track and field athletes. Clin J Sport Med Off J Can Acad Sport Med. 2017;27:278–82.

Rettig AC, Shelbourne KD, McCarroll JR, Bisesi M, Watts J. The natural history and treatment of delayed union stress fractures of the anterior cortex of the tibia. Am J Sports Med. 1988;16:250–5.

Kessler JI, Nikizad H, Shea KG, Jacobs JC, Bebchuk JD, Weiss JM. The demographics and epidemiology of osteochondritis dissecans of the knee in children and adolescents. Am J Sports Med. 2014;42:320–6.

Kumar V, Bhatnagar N, Lodhi JS. Grade I osteochondritis dissecans in a young professional athlete. Indian J Orthop. 2018;52:344–52.

Krause M, Hapfelmeier A, Möller M, Amling M, Bohndorf K, Meenen NM. Healing predictors of stable juvenile osteochondritis dissecans knee lesions after 6 and 12 months of nonoperative treatment. Am J Sports Med. 2013;41:2384–91.

Rothermich MA, Glaviano NR, Li J, Hart JM. Patellofemoral pain. Epidemiology, pathophysiology, and treatment options. Clin Sports Med. 2015;34:313–27.

Uppstrom TJ, Gausden EB, Green DW. Classification and assessment of juvenile osteochondritis dissecans knee lesions. Curr Opin Pediatr. 2016;28:60–7.

Suzue N, Matsuura T, Iwame T, Hamada D, Goto T, Takata Y, et al. Prevalence of childhood and adolescent soccer-related overuse injuries. J Med Investig JMI. 2014;61:369–73.

Yang JS, Bogunovic L, Wright RW. Nonoperative treatment of osteochondritis dissecans of the knee. Clin Sports Med. 2014;33:295–304.

Download references

Acknowledgements

The authors acknowledge for the kind support of Professor Biagio Moretti and Dr. Lorenzo Moretti, Department of Orthopedics and Traumatology of University Hospital “Policlinico” of Bari, Bari, Italy.

The authors declare that they did not have any funding.

Availability of data and materials

Not applicable

Author information

Authors and affiliations.

Department of Musculoskeletal Disorders, School of Medicine and Dentistry, University of Salerno, Salerno, Italy

R. Aicale, D. Tarantino & N. Maffulli

Barts and the London School of Medicine and Dentistry, Centre for Sports and Exercise Medicine, Mile End Hospital, Queen Mary University of London, 275 Bancroft Road, London, E1 4DG, England

N. Maffulli

You can also search for this author in PubMed   Google Scholar

Contributions

Two medical doctors (RA and DT) performed the search of articles in the available scientific literature, while the researcher with more experience (NM) solved cases of doubt. At the beginning of the procedure, each examiner read the abstracts of all articles and selected the relevant ones. After 1 week, the same studies were read again to establish the agreement of the researchers on the selection. Each author has contributed to writing the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to R. Aicale .

Ethics declarations

Ethics approval and consent to participate, consent for publication, competing interests.

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License ( http://creativecommons.org/licenses/by/4.0/ ), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated.

Reprints and permissions

About this article

Cite this article.

Aicale, R., Tarantino, D. & Maffulli, N. Overuse injuries in sport: a comprehensive overview. J Orthop Surg Res 13 , 309 (2018). https://doi.org/10.1186/s13018-018-1017-5

Download citation

Received : 21 June 2018

Accepted : 21 November 2018

Published : 05 December 2018

DOI : https://doi.org/10.1186/s13018-018-1017-5

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Stress fracture
• Overuse injuries
• Osteochondritis dissecans
• Tendinopathy

Journal of Orthopaedic Surgery and Research

ISSN: 1749-799X

• Submission enquiries: [email protected]

• Review Paper
• Open access
• Published: 14 April 2021

Machine learning methods in sport injury prediction and prevention: a systematic review

• Hans Van Eetvelde   ORCID: orcid.org/0000-0003-1421-7721 1 ,
• Luciana D. Mendonça   ORCID: orcid.org/0000-0002-4495-1807 2 , 3 , 4 ,
• Christophe Ley   ORCID: orcid.org/0000-0002-2290-8437 1 ,
• Romain Seil   ORCID: orcid.org/0000-0001-8806-9384 5 &
• Thomas Tischer   ORCID: orcid.org/0000-0002-3942-0235 6  

Journal of Experimental Orthopaedics volume  8 , Article number:  27 ( 2021 ) Cite this article

29k Accesses

92 Citations

22 Altmetric

Metrics details

Injuries are common in sports and can have significant physical, psychological and financial consequences. Machine learning (ML) methods could be used to improve injury prediction and allow proper approaches to injury prevention. The aim of our study was therefore to perform a systematic review of ML methods in sport injury prediction and prevention.

A search of the PubMed database was performed on March 24th 2020. Eligible articles included original studies investigating the role of ML for sport injury prediction and prevention. Two independent reviewers screened articles, assessed eligibility, risk of bias and extracted data. Methodological quality and risk of bias were determined by the Newcastle–Ottawa Scale. Study quality was evaluated using the GRADE working group methodology.

Eleven out of 249 studies met inclusion/exclusion criteria. Different ML methods were used (tree-based ensemble methods ( n  = 9), Support Vector Machines ( n  = 4), Artificial Neural Networks ( n  = 2)). The classification methods were facilitated by preprocessing steps ( n  = 5) and optimized using over- and undersampling methods ( n  = 6), hyperparameter tuning ( n  = 4), feature selection ( n  = 3) and dimensionality reduction ( n  = 1). Injury predictive performance ranged from poor (Accuracy = 52%, AUC = 0.52) to strong (AUC = 0.87, f1-score = 85%).

Conclusions

Current ML methods can be used to identify athletes at high injury risk and be helpful to detect the most important injury risk factors. Methodological quality of the analyses was sufficient in general, but could be further improved. More effort should be put in the interpretation of the ML models.

Injuries are common in individual and team sports and can have significant physical, psychosocial and financial consequences [ 3 , 13 , 22 ]. Understanding injury risk factors and their interplay is thereby a key component of preventing future injuries in sport [ 4 ]. An abundance of research has attempted to identify injury risk factors [ 4 , 6 , 28 ]. However, sports injuries are a consequence of complex interactions of multiple risk factors and inciting events making a comprehensive model necessary [ 6 , 28 ]. It has to account for the events leading to the injury situation, as well as to include a description of body and joint biomechanics at the time of injury [ 4 ]. Due to the many interactions between intrinsic and extrinsic risk factors as well as their sometimes highly unpredictable nature (e.g., contact with another player), the ability to foresee the occurrence of an inciting injury event is challenging. Therefore, predictive modelling should not only focus on the prediction of the occurrence of an injury itself but, moreover, it should try to identify injury risk at an individual level and implement interventions to mitigate the level of risk [ 28 ]. In order to plan effective preventive intervention, it is therefore important to be aware both of the various isolated risk factors and their interaction [ 6 ].

In recent years, the use of advanced Artificial Intelligence (AI) methods has appeared in sports medicine to tackle this challenging multi-faceted task [ 1 , 5 , 14 , 16 ]. AI methods have already been used successfully in sports science within the realm of game analysis, tactics, performance analysis and outcome predictions [ 12 , 17 , 21 ] and are about to start transforming clinical medicine [ 9 , 31 , 33 , 39 , 42 ]. However, for clinicians, the application and the understanding of AI is often difficult [ 24 ]. Therefore, the explanations of the core terms for AI application are provided in Supplementary File S 1 .

AI is mostly narrowed down to Machine Learning (ML) methods although it is a very broad concept comprising every aspect of mimicking human intelligence. ML is the study of algorithms that can automatically learn from data to make new decisions [ 23 ]. Current ML methods include Neural Networks, Support Vector Machines, or Random Forests which are part of a 'Machine Learning pipeline' (Fig.  1 ). The available data for the ML model has to be of high quality and can be any data deemed useful for the purpose of injury prediction. This data is split in two parts (data splitting), the so-called training and test data. First, the algorithm has to learn the relationship between outcome of interest (injury or not) and the potential contributing factors (also called predictors/features/covariates/explanatory variables) from the training data set. The test data can then be used to test the prediction capacity of the algorithm learned from the training data. It is important that this quality check is not achieved on the training data, but on unseen data, hence the data splitting at the beginning. The quality and size of the data sets are important parameters for the quality of the results. To improve the quality of these large and complex datasets and to ensure optimal operation of the ML algorithms, data preprocessing methods (imputation, standardization, discretization), dimension reduction and feature selection can be applied (see Supplementary File S 1 ). Most ML procedures further require parameter tuning, a sort of optimization of parameters which cannot be estimated directly from the data (e.g., number of trees to be used in a Random Forest). When the entire ML pipeline is fitted on the training data, the outcome of the test data is predicted. Since we know the true outcome of the test data, this allows us to evaluate our established prediction model. Finally, well-performing models provide an idea of the most important risk factors, by observing which factors have the largest influence in these models.

Schematic figure of the Machine Learning approach. The entire Machine Learning process is shown. Parts in dotted shapes are optional or not always necessary

Considering that sport injury prediction and prevention are trending topics in sport science [ 12 , 13 , 16 ], the intention of this systematic review is to synthesize the evidence of sophisticated ML algorithms in sport injury prediction and prevention. Our systematic review differs from the one by Claudino et al. [ 12 ] in that we focus on injury prevention and risk factor identification together with a deeper examination of the used ML analyses. The following three topics are assessed:

Identify the currently used definition of ML as well as predominantly used ML methods.

Identify the accuracy of the currently used ML methods to predict injury.

Evaluate the used methods for sport injury prevention purposes.

This systematic review was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [ 30 ]. The review protocol was prospectively registered at PROSPERO (International prospective register of systematic reviews—ie, CRD42020177708).

Search strategy and inclusion/exclusion criteria

A systematic electronic search of the PubMed database was executed on March 24th 2020 to identify studies investigating Machine Learning methods in injury prediction and prevention. The following search term was used in all fields: (“deep learning” OR “artificial intelligence” OR “machine learning” OR “neural network” OR „neural networks “ OR „support vector machines “ OR „nearest neighbor “ OR „nearest neighbors “ OR „random forest “ OR „random forests “ OR „trees" OR „elastic net “ OR „ridge “ OR „lasso “ OR „boosting “ OR „predictive modeling “ OR “learning algorithms” OR „bayesian logistic regression “) AND (“sport” OR „sports “ OR “athlete” OR “athletes”) AND („injury “ OR "injuries"). We did not use limits to perform the search and no date restrictions were applied. Inclusion criteria were as follows: (i) Original studies investigating the role of machine learning for sport injury prediction and sport injury prevention, (ii) English-language studies, (iii) studies published online or in print in a peer-reviewed journal. Injury prediction had to refer to predicting either the occurrence, the severity, or the type of injuries on the basis of risk factors. The exclusion criteria were as follows: (i) not being sport-specific, (ii) not covering injury prevention or injury prediction, (iii) meeting abstracts and proceedings. Also, studies were excluded if the used approach was rather statistical than ML. This explains why, for example, two papers from Hasler et al. [ 19 , 20 ] and one from Mendonça et al. [ 29 ] were not included here.

Study selection

The titles and abstracts of all articles were screened for relevance according to the inclusion and exclusion criteria (L.M. and T.T.). If no abstract was available, the full-text article was obtained to assess the relevance of the study. The full text was subsequently reviewed for possible inclusion in the systematic review for all articles that were not excluded during the initial screening process. A third reviewer (H.E.) resolved between-reviewer discrepancies. In addition to the electronic search, the reference lists of all included articles and review articles were manually searched (C.L., T.T., H.E.) for additional relevant articles. Moreover, if any systematic reviews on ML in sport injury prediction and prevention were identified during the screening process, the reference list was screened to identify any further studies.

Methodological quality

Methodological quality and risk of bias of included studies were determined by the Newcastle–Ottawa Scale (NOS) [ 45 ]. Eligible studies were independently rated by two authors blind to the study authors and institutions (L.M. and T.T.), with discrepancies resolved by a third author (H.E.). The NOS contains eight categories relating to methodological quality and each study was given an eventual score out of a maximum of 8 points. A score of 0–3 points equated to a low quality study, a score of 4–6 points equated to a moderate quality study, with a score of 7–8 points required for a study to be given a score of high quality.

Data extraction

Characteristics of all included studies (i.e. participants, type of study, sample size, etc.) and about ML used (i.e. data pre-processing, classification method, etc.) were extracted independently by two reviewers (H.E. and C.L.), with a third (L.M.) resolving potential discrepancies.

Data analysis

Two independent reviewers (L.M. and T.T.) assessed the quality of evidence using the GRADE methodology [ 18 ]. In the current review, evidence started at moderate certainty, since investigation of publication bias was not possible due to the small number of included trials. Then it was downgraded by one level for imprecision when the analysed sample was < 300 participants (serious imprecision was downgraded by two levels); and by one level for risk of bias when the mean NOS Score was < 6 out of 9. Between-reviewer discrepancies were resolved by a third investigator (R.S.).

In the scope of this systematic literature review, 246 articles were found, and an additional three articles added by hand search from which a total of 11 articles were included according to the strict inclusion/exclusion criteria for this systematic review (Fig.  2 ).

PRISMA flow chart

Study characteristics

Table 1 lists all details of the included studies. Studies were prospective cohort studies ( n  = 9) [ 2 , 11 , 25 , 27 , 32 , 35 , 36 , 38 , 41 ] or case–control studies ( n  = 2) [ 41 , 46 ]. Most of them were performed in soccer ( n  = 4) [ 2 , 32 , 35 , 36 ] and Australian Football ( n  = 3).[ 11 , 27 , 38 ] Two studies incorporated athletes from multiple sports [ 25 , 34 ]. The number of participants ranged between 25 and 363. In seven studies, the athlete was the unit of observation [ 2 , 25 , 32 , 34 , 35 , 38 , 46 ]. In the remaining 4 studies, there were multiple observations per player [ 11 , 27 , 36 , 41 ]. Both occurrence ( n  = 11) [ 2 , 11 , 25 , 27 , 32 , 34 , 35 , 36 , 38 , 41 , 46 ] and type of injury ( n  = 2) (acute/overuse [ 35 ], contact/non-contact [ 27 ]) were evaluated, whereby lower limb muscle injury was the most often assessed outcome [ 2 , 25 , 38 ] and only one publication investigated specifically upper limb injuries [ 46 ].

Study Quality

The methodological quality of the included studies ranged from 5 to 8 in the NOS scale (Table 2 ). All studies had proper ascertainment of outcome/exposure; follow-up long enough for outcomes to occur, same method of ascertainment for cases and controls; and adequacy of follow-up of cohorts/non-response rate. Seven studies (63.63%) were downgraded for methodological quality due to imprecision and three (27.27%) because of risk of bias (Table 3 ).

Data analysis characteristics

In all 11 papers, the outcome variable was the occurrence of injury or the type of injury, which are categorical variables, making the base models classification models. From the 11 considered papers, 9 papers used tree-based models [ 2 , 11 , 25 , 32 , 34 , 35 , 36 , 38 , 41 ], 4 papers used Support Vector Machines [ 11 , 34 , 38 , 46 ] and 2 papers used Artificial Neural Networks [ 27 , 38 ]. Eight out of 9 papers using tree-based models applied a bagging strategy [ 2 , 11 , 25 , 32 , 34 , 36 , 38 , 41 ], whereof 5 used a Random Forest approach [ 11 , 34 , 36 , 38 , 41 ]. Four papers used boosting algorithms to construct tree ensemble methods [ 2 , 25 , 32 , 35 ].

The training, validation and test strategy for the used ML approaches varied largely between the different studies. For the evaluation and comparison of the methods, 7 papers [ 2 , 25 , 27 , 32 , 34 , 36 , 46 ] used cross-validation and 4 [ 11 , 35 , 38 , 41 ] used a single data-splitting approach. Of the former, four [ 2 , 25 , 32 , 36 ] used stratified cross-validation, which may be especially of interest in unbalanced data, because it ensures that in both training and test set the number of positive cases (injuries) is sufficiently high. In 4 papers [ 11 , 34 , 35 , 38 ] the training dataset was split further for tuning the hyperparameters. In 3 papers [ 11 , 36 , 37 ] the authors repeated their entire analysis a large number of times to adjust for the randomness in the resampling and under/oversampling methods.

Three of the discussed papers used feature selection methods [ 34 , 36 , 46 ]. Rodas et al. [ 34 ] used the LASSO method for selecting significant features, Rossi et al. [ 36 ] eliminated features by applying cross-validation on a separate part of the data, and Whiteside et al. [ 46 ] evaluated all possible feature subsets. Carey et al. [ 11 ] used Principal Component Analysis for reducing the dimensionality of the data instead of feature selection.

To adjust for imbalanced data, the training datasets were over-and/or undersampled in 6 papers [ 2 , 11 , 25 , 32 , 36 , 38 ]. All of them used oversampling of the minority class (injuries) and 4 of them applied undersampling of the majority class (non-injuries) [ 2 , 11 , 25 , 32 ].

Data pre-processing was used in some papers to optimize the performance of the classification methods. To solve the missing values problem, three papers [ 2 , 25 , 34 ] mentioned using imputation methods. In three papers [ 2 , 25 , 32 ], the continuous variables were transformed into categorical variables, using cut-off values found in the literature or based on the data. There was only one paper [ 37 ] that mentioned a standardization of the continuous variables.

Some of the studies had small deficits in the Machine Learning Pipeline approach. Four papers in this review had multiple observations per athlete [ 11 , 27 , 36 , 41 ] and it seems that players may appear in both the training and test datasets, which would be a violation of the rule that the training and test dataset should be independent from each other. The results of these studies can therefore not be generalized to a bigger population. The other mistakes were made in the preprocessing phase. Four papers [ 2 , 25 , 32 , 38 ] seemed to perform discretization or standardization on the entire dataset (including test dataset), which in that case would be an example of data leakage, i.e. using the test data in the training process. This should be avoided since it does not reflect reality as the test dataset has to be seen as future data. On the other hand, Ruddy et al. [ 38 ] independently standardized the training and test dataset. Applying different transformations on the training and test dataset will cause non-optimal operation of the classifier and can lead to lower predictive performance. A structured overview of the data analysis characteristics can be found in Table 4 .

Performance in predicting injury occurrence

In Table 5 , the study results characteristics are given for each of the included papers. For predicting the occurrence of the outcome (injury in general, muscle injury, …), seven papers used Area Under the ROC Curve (AUC) as an evaluation measure [ 2 , 11 , 25 , 32 , 36 , 38 , 41 ], while the remaining four papers used only metrics based on the confusion matrix, e.g. accuracy, sensitivity, specificity, precision and f1-score. Eight out of eleven studies [ 2 , 25 , 27 , 32 , 35 , 36 , 41 , 46 ] reported appropriate to good performance of the Machine Learning prediction methods. AUC values for predicting the outcome ranged between 0.64 and 0.87, and high values were found for accuracy (75%—82.9%), sensitivity (55.6%—94.5%), specificity (74.2%—87%) and precision (50%—85%). Three papers [ 11 , 34 , 38 ] reported low prediction potential for their built ML models, showing low AUC (0.52—0.65) and accuracy (52%) values.

Most important injury predictors

Analysed risk factors included both modifiable (training load, psychological and neuromuscular assessment, stress level, …) and non-modifiable (demographics, genetic markers, anthropometric measurements, previous injury, …) factors (for more details, see Table 1 ). In 4 papers [ 2 , 25 , 32 , 41 ], the authors have counted the number of appearances of each feature in the final ensemble of decision trees. Two papers [ 34 , 46 ] counted the number of times that a feature is selected by their feature selection procedure. Rossi et al. [ 36 ] used the decrease in Gini coefficient to measure the importance of variables and Rommers et al. [ 35 ] used a SHAP summary plot [ 26 ]. This plot was based on the Shapley values in game theory and shows the importance of the variables, as well as the relation between high/low feature values and high/low injury risk. Because of the wide variety of features used over the different papers, not much consistency was found in the reported important predictors. The features that were reported twice as important were previous injury [ 25 , 36 ], higher training load [ 36 , 41 ], and higher body size (in youth players only) [ 32 , 35 ]. Note that lower training load after previous injury might indicate a not fully recovered athlete and can hence be considered being a risk factor after previous injury [ 36 ].

The 11 studies included in this systematic review showed that ML methods can be successfully applied for sport injury prediction. The most promising results to predict injury risk were obtained in elite youth football players based on anthropometric, motor coordination and physical performance measures with a high accuracy of 85% [ 35 ], and in professional soccer based on a pre-season screening evaluation with a high sensitivity (77.8%) and specificity (83.8%).[ 2 ] This is in opposition with several authors who found that screening tests were not successful in predicting sports injuries [ 40 , 43 ]. These results are promising in the sense that future models might help coaches, physical trainers and medical practitioners in the decision-making process for injury prevention.

Data inclusion was still limited in the analysed studies, where only selected variables were included (e.g., only anthropometric, motor coordination and physical performance measures in the study by Rommers et al. [ 35 ]). Nevertheless, the achieved accuracy was quite high and future prediction might become even higher by using smart machine learning approaches or by incorporating more data (e.g., using sensors, more intense monitoring of athletes) [ 44 ]. Future studies will need to refine the target of injury prediction with AI/ML. This can either be achieved with an increase of the number of different injuries affecting a specific population or a study cohort [ 35 ] or with a targeted inclusion of specific injuries with a high injury incidence like hamstring injuries in football or athletics [ 2 , 38 ], or a high injury burden like anterior cruciate ligament (ACL) injuries in pivoting sports or ulnar collateral injuries in baseball [ 46 ]. The types and number of injury risk factors to be included in these studies are manifold and vary for each target. Large datasets may help the sports medicine community to improve the understanding of the respective influence of each factor on injury occurrence as well as their specific interactions in a given environment, allowing for a more systemic approach of sports injury prevention [ 6 , 7 , 8 , 15 ].

In the new field of ML for sports injury prevention, the level of quality of the published studies is of utmost importance. The analysis of the methodological quality of the 11 included studies indicates that they had very-low to moderate methodological quality according to the GRADE analysis. Imprecision (i.e., a study including relatively few participants/events) is an issue that may be improved with multicentric studies. Only 3 studies [ 11 , 36 , 41 ] had a NOS score under 6 and only 1 study [ 32 ] scored in comparability. In fact, the main reason to a lower NOS score was lack of comparability, which indicates that either cases and controls or exposed and non-exposed individuals were not matched in the design and/or confounders were not adjusted for in the analysis. Oliver et al., the only paper considering comparability, recruited 6 professional football teams of the English Premier League and Championship and followed 355 athletes [ 32 ]. Injured and non-injured players were compared in all continuous variables and all 95% CI presented had proper range, indicating adequate matching between groups. Future studies should be aware of this common limitation and include the comparison between groups.

In terms of ML methodology, the following observations can be made from this review. (i) Tree-based models are currently the most popular ML models in sports medicine. They are easy to visualize and interpret, and they can be extended to ensemble methods for boosting and bagging purposes or adapted to be cost-sensitive. The two publications that did not use tree-based models were the first to be published on the subject [ 27 , 46 ], thereby confirming the trend that more recent studies seem to adhere to this methodology. (ii) Concerning training and evaluating the ML models, a big variety between the 11 papers could be noticed. It was surprising to see that only 4 papers [ 11 , 34 , 35 , 38 ] mentioned having tuned the hyperparameters to optimize the performance of the ML methods, since tuning hyperparameters is recommendable (though not mandatory) in order to take the most out of the ML methods. The other studies may have used values from the literature or the default values from the used software, which may have led to a failure to identify the optimal model. (iii) The findings from this review further reveal that future studies involving ML approaches in the field of sports injury prevention should aim for a higher methodological quality. One of the identified deficits of the analysed studies was the dependency between training and test datasets.

The predictive performance of the considered publications was very heterogeneous. It should be emphasized that the reported predictive performances cannot be seen as a quality measure of the ML analysis per se, because they are depending on many other factors, like the kind of included risk factors, the design of the study, the sample size or the unit of observation. This also appeared when the publications of Ayala et al. [ 2 ], Lopez-Valenciano et al. [ 25 ] and Oliver et al. [ 32 ] were compared to each other. They used similar preprocessing and processing steps and classification trees, but reported very different performance values (AUC ranging from 0.663 to 0.873). Furthermore, the reported measures (AUC, accuracy, sensitivity, specificity) might not be the best measures to evaluate the prediction models, since these measures only see black and white (injured or not injured), while probabilistic scoring rules, such as the Brier Score and the Logarithmic Loss, would be able to evaluate the exactness of a predicted probability (e.g. this player has 30% chance to get injured) as is stated by Carey et al.[ 10 ]. From a clinical point of view, it could be more informative to know the probability of injury instead of only the classification into a high or low risk profile.

When dealing with injury prediction and prevention, it is important to identify especially modifiable risk factors, which can be intrinsic or extrinsic [ 28 ]. While some studies did not provide any information on the relative importance or influence of an individual risk factor, others used the number of times that a considered variable appeared in the ensemble of decision trees. Rossi et al. [ 36 ] measured how much the predictive performance of an algorithm would decrease if a specific variable would be left out as a predictor. Rommers et al. [ 35 ] provided a visualisation of the influence of the risk factors on the predicted injury risk. Therefore, it appeared that more efforts should be done to understand the relative weight of individual risk factors on the injury risk. This approach may help guiding practitioners to apply targeted interventions to the athletes at high injury risk.

Limitations of the systematic review

Besides investigating the outcome of machine learning algorithms in injury prediction and prevention, this systematic review also focused on the methodology of AI/ML studies, which makes some parts probably challenging to read for sports medicine clinicians. To avoid misinterpretation, a brief summary of AI/ML methods was included. It is important to stress that a previous review of Claudino et al. [ 12 ] about the use of AI in team sports provided a first overview of the topic, however it included methods that were used in a clearly statistical way, such as Bayesian logistic regression and single decision tree classifiers. Using this categorization implies that studies which were performed before the era of AI/ML and including statistical methods like linear or logistic regression would need to be considered to get a complete overview of the topic. This would seriously dilute true ML approaches. Another limitation is the fact that only PubMed database was included in this review. Even though, more relevant studies were found compared to reviews using other databases, such as e.g. in Claudino et al. [ 12 ].

This systematic review showed that ML methods may be used to identify athletes at high injury risk during sport participation and that it may be helpful to identify risk factors. However, although the majority of the analysed studies did apply machine learning methods properly to predict injuries, the methodological study quality was moderate to very low. Sports injury prediction is a growing area and further developments in this promising field should be encouraged with respect to the big potential of AI/ML methods.

Availability of data and materials

Not applicable.

Adetiba E, Iweanya VC, Popoola SI, Adetiba JN, Menon C (2017) Automated detection of heart defects in athletes based on electrocardiography and artificial neural network. Cogent Eng 4:1411220

Article   Google Scholar  

Ayala F, López-Valenciano A, Gámez Martín JA, De Ste CM, Vera-Garcia FJ, García-Vaquero MDP, Ruiz-Pérez I, Myer GD (2019) A Preventive Model for Hamstring Injuries in Professional Soccer: Learning Algorithms. Int J Sports Med 40:344–353. https://doi.org/10.1055/a-0826-1955

Article   PubMed   Google Scholar  

Bahr R, Clarsen B, Ekstrand J (2018) Why we should focus on the burden of injuries and illnesses, not just their incidence. Br J Sports Med 52:1018–1021. https://doi.org/10.1136/bjsports-2017-098160

Bahr R, Krosshaug T (2005) Understanding injury mechanisms: a key component of preventing injuries in sport. Br J Sports Med 39:324–329. https://doi.org/10.1136/bjsm.2005.018341

Article   CAS   PubMed   PubMed Central   Google Scholar  

Bartlett JD, O’Connor F, Pitchford N, Torres-Ronda L, Robertson SJ (2017) Relationships Between Internal and External Training Load in Team-Sport Athletes: Evidence for an Individualized Approach. Int J Sports Physiol Perform 12:230–234. https://doi.org/10.1123/ijspp.2015-0791

Bittencourt NFN, Meeuwisse WH, Mendonça LD, Nettel-Aguirre A, Ocarino JM, Fonseca ST (2016) Complex systems approach for sports injuries: moving from risk factor identification to injury pattern recognition-narrative review and new concept. Br J Sports Med 50:1309–1314. https://doi.org/10.1136/bjsports-2015-095850

Article   CAS   PubMed   Google Scholar  

Bolling C, van Mechelen W, Pasman HR, Verhagen E (2018) Context Matters: Revisiting the First Step of the “Sequence of Prevention” of Sports Injuries. Sports Med Auckl NZ 48:2227–2234. https://doi.org/10.1007/s40279-018-0953-x

Bolling C, Mellette J, Pasman HR, van Mechelen W, Verhagen E (2019) From the safety net to the injury prevention web: applying systems thinking to unravel injury prevention challenges and opportunities in Cirque du Soleil. BMJ Open Sport Exerc Med 5:e000492. https://doi.org/10.1136/bmjsem-2018-000492

Article   PubMed   PubMed Central   Google Scholar  

Cabitza F, Locoro A, Banfi G (2018) Machine Learning in Orthopedics: A Literature ReviewFront Bioeng Biotechnol 6. https://doi.org/10.3389/fbioe.2018.00075

Carey DL, Crossley KM, Whiteley R, Mosler A, Ong K-L, Crow J, Morris ME (2018) Modeling Training Loads and Injuries: The Dangers of Discretization. Med Sci Sports Exerc 50:2267–2276. https://doi.org/10.1249/MSS.0000000000001685

Carey DL, Ong K, Whiteley R, Crossley KM, Crow J, Morris ME (2018) Predictive modelling of training loads and injury in Australian football. Int J ComputSci Sport 17:49–66

Claudino JG, de Capanema D, O, de Souza TV, Serrão JC, Machado Pereira AC, Nassis GP, (2019) Current Approaches to the Use of Artificial Intelligence for Injury Risk Assessment and Performance Prediction in Team Sports: a Systematic Review. Sports Med - Open 5:28. https://doi.org/10.1186/s40798-019-0202-3

Emery CA, Pasanen K (2019) Current trends in sport injury prevention. Best Pract Res ClinRheumatol 33:3–15. https://doi.org/10.1016/j.berh.2019.02.009

Ertelt T, Solomonovs I, Gronwald T (2018) Enhancement of force patterns classification based on Gaussian distributions. J Biomech 67:144–149. https://doi.org/10.1016/j.jbiomech.2017.12.006

Fonseca ST, Souza TR, Verhagen E, van Emmerik R, Bittencourt NFN, Mendonça LDM, Andrade AGP, Resende RA, Ocarino JM (2020) Sports Injury Forecasting and Complexity: A Synergetic Approach. Sports Med Auckl NZ 50:1757–1770. https://doi.org/10.1007/s40279-020-01326-4

Gastin PB, Hunkin SL, Fahrner B, Robertson S (2019) Deceleration, Acceleration, and Impacts Are Strong Contributors to Muscle Damage in Professional Australian Football. J Strength Cond Res 33:3374–3383. https://doi.org/10.1519/JSC.0000000000003023

Groll A, Ley C, Schauberger G, Eetvelde HV (2019) A hybrid random forest to predict soccer matches in international tournaments. J Quant Anal Sports 15:271–287. https://doi.org/10.1515/jqas-2018-0060

Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, Schünemann HJ (2008) GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 336:924–926. https://doi.org/10.1136/bmj.39489.470347.AD

Hasler RM, Berov S, Benneker L, Dubler S, Spycher J, Heim D, Zimmermann H, Exadaktylos AK (2010) Are there risk factors for snowboard injuries? A case-control multicentre study of 559 snowboarders. Br J Sports Med 44:816–821. https://doi.org/10.1136/bjsm.2010.071357

Hasler RM, Dubler S, Benneker LM, Berov S, Spycher J, Heim D, Zimmermann H, Exadaktylos AK (2009) Are there risk factors in alpine skiing? A controlled multicentre survey of 1278 skiers. Br J Sports Med 43:1020–1025. https://doi.org/10.1136/bjsm.2009.064741

Hubáček O, Šourek G, Železný F (2019) Learning to predict soccer results from relational data with gradient boosted trees. Mach Learn 108:29–47. https://doi.org/10.1007/s10994-018-5704-6

Klein C, Luig P, Henke T, Platen P (2020) Injury burden differs considerably between single teams from German professional male football (soccer): surveillance of three consecutive seasons. Knee Surg Sports TraumatolArthrosc Off J ESSKA 28:1656–1664. https://doi.org/10.1007/s00167-019-05623-y

Kuhn M, Johnson K (2013) Applied predictive modeling. Springer, New York

Liu Y, Chen P-HC, Krause J, Peng L (2019) How to Read Articles That Use Machine Learning: Users’ Guides to the Medical Literature. JAMA 322:1806–1816. https://doi.org/10.1001/jama.2019.16489

López-Valenciano A, Ayala F, PuertaJosM DE, Ste Croix MBA, Vera-Garcia FJ, Hernández-Sánchez S, Ruiz-Pérez I, Myer GD (2018) A Preventive Model for Muscle Injuries: A Novel Approach based on Learning Algorithms. Med Sci Sports Exerc 50:915–927. https://doi.org/10.1249/MSS.0000000000001535

Lundberg SM, Lee S-I (2017) A Unified Approach to Interpreting Model Predictions. In: Guyon I, Luxburg UV, Bengio S, Wallach H, Fergus R, Vishwanathan S, Garnett R (eds) 30th Conference on Neural Information Processing Systems (NIPS 2017). Curran Associates, Inc., LongBeach, p 4765–4774

McCullagh J, Whitfort T (2013) An investigation into the application of Artificial Neural Networks to the prediction of injuries in sport. Int J Sport Health Sci 7:356–360

Google Scholar  

Meeuwisse WH, Tyreman H, Hagel B, Emery C (2007) A dynamic model of etiology in sport injury: the recursive nature of risk and causation. Clin J Sport Med Off J Can Acad Sport Med 17:215–219. https://doi.org/10.1097/JSM.0b013e3180592a48

Mendonça LD, Ocarino JM, Bittencourt NFN, Macedo LG, Fonseca ST (2018) Association of Hip and Foot Factors With Patellar Tendinopathy (Jumper’s Knee) in Athletes. J Orthop Sports PhysTher 48:676–684. https://doi.org/10.2519/jospt.2018.7426

Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group (2009) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 6:e1000097. https://doi.org/10.1371/journal.pmed.1000097

Myers TG, Ramkumar PN, Ricciardi BF, Urish KL, Kipper J, Ketonis C (2020) Artificial Intelligence and Orthopaedics: An Introduction for Clinicians. JBJS 102:830–840. https://doi.org/10.2106/JBJS.19.01128

Oliver JL, Ayala F, De Ste Croix MBA, Lloyd RS, Myer GD, Read PJ (2020) Using machine learning to improve our understanding of injury risk and prediction in elite male youth football players. J Sci Med Sport. https://doi.org/10.1016/j.jsams.2020.04.021

Parker W, Forster BB (2019) Artificial intelligence in sports medicine radiology: what’s coming? Br J Sports Med 53:1201–1202. https://doi.org/10.1136/bjsports-2018-099999

Rodas G, Osaba L, Arteta D, Pruna R, Fernández D, Lucia A (2019) Genomic Prediction of Tendinopathy Risk in Elite Team Sports. Int J Sports Physiol Perform: 1–7. https://doi.org/10.1123/ijspp.2019-0431

Rommers N, RÖssler R, Verhagen E, Vandecasteele F, Verstockt S, Vaeyens R, Lenoir M, D’Hondt E, Witvrouw E, (2020) A Machine Learning Approach to Assess Injury Risk in Elite Youth Football Players. Med Sci Sports Exerc 52:1745–1751. https://doi.org/10.1249/MSS.0000000000002305

Rossi A, Pappalardo L, Cintia P, Iaia FM, Fernàndez J, Medina D (2018) Effective injury forecasting in soccer with GPS training data and machine learning. PLoS ONE 13:e0201264. https://doi.org/10.1371/journal.pone.0201264

Ruddy JD, Cormack SJ, Whiteley R, Williams MD, Timmins RG, Opar DA (2019) Modeling the Risk of Team Sport Injuries: A Narrative Review of Different Statistical Approaches. Front Physiol 10:829. https://doi.org/10.3389/fphys.2019.00829

Ruddy JD, Shield AJ, Maniar N, Williams MD, Duhig S, Timmins RG, Hickey J, Bourne MN, Opar DA (2018) Predictive Modeling of Hamstring Strain Injuries in Elite Australian Footballers. Med Sci Sports Exerc 50:906–914. https://doi.org/10.1249/MSS.0000000000001527

Shah P, Kendall F, Khozin S, Goosen R, Hu J, Laramie J, Ringel M, Schork N (2019) Artificial intelligence and machine learning in clinical development: a translational perspective. NPJ Digit Med 2:69. https://doi.org/10.1038/s41746-019-0148-3

Tervo T, Ermling J, Nordström A, Toss F (2020) The 9+ screening test score does not predict injuries in elite floorball players. Scand J Med Sci Sports 30:1232–1236. https://doi.org/10.1111/sms.13663

Thornton HR, Delaney JA, Duthie GM, Dascombe BJ (2017) Importance of Various Training-Load Measures in Injury Incidence of Professional Rugby League Athletes. Int J Sports Physiol Perform 12:819–824. https://doi.org/10.1123/ijspp.2016-0326

Topol EJ (2019) High-performance medicine: the convergence of human and artificial intelligence. Nat Med 25:44–56. https://doi.org/10.1038/s41591-018-0300-7

Trinidad-Fernandez M, Gonzalez-Sanchez M, Cuesta-Vargas AI (2019) Is a low Functional Movement Screen score (≤14/21) associated with injuries in sport? A systematic review and meta-analysis. BMJ Open Sport Exerc Med 5:e000501. https://doi.org/10.1136/bmjsem-2018-000501

Verhagen E, Bolling C (2015) Protecting the health of the @hlete: how online technology may aid our common goal to prevent injury and illness in sport. Br J Sports Med 49:1174–1178. https://doi.org/10.1136/bjsports-2014-094322

Wells G, Shea B, O’Connell D, Robertson J, Peterson J, Welch V, Losos M, Tugwell P The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomized Studies in Meta- Analysis. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp . Accessed 28 June 2020

Whiteside D, Martini DN, Lepley AS, Zernicke RF, Goulet GC (2016) Predictors of Ulnar Collateral Ligament Reconstruction in Major League Baseball Pitchers. Am J Sports Med 44:2202–2209. https://doi.org/10.1177/0363546516643812

Download references

Acknowledgements

Not applicable

No special funding existed for the performed study.

Author information

Authors and affiliations.

Department of Applied Mathematics, Computer Science and Statistics, Ghent University, Krijgslaan 281-S9, 9000, Ghent, Belgium

Hans Van Eetvelde & Christophe Ley

Graduate Program in Rehabilitation and Functional Performance, Universidade Federal Dos Vales Do Jequitinhonha E Mucuri (UFVJM), Diamantina, Minas Gerais, Brazil

Luciana D. Mendonça

Department of Physical Therapy and Motor Rehabilitation, Faculty of Medicine and Health Sciences, Ghent University, Ghent, Belgium

Ministry of Education of Brazil, CAPES Foundation, Brasília, Distrito Federal, Brazil

Department of Orthopaedic Surgery, Centre Hospitalier Luxembourg and Luxembourg Institute of Health, Luxembourg, Luxembourg

Romain Seil

Department of Orthopaedic Surgery, University of Rostock, Rostock, Germany

Thomas Tischer

You can also search for this author in PubMed   Google Scholar

Contributions

TT, HVE and LDM designed the search strategy and the overall setup of the systematic review. HVE and CL performed the technical analysis of the included studies. TT and LDM performed the methodological quality analysis of the articles. All authors drafted the manuscript and performed multiple revisions. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Hans Van Eetvelde .

Ethics declarations

Ethics approval and consent to participate, consent for publication, competing interests.

The authors declare that they have no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: s1..

Definitions of core terms important for AI application in sport injury prediction and prevention.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Van Eetvelde, H., Mendonça, L.D., Ley, C. et al. Machine learning methods in sport injury prediction and prevention: a systematic review. J EXP ORTOP 8 , 27 (2021). https://doi.org/10.1186/s40634-021-00346-x

Download citation

Received : 25 January 2021

Accepted : 15 March 2021

Published : 14 April 2021

DOI : https://doi.org/10.1186/s40634-021-00346-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Machine Learning
• Injury prediction
• Injury prevention
• Sport injury

Log in using your username and password

• Search More Search for this keyword Advanced search
• Latest content
• Current issue
• For authors
• New editors
• BMJ Journals More You are viewing from: Google Indexer

You are here

• Volume 34, Issue 2
• Sports injury incidence
• Article Text
• Article info
• Citation Tools
• Rapid Responses
• Article metrics

• Lisa Hodgson Phillips
• Centre for Sports Medicine, Department of Orthopaedic and Accident Surgery, Queen's Medical Centre, University Hospital, Nottingham NG7 2UH, United Kingdom
• Correspondence to: L Hodgson Phillips

https://doi.org/10.1136/bjsm.34.2.133

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

The aim of this paper is to give a “medical” viewpoint on sports injury data collection and analysis, and to emphasise the importance of epidemiological sports data collection with regard to incidence rates and exposure risk hours and highlight the need for uniform definitions within and across sport. It is designed not as a statistical or epidemiological paper but as a resource to be used by those involved in sports injury research so that they may confidently analyse, evaluate, and compare existing research and to enable them to collect accurate sports injury data in their own field.

Introduction

Sports injuries occur when athletes are exposed to their given sport and they occur under specific conditions, at a known time and place.

The last point should relate to time missed in training days as well as competitive participation and may also consider time lost to work in the case of a semiprofessional athlete. The knowledge gained from asking these questions may help us to predict and thus prevent injury.

In sports medicine, we are thus all epidemiologists “concerned with quantifying injury occurrence with respect to who is affected by injury, where and when injuries occur and what is their outcome—for the purposes of explaining why and how injuries occur and identifying strategies to control and prevent them”. 1

To interpret the literature, we must be able to discern good studies from bad, to verify whether conclusions of a particular study are valid, and to understand the limitations of a study. 2 Many studies are limited because the data collection is for injured athletes alone or risk factors alone, which does not allow the use of the epidemiological concept of athletes being at risk. There is no common operational definition of sports injury in existence at present and furthermore no set definition of severity. Some studies classify a severe injury as one that results in five games being missed, whereas others classify a severe injury as one that requires five weeks out of competition to heal; these clearly are not compatible for comparisons of sports for which more than one game is played each week. 3, 4 There is currently no set format for data collection across sports, and the size of the samples vary: some studies refer to only one team and others use multiple teams. 3, 5– 7 Therefore methodological factors alter the perception and interpretation of incidence rates.

When examining sports injury data the questions typically asked are:

is there a greater risk in one certain sport?

is there a common site and type of injury in a given sport?

who is at most risk in a team sport?

what is the participation time missed as the result of that specific injury?

The fundamental unit of measurement is rate. To calculate a valid injury rate, the number of injuries experienced (numerator data) is linked to a suitable denominator measure of the amount of athletic exposure to the risk of injury. Thus a rate consists of a denominator and a numerator over a period of time. Denominator data can be a number of different things; they could be the number of athletes in a club or team, the number of games played, the number of minutes played, or the number of player appearances. The choice of the denominator affects the numerical value of the derived data and also their interpretation. For example, injuries can be expressed as the number of injuries per game, an injury every so many minutes of play, or the number of injuries per ( x ) player appearances. 8

Incidence rates

Incidence is the most basic expression of risk. Incidence rates pertain to the number of new injuries that occur in a population at risk over a specified time period or the number of new injuries during a period divided by the total number of sportspeople at that period. Thus the epidemiological concept of athletic exposure in games or training is multiplied by the number of players participating. Incidence rates that do not consider exposure do not reliably indicate the problem and cannot be used to compare injury incidence.

Determining incidence rates

Accurate and consistent medical diagnosis is imperative. Diagnoses may be made by the doctor or physiotherapist but must be consistent throughout, with the use of set codes for site, nature, and severity of injury. All injuries should be recorded, including transient injuries—that is, those that require medical attention but result in no time lost to training or playing. Time lost from participation must be recorded accurately, using both training and game/competitive participation data, in days lost as well as games and weeks lost. Many studies exclude training injuries and training time lost, using only those injuries that occur in a game or that require a competitive game to be missed. 3, 9 These studies lose valuable data and fail to portray the true injury picture of the sport. If training information is excluded, then the data only represent the tip of the iceberg—submerged missed data may include the effects of training injuries or, more importantly, the training time lost on the player, his/her fitness, and ultimately his/her career. The same argument can be used to show the importance of including transient injuries in the data analysis. Excluding these injuries gives a false picture of the injuries sustained in a given sport.

Coding of injury diagnosis

Coding and recording of injuries should be through the consistent use of a set of established definitions of injury, which are expansive and descriptive to avoid subjectivity. Standard classifications of diagnoses are in existence such as the International Classification of Diseases; however, these are often not specific enough and thus not of any use for sports injury data collection. In contrast, there is the Orchard Codes system, which is very descriptive and expansive and may be used in this type of research. A single person should record the information where possible to achieve greatest intra-rater reliability. Time lost from sport participation must be considered an objective measure that is not sensitive to the concept of returning to play when the athlete is not fully healed and must always be referred to as a filter when conclusions on sports injury data are drawn. Athletes are often paid professionals and as such do not wish to miss a training or competitive/playing session, which could result in loss of their team place in the next game or their wage at the end of the week. Athletes are eager to participate and thus always challenge the healing process as they almost always aim to return to competition much sooner than the lay person. 10 We do not have any reliable criteria on return to sport.

Study design

The US Preventative Services Task Force in 1989 established a hierarchy of evidence in which greater weight was given to study designs in decreasing order of importance. 11 Random control trials were rated first; these expose some subjects, but not others, to an intervention—for example, risk of injury. Therefore this type of research is more clinical in nature and not typically appropriate for the study of injury patterns. Cohort studies were rated next; this type of study monitors both injured and non-injured athletes, thereby providing results on the effects of participation, and are ideally prospective in nature. Case-control was the third type of study, monitoring only those athletes who suffered an injury and are typically more retrospective in nature. These make up the vast majority of sports injury studies at present; however, we should recognise that multiple anecdotes do not add up to an evidence base.

Weaknesses in sports injury epidemiology research

Retrospective data are used which may lead to bias.

Multiple injury recorders leading to a lower inter-rater reliability.

Single or part season's data analysed.

Single team analysed.

Injury cases documented are not adjusted for exposure risk hours of training or playing.

Comparisons made with other studies that have not used the same injury coding or methodology (may not even be of the same sport).

Studies should have validity and reliability. The former is defined as the extent to which you measure what you intended to measure and is usually compared against a yardstick. Sports injury incidence at present has no yardstick against which comparisons can be made. Reliability is the ability to produce the same results on more than one occasion and is dependent on inter-rater or intra-rater data collection. For accurate injury incidence, reliability is imperative. 12

Sample size influences results. It is impossible to compare studies in which various sample sizes—that is, one team or many—have been used, unless adjustments for exposure have been made and this is clearly stated in the methods. Studies on one particular sports team, however, can be powerful if the number of injuries incurred is large enough to show statistical significance. 13

The type of statistical analysis is directly related to the methodology of the study. For example, the χ 2 test can be used to assess the differences between observed and expected injuries in a season or number of seasons. Multiple regression and multiple variate analyses may be chosen to assess the influence of independent factors on the injuries incurred—for example, the player position or the hardness of the ground. The calculation of incidence rates has been identified as a critical feature of sound epidemiological sports injury studies. 14

As a footnote, it should be mentioned that any patient injury information collected must always be confidential.

Exposure risk hours and rates per 1000 hours

The way in which incidence is expressed has also been shown to affect the calculation/interpretation of incidence rates. Increasingly, incidence rates in all sports are being expressed as rates per 1000 hours. This is a good approach and allows some comparison across sports. However, a further refinement of the calculation of incidence rates is to measure the actual exposure time at risk. Thus expected injuries are calculated using player exposure/risk hours. These risk hours should ideally include training time as well as competitive participation. 13, 15

The following is an example of how exposure/risk hours are calculated in a team sport, specifically rugby league. There are 13 players of one team on the field at any one time. The duration of the game is 80 minutes (1.33 hours). Thus there are 17.33 player exposure/risk hours per team per game of rugby league (13 × 1.33). Over an average season—for example, 30 games—there may be 520 player exposure/risk hours (13 × 1.33 × 30).

To calculate the incidence in relation to these exposure hours, the total number of injuries recorded over a period is divided by the total exposure for that period, and the result multiplied by 1000 to obtain the rate per 1000 hours. This period could be one game, several games, or a whole season or number of seasons. To see if there are significant differences across games or seasons, observed and expected injuries can be used. Observed injuries are those recorded over the period under consideration. Expected injuries are calculated by dividing the total injuries—for example, over four seasons—by the total exposure—for example, for the same four seasons—and multiplying the result by the exposure for the period under consideration—for example, one season only—giving an expected injury case for that one season. Significance tests may then be applied.

The relevance of recording and analysing data in this way is shown below taking data from a previous study. 13 Figure 1 shows the number of injury cases recorded over four rugby league seasons at one British professional rugby league club (1993–1996 inclusive). On initial observation, there does not appear to be a significant difference across the four seasons, and the observer may even say that the injuries were in fact lower over the last two seasons. However, in fig 2, which is for the same four seasons but the data are adjusted for exposure/risk hours and presented as rates per 1000 hours, the true picture is disclosed. An obvious increase in injury incidence is seen. In truth, in the 1996 season, the incidence of injury was almost double that of the first season recorded (1993/1994). Excluding exposure time at risk prevents the true picture from being seen. This can be highlighted by the fact that, during the 1993/1994 season, there were 35 games played (605.15 exposure hours) and in 1996 only 21 games were played (363.09 exposure hours); however, observe the difference in injury incidence again. Not adjusting for exposure/risk hours but only commenting on total injury cases is a fatal flaw in sports injury data presentation.

• Download figure
• Open in new tab
• Download powerpoint

Game injury statistics showing number of injury cases per season.

Game injury statistics showing rates per 1000 hours per season.

Strengths in sports injury epidemiology research

Using one recorder to diagnose and document injuries gives a high intra-rater reliability.

Incidence rates are used and adjusted for exposure.

Training injuries are included.

Time lost to competitive participation plus time lost to training and work also documented.

Prospective studies conducted using descriptive set injury coding definitions and methodology.

Filters recognised and referred to.

Comparisons made with similar studies but acknowledging the differences in diagnostic coding and definitions of severity.

Acknowledging where professional sport is compared with amateur sport.

Using more than one team where possible: improved generalisability.

If we apply the above to what we already know clinically, we may help to predict and prevent future injury occurrence. Thus accurate data collection could be essential in the prevention of injuries. If specific influences are identified as a contributing factor to the risk of injury and supported by scientific data collection, then the rules of the sport may be changed to prevent this happening again. This will have the effect of making our athletes as injury free as possible and may even help to lengthen their time in competitive participation.

Ideal study

Cohort design (injured and non-injured athletes observed).

Conducted over several teams.

Longitudinal prospective data collection.

One recorder where possible (high intra-rater reliability).

Uniformity of injury definition across sports.

Specific definitions of injury severity so comparisons between studies can be made accurately.

Exposure hours used to express incidence rates for competitive participation and training.

Acknowledgement of existing filters.

Acknowledgments

I would like to thank Dr Mark E Batt for his help and advice on preparing this paper.

• ↵ Caine CG, Caine DJ, Lindner KJ. The epidemiological approach to sports injuries. In: Caine DJ, Caine CG, Lindner KJ, eds. Epidemiology of sports injuries. Champaign, IL: Human Kinetics, 1996:1–13.
• ↵ Campbell M, Machin D. Medical statistics . Chichester: John Wiley & Sons, 1990:1.
• ↵ Gibbs N. Injuries in professional rugby league. A 3 year prospective study. Am J Sports Med 1993 ; 21 : 696 –700. OpenUrl Abstract / FREE Full Text
• ↵ Estell J, Shenstone B, Barnsley L. Frequency of injuries in different age groups in an elite rugby league club. Aust J Sci Med Sport 1995 ; 27 : 95 –7. OpenUrl PubMed
• ↵ Walker RD. Sports injuries: rugby league may be less dangerous than rugby union. The Practitioner 1985 ; 229 : 205 –6. OpenUrl PubMed Web of Science
• Garraway M, Macleod D. Epidemiology of rugby football injuries. Lancet 1995 ; 345 : 1485 –7. OpenUrl CrossRef PubMed Web of Science
• ↵ Norton R, Wilson MA. Rugby league injuries and patterns. New Zealand Journal of Sports Medicine 1995 ; 22 : 37 –8. OpenUrl
• ↵ Schootman M, Powell JW, Torner JC. Study designs and potential biases in sports injury research: the case control study. Sports Med 1994 ; 18 : 22 –37. OpenUrl CrossRef PubMed Web of Science
• ↵ Alexander D, Kennedy M, Kennedy J. Injuries in rugby league football. Med J Aust 1979 ; 2 : 341 –2. OpenUrl PubMed Web of Science
• ↵ Schootman M, Powell JW, Albright JP. Statistics in sports injury research. In: Delee JC, Drez D, eds. Orthopaedic sports medicine. London: WB Saunders, 1994.
• ↵ US Preventative Services Task Force. Guide to clinical preventative services: an assessment of the effectiveness of 169 interventions. Report of the US Preventative Services Task Force. Sydney: Williams & Wilkins, 1989.
• ↵ Lower T. Injury data collection in the rugby codes. Aust J Sci Med Sport 1995 ; 27 : 43 –7. OpenUrl PubMed
• ↵ Hodgson Phillips L, Standen P, Batt ME. Effects of seasonal change in rugby league on the incidence of injury. Br J Sports Med 1998 ; 32 : 144 –8. OpenUrl Abstract / FREE Full Text
• ↵ Wallace RB. The numerator, denominator and the population at risk. Am J Sports Med 1988 ; 15 : 55 –6. OpenUrl
• ↵ Hodgson Phillips L. The effects of seasonal change in rugby league on the incidence of injury. MSc thesis, 1996.

Read the full text or download the PDF:

• Search Menu

Sign in through your institution

• Browse content in Arts and Humanities
• Browse content in Archaeology
• Anglo-Saxon and Medieval Archaeology
• Archaeological Methodology and Techniques
• Archaeology by Region
• Archaeology of Religion
• Archaeology of Trade and Exchange
• Biblical Archaeology
• Contemporary and Public Archaeology
• Environmental Archaeology
• Historical Archaeology
• History and Theory of Archaeology
• Industrial Archaeology
• Landscape Archaeology
• Mortuary Archaeology
• Prehistoric Archaeology
• Underwater Archaeology
• Zooarchaeology
• Browse content in Architecture
• Architectural Structure and Design
• History of Architecture
• Residential and Domestic Buildings
• Theory of Architecture
• Browse content in Art
• Art Subjects and Themes
• History of Art
• Industrial and Commercial Art
• Theory of Art
• Biographical Studies
• Byzantine Studies
• Browse content in Classical Studies
• Classical History
• Classical Philosophy
• Classical Mythology
• Classical Literature
• Classical Reception
• Classical Art and Architecture
• Classical Oratory and Rhetoric
• Greek and Roman Papyrology
• Greek and Roman Epigraphy
• Greek and Roman Law
• Greek and Roman Archaeology
• Late Antiquity
• Religion in the Ancient World
• Digital Humanities
• Browse content in History
• Colonialism and Imperialism
• Diplomatic History
• Environmental History
• Genealogy, Heraldry, Names, and Honours
• Genocide and Ethnic Cleansing
• Historical Geography
• History by Period
• History of Emotions
• History of Agriculture
• History of Education
• History of Gender and Sexuality
• Industrial History
• Intellectual History
• International History
• Labour History
• Legal and Constitutional History
• Local and Family History
• Maritime History
• Military History
• National Liberation and Post-Colonialism
• Oral History
• Political History
• Public History
• Regional and National History
• Revolutions and Rebellions
• Slavery and Abolition of Slavery
• Social and Cultural History
• Theory, Methods, and Historiography
• Urban History
• World History
• Browse content in Language Teaching and Learning
• Language Learning (Specific Skills)
• Language Teaching Theory and Methods
• Browse content in Linguistics
• Applied Linguistics
• Cognitive Linguistics
• Computational Linguistics
• Forensic Linguistics
• Grammar, Syntax and Morphology
• Historical and Diachronic Linguistics
• History of English
• Language Evolution
• Language Reference
• Language Acquisition
• Language Variation
• Language Families
• Lexicography
• Linguistic Anthropology
• Linguistic Theories
• Linguistic Typology
• Phonetics and Phonology
• Psycholinguistics
• Sociolinguistics
• Translation and Interpretation
• Writing Systems
• Browse content in Literature
• Bibliography
• Children's Literature Studies
• Literary Studies (Romanticism)
• Literary Studies (American)
• Literary Studies (Asian)
• Literary Studies (European)
• Literary Studies (Eco-criticism)
• Literary Studies (Modernism)
• Literary Studies - World
• Literary Studies (1500 to 1800)
• Literary Studies (19th Century)
• Literary Studies (20th Century onwards)
• Literary Studies (African American Literature)
• Literary Studies (British and Irish)
• Literary Studies (Early and Medieval)
• Literary Studies (Fiction, Novelists, and Prose Writers)
• Literary Studies (Gender Studies)
• Literary Studies (Graphic Novels)
• Literary Studies (History of the Book)
• Literary Studies (Plays and Playwrights)
• Literary Studies (Poetry and Poets)
• Literary Studies (Postcolonial Literature)
• Literary Studies (Queer Studies)
• Literary Studies (Science Fiction)
• Literary Studies (Travel Literature)
• Literary Studies (War Literature)
• Literary Studies (Women's Writing)
• Literary Theory and Cultural Studies
• Mythology and Folklore
• Shakespeare Studies and Criticism
• Browse content in Media Studies
• Browse content in Music
• Applied Music
• Dance and Music
• Ethics in Music
• Ethnomusicology
• Gender and Sexuality in Music
• Medicine and Music
• Music Cultures
• Music and Media
• Music and Religion
• Music and Culture
• Music Education and Pedagogy
• Music Theory and Analysis
• Musical Scores, Lyrics, and Libretti
• Musical Structures, Styles, and Techniques
• Musicology and Music History
• Performance Practice and Studies
• Race and Ethnicity in Music
• Sound Studies
• Browse content in Performing Arts
• Browse content in Philosophy
• Aesthetics and Philosophy of Art
• Epistemology
• Feminist Philosophy
• History of Western Philosophy
• Metaphysics
• Moral Philosophy
• Non-Western Philosophy
• Philosophy of Language
• Philosophy of Mind
• Philosophy of Perception
• Philosophy of Science
• Philosophy of Action
• Philosophy of Law
• Philosophy of Religion
• Philosophy of Mathematics and Logic
• Practical Ethics
• Social and Political Philosophy
• Browse content in Religion
• Biblical Studies
• Christianity
• East Asian Religions
• History of Religion
• Judaism and Jewish Studies
• Qumran Studies
• Religion and Education
• Religion and Health
• Religion and Politics
• Religion and Science
• Religion and Law
• Religion and Art, Literature, and Music
• Religious Studies
• Browse content in Society and Culture
• Cookery, Food, and Drink
• Cultural Studies
• Customs and Traditions
• Ethical Issues and Debates
• Hobbies, Games, Arts and Crafts
• Natural world, Country Life, and Pets
• Popular Beliefs and Controversial Knowledge
• Sports and Outdoor Recreation
• Technology and Society
• Travel and Holiday
• Visual Culture
• Browse content in Law
• Arbitration
• Browse content in Company and Commercial Law
• Commercial Law
• Company Law
• Browse content in Comparative Law
• Systems of Law
• Competition Law
• Browse content in Constitutional and Administrative Law
• Government Powers
• Judicial Review
• Local Government Law
• Military and Defence Law
• Parliamentary and Legislative Practice
• Construction Law
• Contract Law
• Browse content in Criminal Law
• Criminal Procedure
• Criminal Evidence Law
• Sentencing and Punishment
• Employment and Labour Law
• Environment and Energy Law
• Browse content in Financial Law
• Banking Law
• Insolvency Law
• History of Law
• Human Rights and Immigration
• Intellectual Property Law
• Browse content in International Law
• Private International Law and Conflict of Laws
• Public International Law
• IT and Communications Law
• Jurisprudence and Philosophy of Law
• Law and Politics
• Law and Society
• Browse content in Legal System and Practice
• Courts and Procedure
• Legal Skills and Practice
• Primary Sources of Law
• Regulation of Legal Profession
• Medical and Healthcare Law
• Browse content in Policing
• Criminal Investigation and Detection
• Police and Security Services
• Police Procedure and Law
• Police Regional Planning
• Browse content in Property Law
• Personal Property Law
• Study and Revision
• Terrorism and National Security Law
• Browse content in Trusts Law
• Wills and Probate or Succession
• Browse content in Medicine and Health
• Browse content in Allied Health Professions
• Arts Therapies
• Clinical Science
• Dietetics and Nutrition
• Occupational Therapy
• Operating Department Practice
• Physiotherapy
• Radiography
• Speech and Language Therapy
• Browse content in Anaesthetics
• General Anaesthesia
• Neuroanaesthesia
• Clinical Neuroscience
• Browse content in Clinical Medicine
• Acute Medicine
• Cardiovascular Medicine
• Clinical Genetics
• Clinical Pharmacology and Therapeutics
• Dermatology
• Endocrinology and Diabetes
• Gastroenterology
• Genito-urinary Medicine
• Geriatric Medicine
• Infectious Diseases
• Medical Toxicology
• Medical Oncology
• Pain Medicine
• Palliative Medicine
• Rehabilitation Medicine
• Respiratory Medicine and Pulmonology
• Rheumatology
• Sleep Medicine
• Sports and Exercise Medicine
• Community Medical Services
• Critical Care
• Emergency Medicine
• Forensic Medicine
• Haematology
• History of Medicine
• Browse content in Medical Skills
• Clinical Skills
• Communication Skills
• Nursing Skills
• Surgical Skills
• Browse content in Medical Dentistry
• Oral and Maxillofacial Surgery
• Paediatric Dentistry
• Restorative Dentistry and Orthodontics
• Surgical Dentistry
• Medical Ethics
• Medical Statistics and Methodology
• Browse content in Neurology
• Clinical Neurophysiology
• Neuropathology
• Nursing Studies
• Browse content in Obstetrics and Gynaecology
• Gynaecology
• Occupational Medicine
• Ophthalmology
• Otolaryngology (ENT)
• Browse content in Paediatrics
• Neonatology
• Browse content in Pathology
• Chemical Pathology
• Clinical Cytogenetics and Molecular Genetics
• Histopathology
• Medical Microbiology and Virology
• Patient Education and Information
• Browse content in Pharmacology
• Psychopharmacology
• Browse content in Popular Health
• Caring for Others
• Complementary and Alternative Medicine
• Self-help and Personal Development
• Browse content in Preclinical Medicine
• Cell Biology
• Molecular Biology and Genetics
• Reproduction, Growth and Development
• Primary Care
• Professional Development in Medicine
• Browse content in Psychiatry
• Addiction Medicine
• Child and Adolescent Psychiatry
• Forensic Psychiatry
• Learning Disabilities
• Old Age Psychiatry
• Psychotherapy
• Browse content in Public Health and Epidemiology
• Epidemiology
• Public Health
• Browse content in Radiology
• Clinical Radiology
• Interventional Radiology
• Nuclear Medicine
• Radiation Oncology
• Reproductive Medicine
• Browse content in Surgery
• Cardiothoracic Surgery
• Gastro-intestinal and Colorectal Surgery
• General Surgery
• Neurosurgery
• Paediatric Surgery
• Peri-operative Care
• Plastic and Reconstructive Surgery
• Surgical Oncology
• Transplant Surgery
• Trauma and Orthopaedic Surgery
• Vascular Surgery
• Browse content in Science and Mathematics
• Browse content in Biological Sciences
• Aquatic Biology
• Biochemistry
• Bioinformatics and Computational Biology
• Developmental Biology
• Ecology and Conservation
• Evolutionary Biology
• Genetics and Genomics
• Microbiology
• Molecular and Cell Biology
• Natural History
• Plant Sciences and Forestry
• Research Methods in Life Sciences
• Structural Biology
• Systems Biology
• Zoology and Animal Sciences
• Browse content in Chemistry
• Analytical Chemistry
• Computational Chemistry
• Crystallography
• Environmental Chemistry
• Industrial Chemistry
• Inorganic Chemistry
• Materials Chemistry
• Medicinal Chemistry
• Mineralogy and Gems
• Organic Chemistry
• Physical Chemistry
• Polymer Chemistry
• Study and Communication Skills in Chemistry
• Theoretical Chemistry
• Browse content in Computer Science
• Artificial Intelligence
• Computer Architecture and Logic Design
• Game Studies
• Human-Computer Interaction
• Mathematical Theory of Computation
• Programming Languages
• Software Engineering
• Systems Analysis and Design
• Virtual Reality
• Browse content in Computing
• Business Applications
• Computer Security
• Computer Games
• Computer Networking and Communications
• Digital Lifestyle
• Graphical and Digital Media Applications
• Operating Systems
• Browse content in Earth Sciences and Geography
• Atmospheric Sciences
• Environmental Geography
• Geology and the Lithosphere
• Maps and Map-making
• Meteorology and Climatology
• Oceanography and Hydrology
• Palaeontology
• Physical Geography and Topography
• Regional Geography
• Soil Science
• Urban Geography
• Browse content in Engineering and Technology
• Agriculture and Farming
• Biological Engineering
• Civil Engineering, Surveying, and Building
• Electronics and Communications Engineering
• Energy Technology
• Engineering (General)
• Environmental Science, Engineering, and Technology
• History of Engineering and Technology
• Mechanical Engineering and Materials
• Technology of Industrial Chemistry
• Transport Technology and Trades
• Browse content in Environmental Science
• Applied Ecology (Environmental Science)
• Conservation of the Environment (Environmental Science)
• Environmental Sustainability
• Environmentalist Thought and Ideology (Environmental Science)
• Management of Land and Natural Resources (Environmental Science)
• Natural Disasters (Environmental Science)
• Nuclear Issues (Environmental Science)
• Pollution and Threats to the Environment (Environmental Science)
• Social Impact of Environmental Issues (Environmental Science)
• History of Science and Technology
• Browse content in Materials Science
• Ceramics and Glasses
• Composite Materials
• Metals, Alloying, and Corrosion
• Nanotechnology
• Browse content in Mathematics
• Applied Mathematics
• Biomathematics and Statistics
• History of Mathematics
• Mathematical Education
• Mathematical Finance
• Mathematical Analysis
• Numerical and Computational Mathematics
• Probability and Statistics
• Pure Mathematics
• Browse content in Neuroscience
• Cognition and Behavioural Neuroscience
• Development of the Nervous System
• Disorders of the Nervous System
• History of Neuroscience
• Invertebrate Neurobiology
• Molecular and Cellular Systems
• Neuroendocrinology and Autonomic Nervous System
• Neuroscientific Techniques
• Sensory and Motor Systems
• Browse content in Physics
• Astronomy and Astrophysics
• Atomic, Molecular, and Optical Physics
• Biological and Medical Physics
• Classical Mechanics
• Computational Physics
• Condensed Matter Physics
• Electromagnetism, Optics, and Acoustics
• History of Physics
• Mathematical and Statistical Physics
• Measurement Science
• Nuclear Physics
• Particles and Fields
• Plasma Physics
• Quantum Physics
• Relativity and Gravitation
• Semiconductor and Mesoscopic Physics
• Browse content in Psychology
• Affective Sciences
• Clinical Psychology
• Cognitive Psychology
• Cognitive Neuroscience
• Criminal and Forensic Psychology
• Developmental Psychology
• Educational Psychology
• Evolutionary Psychology
• Health Psychology
• History and Systems in Psychology
• Music Psychology
• Neuropsychology
• Organizational Psychology
• Psychological Assessment and Testing
• Psychology of Human-Technology Interaction
• Psychology Professional Development and Training
• Research Methods in Psychology
• Social Psychology
• Browse content in Social Sciences
• Browse content in Anthropology
• Anthropology of Religion
• Human Evolution
• Medical Anthropology
• Physical Anthropology
• Regional Anthropology
• Social and Cultural Anthropology
• Theory and Practice of Anthropology
• Browse content in Business and Management
• Business Ethics
• Business Strategy
• Business History
• Business and Technology
• Business and Government
• Business and the Environment
• Comparative Management
• Corporate Governance
• Corporate Social Responsibility
• Entrepreneurship
• Health Management
• Human Resource Management
• Industrial and Employment Relations
• Industry Studies
• Information and Communication Technologies
• International Business
• Knowledge Management
• Management and Management Techniques
• Operations Management
• Organizational Theory and Behaviour
• Pensions and Pension Management
• Public and Nonprofit Management
• Strategic Management
• Supply Chain Management
• Browse content in Criminology and Criminal Justice
• Criminal Justice
• Criminology
• Forms of Crime
• International and Comparative Criminology
• Youth Violence and Juvenile Justice
• Development Studies
• Browse content in Economics
• Agricultural, Environmental, and Natural Resource Economics
• Asian Economics
• Behavioural Finance
• Behavioural Economics and Neuroeconomics
• Econometrics and Mathematical Economics
• Economic History
• Economic Systems
• Economic Methodology
• Economic Development and Growth
• Financial Markets
• Financial Institutions and Services
• General Economics and Teaching
• Health, Education, and Welfare
• History of Economic Thought
• International Economics
• Labour and Demographic Economics
• Law and Economics
• Macroeconomics and Monetary Economics
• Microeconomics
• Public Economics
• Urban, Rural, and Regional Economics
• Welfare Economics
• Browse content in Education
• Adult Education and Continuous Learning
• Care and Counselling of Students
• Early Childhood and Elementary Education
• Educational Equipment and Technology
• Educational Strategies and Policy
• Higher and Further Education
• Organization and Management of Education
• Philosophy and Theory of Education
• Schools Studies
• Secondary Education
• Teaching of a Specific Subject
• Teaching of Specific Groups and Special Educational Needs
• Teaching Skills and Techniques
• Browse content in Environment
• Applied Ecology (Social Science)
• Climate Change
• Conservation of the Environment (Social Science)
• Environmentalist Thought and Ideology (Social Science)
• Natural Disasters (Environment)
• Social Impact of Environmental Issues (Social Science)
• Browse content in Human Geography
• Cultural Geography
• Economic Geography
• Political Geography
• Browse content in Interdisciplinary Studies
• Communication Studies
• Museums, Libraries, and Information Sciences
• Browse content in Politics
• African Politics
• Asian Politics
• Chinese Politics
• Comparative Politics
• Conflict Politics
• Elections and Electoral Studies
• Environmental Politics
• Ethnic Politics
• European Union
• Foreign Policy
• Gender and Politics
• Human Rights and Politics
• Indian Politics
• International Relations
• International Organization (Politics)
• International Political Economy
• Irish Politics
• Latin American Politics
• Middle Eastern Politics
• Political Behaviour
• Political Economy
• Political Institutions
• Political Methodology
• Political Communication
• Political Philosophy
• Political Sociology
• Political Theory
• Politics and Law
• Politics of Development
• Public Policy
• Public Administration
• Quantitative Political Methodology
• Regional Political Studies
• Russian Politics
• Security Studies
• State and Local Government
• UK Politics
• US Politics
• Browse content in Regional and Area Studies
• African Studies
• Asian Studies
• East Asian Studies
• Japanese Studies
• Latin American Studies
• Middle Eastern Studies
• Native American Studies
• Scottish Studies
• Browse content in Research and Information
• Research Methods
• Browse content in Social Work
• Addictions and Substance Misuse
• Adoption and Fostering
• Care of the Elderly
• Child and Adolescent Social Work
• Couple and Family Social Work
• Direct Practice and Clinical Social Work
• Emergency Services
• Human Behaviour and the Social Environment
• International and Global Issues in Social Work
• Mental and Behavioural Health
• Social Justice and Human Rights
• Social Policy and Advocacy
• Social Work and Crime and Justice
• Social Work Macro Practice
• Social Work Practice Settings
• Social Work Research and Evidence-based Practice
• Welfare and Benefit Systems
• Browse content in Sociology
• Childhood Studies
• Community Development
• Comparative and Historical Sociology
• Economic Sociology
• Gender and Sexuality
• Gerontology and Ageing
• Health, Illness, and Medicine
• Marriage and the Family
• Migration Studies
• Occupations, Professions, and Work
• Organizations
• Population and Demography
• Race and Ethnicity
• Social Theory
• Social Movements and Social Change
• Social Research and Statistics
• Social Stratification, Inequality, and Mobility
• Sociology of Religion
• Sociology of Education
• Sport and Leisure
• Urban and Rural Studies
• Browse content in Warfare and Defence
• Defence Strategy, Planning, and Research
• Land Forces and Warfare
• Military Administration
• Military Life and Institutions
• Naval Forces and Warfare
• Other Warfare and Defence Issues
• Peace Studies and Conflict Resolution
• Weapons and Equipment

Sports Injury Research

• Cite Icon Cite
• Permissions Icon Permissions

With the increasing focus on tackling obesity and other lifestyle-related illnesses and conditions, participation in sports and physical activity is growing. The consequences are that injuries and unwanted side-effects of healthy activity are becoming major health problems. Prevention is crucial to health gain, both in the short-term (preventing immediate injury), and in the longer term (reducing the risk of recurrence and prolonged periods of impairment). Prevention follows four main steps: 1) the sports injury problem must be described in incidence and severity; 2) the etiological risk factors and mechanisms underlying the occurrence of injury are identified; 3) preventive methods that are likely to work can be developed and introduced; and 4) the effectiveness and cost-effectiveness of such measures are evaluated.

Signed in as

Institutional accounts.

• Google Scholar Indexing
• GoogleCrawler [DO NOT DELETE]

Personal account

• Sign in with email/username & password
• Get email alerts
• Save searches
• Purchase content
• Activate your purchase/trial code
• Add your ORCID iD

Institutional access

Sign in with a library card.

• Sign in with username/password
• Recommend to your librarian
• Institutional account management
• Get help with access

Access to content on Oxford Academic is often provided through institutional subscriptions and purchases. If you are a member of an institution with an active account, you may be able to access content in one of the following ways:

IP based access

Typically, access is provided across an institutional network to a range of IP addresses. This authentication occurs automatically, and it is not possible to sign out of an IP authenticated account.

Choose this option to get remote access when outside your institution. Shibboleth/Open Athens technology is used to provide single sign-on between your institution’s website and Oxford Academic.

• Click Sign in through your institution.
• Select your institution from the list provided, which will take you to your institution's website to sign in.
• When on the institution site, please use the credentials provided by your institution. Do not use an Oxford Academic personal account.
• Following successful sign in, you will be returned to Oxford Academic.

If your institution is not listed or you cannot sign in to your institution’s website, please contact your librarian or administrator.

Enter your library card number to sign in. If you cannot sign in, please contact your librarian.

Society Members

Society member access to a journal is achieved in one of the following ways:

Sign in through society site

Many societies offer single sign-on between the society website and Oxford Academic. If you see ‘Sign in through society site’ in the sign in pane within a journal:

• Click Sign in through society site.
• When on the society site, please use the credentials provided by that society. Do not use an Oxford Academic personal account.

If you do not have a society account or have forgotten your username or password, please contact your society.

Sign in using a personal account

Some societies use Oxford Academic personal accounts to provide access to their members. See below.

A personal account can be used to get email alerts, save searches, purchase content, and activate subscriptions.

Some societies use Oxford Academic personal accounts to provide access to their members.

Viewing your signed in accounts

Click the account icon in the top right to:

• View your signed in personal account and access account management features.
• View the institutional accounts that are providing access.

Signed in but can't access content

Oxford Academic is home to a wide variety of products. The institutional subscription may not cover the content that you are trying to access. If you believe you should have access to that content, please contact your librarian.

For librarians and administrators, your personal account also provides access to institutional account management. Here you will find options to view and activate subscriptions, manage institutional settings and access options, access usage statistics, and more.

Our books are available by subscription or purchase to libraries and institutions.

• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Rights and permissions
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Academia.edu no longer supports Internet Explorer.

To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to  upgrade your browser .

•  We're Hiring!
•  Help Center

Sports Injury

• Most Cited Papers
• Most Downloaded Papers
• Newest Papers
• Save to Library
• Last »
• Stretching and Flexibility training Follow Following
• Stretching Follow Following
• Acupuncture Follow Following
• Flexibility Follow Following
• Sports Law Follow Following
• Sports Nutrition Follow Following
• Sport Media Follow Following
• Sports Coaching and Training Follow Following
• Personal Injury Law Follow Following
• Injury Prevention Follow Following

Enter the email address you signed up with and we'll email you a reset link.

• Academia.edu Publishing
•   We're Hiring!
•   Help Center
• Find new research papers in:
• Health Sciences
• Earth Sciences
• Cognitive Science
• Mathematics
• Computer Science
• Academia ©2024

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Front Sports Act Living

Great Challenges Toward Sports Injury Prevention and Rehabilitation

Pascal edouard.

1 Inter-university Laboratory of Human Movement Science (LIBM EA 7424), University of Lyon, University Jean Monnet, Saint Etienne, France

2 Faculty of Medicine, Department of Clinical and Exercise Physiology, Sports Medicine Unit, University Hospital of Saint-Etienne, Saint-Etienne, France

Kevin R. Ford

3 Department of Physical Therapy, Congdon School of Health Sciences, High Point University, High Point, NC, United States

Introduction

While the positive benefits of sport participation are numerous, unfortunately this is balanced by the negative effects of injury (Engebretsen and Bahr, 2005 ; Merkel, 2013 ; Malm et al., 2019 ). Sports injury prevalence and incidence vary according to sports and population. And unfortunately, short- and long-term negative consequences can impact a variety of different domains (e.g., physical, psychological, sport, professional, financial, or social). Whatever these impacts, there is a need to reduce the occurrence and consequences of sports injury in order to allow a healthy and sustainable sports participation (Engebretsen and Bahr, 2005 ). This is the important challenge of injury prevention and rehabilitation! This challenge includes primary prevention (i.e., to reduce the occurrence of the first injury event), secondary prevention (i.e., to reduce the occurrence of reinjury after a first injury event) and tertiary prevention (i.e., to reduce the occurrence of sequalae) (Commission on Chronic Illness, 1957 ).

The popular saying, “prevention is better than cure,” has been known for quite some time. While large efforts have launched to continue to engage in this challenging goal of injury prevention, it still remains a real problem in sports. This could be due to the complex and multifactorial nature of sports injury (Meeuwisse, 1994 ; Bittencourt et al., 2016 ; Pol et al., 2019 ), which makes its “prevention” / “reduction” difficult. It seems that sports injury is not the result of one unique cause but likely the combination and interactions of several factors (including among others intrinsic and extrinsic risk factors and injury mechanism) (Meeuwisse, 1994 ; Bahr and Krosshaug, 2005 ; Bittencourt et al., 2016 ).

In order to face this challenging problem of sports injury, there is thus even more of a need to understand sports injury: How to monitor sports injury? What are the risk factors? How these factors interact? What is the healing process injured tissue? How can we optimize the process of healing, functional recovery, and safe return to sports? Then, there is a need to continue to reflect and develop strategies that can help to reduce the occurrence and recurrence of sports injuries: How can we play on/modify these factors to reduce the occurrence and/or recurrence of sports injuries? Which strategy or combination of strategies can reduce the occurrence and/or recurrence of sports injuries? Are these strategies efficient to reduce the occurrence and/or recurrence of sports injuries, in the context of scientific studies and in real life? How can we implement these strategies? How can athletes be compliant with these strategies? To answer these questions and reach this great challenge of injury prevention and rehabilitation, we believe that there is not one way, we believe that approaches should be comprehensive, multidisciplinary and holistic, including contributors from different fields, with communication between them and by embracing new fields.

Prevent or Reduce?

Before going to concrete aspects, there is maybe a need to improve knowledge and accuracy on some conceptual and terminological aspects.

“Prevention” is a widely used term, however, this is likely not the most appropriate or feasible term within sports. This term is well-known and recognized as a banner of work which aims to protect the health of athletes, especially injuries and illnesses, perhaps thanks to the important efforts of the Oslo Sports Trauma Research Center and the IOC toward injury and illness prevention (Engebretsen and Bahr, 2005 ; Ljungqvist, 2008 ; Engebretsen et al., 2014 ). Although this term could be useful to describe the field (as we use for the name of our section Injury Prevention and Rehabilitation ), it is maybe not the most appropriate when we want to accurately discuss about the concrete goals. Indeed, “prevention” means no occurrence of injuries, which is probably not possible. So, most appropriate terms would probably be “injury control” or “injury risk management” or “injury risk reduction” (Avery, 1995 ; Webster and Hewett, 2018 ). Other terms are also used on our field that deserve to have clear definitions for proper use, such as for instance “efficacy,” “effectiveness,” “compliance,” “prediction,” “prognostic.” Therefore, we believe that some discussions, researches and/or consensus should clarify these aspects.

In addition, sports injury prevention research is often modeled around the classic four-step sequence presented by van Mechelen et al. ( 1992 ) nearly 30 years ago. This model has provided a conceptual framework to monitor progress and effectiveness of decreasing the incidence of a variety of sport injuries (Edouard et al., 2015 ; Hewett et al., 2016 ). The “sequence of prevention” conceptual framework was extended in 2006 by Finch ( 2006 ) to phases related to the implementation of prevention measures and evaluation of real-world impact. Recently, Bolling et al. ( 2018 ) revised the four-steps sequence by improving the first step of the sequence extended to exploration of the context of the sports injury. Other frameworks have been developed to detail some steps of the sequence or some specific aspects, for instance, concept of sports injury (Timpka et al., 2014b ), etiology of sports injury (Meeuwisse, 1994 ), understanding injury mechanisms (Bahr and Krosshaug, 2005 ), a biomechanics-focused model (Hewett and Bates, 2017 ), complex systems approach (Bittencourt et al., 2016 ; Pol et al., 2019 ), risk factor-based categorization of the prevention (Jacobsson and Timpka, 2015 ), prevention measure implementation (Tee et al., 2020 ), and individualized approach (Roe et al., 2017 ). These conceptual frameworks of sports injury prevention research can continue to benefit from improvements or details to help researchers and/or practitioners.

Primary and Secondary Prevention = Same Fight!

Methodology used in primary prevention could seamlessly assist secondary prevention and vice versa (Hewett and Bates, 2017 ; Cools et al., 2020 ). In addition, given the high prevalence of sports injury, a large percentage of athletes will participate in sport with history of previous injury. Therefore, the need for secondary prevention is ongoing and increasingly more important as athletes age. However, primary and secondary approaches are sometimes compartmentalized; sports scientists and coaches may be more involved with primary prevention, while health professionals involved with secondary prevention. Consequently, scientific literature may also be compartmentalized. Therefore, we strongly support that all knowledge regarding both primary and secondary prevention should be directly translated to all stakeholders (applied, clinical). In addition, we suggest increased communication and collaboration between professionals and community to reach success in this challenge.

Specificity of Sports Rehabilitation

Secondary prevention can be addressed through rehabilitation (Hewett and Bates, 2017 ; Cools et al., 2020 ). This particular phase of the sports injury management has some specificities. It aims to orient/guide the injured tissue healing process, restore the function, and help the patient/athlete return to sporting activities while at the same time minimizing the risk of reinjury. This multi-goal management is currently approached mainly through biological/physical aspects (e.g., physiological, biomechanical…). However, psychological, social and contextual factors play a critical role in the recovery of patients/athletes after sports injury, and should be taken into account in this phase of the sports injury management.

Sports rehabilitation should thus be done in a multifactorial biopsychosocial approach, bringing the patient/athlete from injury to return to his desired activity, by taking into account the consequences of the sports injury at these different levels (Ardern et al., 2016 ; Van Melick et al., 2016 ; Cools et al., 2020 ).

Unity is Strength: Need of Multidisciplinary Teamwork!

To face the problem of sports injury, everyone is needed! Each person has a different experience, expertise, and view of the problem. So, it is important to encourage and act on the input from all parties involved. This implies a multidisciplinary approach, with inputs from several fields (e.g., sports medicine, sports and exercise science, physical conditioning and training, biomechanics, nutrition, physiology, psychology, sociology, data science…). This implies for instance at a field level that health professionals and coaching staffs, who are facing the same problem of sports injury, share their points of view, arguments, proposals of management in order to find the optimal solution for athletes. Likewise, this should be extended to other fields working with athletes in order to create a cohesive multidisciplinary team. This approach should be favored at the field/clinical and research levels.

Such an approach implies communication to go beyond discussions simply within a field and extend to discussions between diverse fields of interest. This also means for athletes' monitoring or research purpose collecting data from different fields, and probably makes choice or compromise given the amount of data this can represent. These discussions or choices are probably not easy because of some conceptual or language barriers, potential for competition, or perceived skepticism. There will maybe a need to structure discussion / choice, and there is a need to clarify the responsibility of each other, especially when coming the decision. But we believe that this is a relevant orientation to overcome the great challenge of sports injury prevention and rehabilitation. We suggest this would be a win-win approach for all stakeholders. The resulting benefits of discussion, exchange, and collaboration would be greater than the sum of each individual input.

Need for a Holistic and Individual Approach

Recent evidence supports that several factors of varying types can play a role in the occurrence of injury or reinjury (Kerkhoffs et al., 2012 ; Hewett et al., 2016 ; Green et al., 2020 ). In addition, each athlete will respond differently to these factors and combination of factors; each will not have an injury for the same reasons. Hence, patients will respond differently to the injury and its consequences. This is supported in contemporary sports injury management which should utilize a bio-psycho-social approach at an individual level and grounded in evidenced-based practice. Thus, efforts should be made in sports injury research to provide knowledge and evidence in each of these different fields, and if possible, combining all these fields.

Given the complexity of sports injury, the research approach currently simplifies the problem, but there will need to go deeper in complex multifactorial individualized approach to better meet the “reality” of sports injury (Meeuwisse, 1994 ; Bittencourt et al., 2016 ; Pol et al., 2019 ).

There is thus a need for a more complex approach, a comprehensive holistic and individual approach, as for understanding the determinants of the sports injury as for the development of strategies that aim to reduce the occurrence of injury or reinjury. Examples are proposed through conceptual or perspective articles (Mendiguchia and Brughelli, 2011 ; Mendiguchia et al., 2017 ; Buckthorpe et al., 2019 ), and there is now a need to provide supporting evidence of the theses approaches.

Improve Methodological and Analytical Approaches

One of the challenges in injury prevention research is to capture the outcome, i.e., sports injury. Efforts have been done to develop and improve methodology for recording and reporting injuries (Hagglund et al., 2005 ; Fuller et al., 2006 ; Junge et al., 2008 ; Timpka et al., 2014a ; Bahr et al., 2020 ) and should continue to most accurately capture injuries and their complexities.

Alternative analytical approaches of effectiveness of injury prevention measures can use as outcome the consequences of the sports injury at physical, psychological, social or financial levels. Injury prevention is of course useful to reduce the occurrence or reoccurrence of injuries, but also that of sequelae (Engebretsen and Bahr, 2005 ) or of the financial impact (Krist et al., 2013 ). Taking into account the economical burden of sports injuries (Hickey et al., 2014 ; Hespanhol Junior et al., 2017 ) could also be a way to improve stakeholders adherence to prevention and increase means for sports injury prevention and rehabilitation at the practical and research levels.

The multifactorial biopsychosocial approach leads to the need of adding in the measurements, data collection or monitoring, information related to the sports injury and the injured athletes, taking into account their multifactorial and complex nature, as well as about the context including individual, socio-cultural and environmental/policy levels (Bolling et al., 2018 ).

The multifactorial approach leads to multimodal methodological approaches. Traditionally quantitative analyses are used in sports injury prevention and rehabilitation research. There is thus a need to improve knowledge through qualitative approach (Bolling et al., 2018 , 2019a , b ). There is also a need for more behavioral approach when it comes to actual sports injury prevention (Verhagen et al., 2010 ) and when we aim increase compliance to prevention measures.

The multifactorial approach leads to analytical challenges. Indeed, this implies increasing the magnitude and type of data, which is of interest to fit the complex nature of sports injury, but can be difficult to managed by traditional analytical approaches, and for sure imply the collaboration with statistical and data science community (Casals and Finch, 2018 ; Nielsen et al., 2020b ). To analyse complex interactions between factors and/or between sports injury and factors, there is a need for new analytical advances (Bittencourt et al., 2016 ; Nielsen et al., 2020b ). As a consequence, other fields of data analyses, such as for instance machine learning, will continually be embraced in the future (Bittencourt et al., 2016 ; Ruddy et al., 2019 ). These analytical approaches may help analyse complex interactions as well as estimating the risk of sports injury occurrence, with application to understand the sports injury as well as to reduce their occurrence or recurrence (Bittencourt et al., 2016 ; Ruddy et al., 2019 ).

In addition, there is a need to use appropriate methodologies to analyse the efficacy of each of these strategies. Randomized Controlled Trial is currently the gold standard to analyse the efficacy of an intervention (Philipps et al., 2009 ), it is the design that should allow the highest level of evidence by minimizing the risk of bias. However, such design may not be the most relevant to reflect the reality of sports given, among others, the risk of low compliance (Nielsen et al., 2020a ). We could benefit from improvement in methodological design inspired from other research fields. In addition, usual analytical approaches, such as intention-to-treat, per protocol or as treated analyses, can lead to bias, especially in the context of low compliance (Edouard et al., in revision). Therefore, there is a need to explore other analytical approaches, as IV analysis, or other G-estimation, which can address some of the problems that arise from low compliance without losing the value of randomization and can also be helpful in observational studies (Edouard et al., in revision).

Conclusions

Although injury prevention and rehabilitation are not new disciplines, there is still an unmet need to improve knowledge toward theoretical understanding on epidemiology, risk factors, and injury mechanisms, as well as on practical strategies that can reduce the risk of sports injury or reinjury and of sequalae after injuries. Given the complex nature of injury, a holistic multifactorial biopsychosocial approach is needed through comprehensive, multidisciplinary and individualized approach to reach this great challenge. We therefore hope that this new section Injury Prevention and Rehabilitation of the Frontiers in Sports and Active Living can contribute to this improvement of knowledge, but also positively impact the sustainable and safe participation and short and long-term health of athletes.

Author Contributions

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding. This work was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under award number R21AR069873 (to KF).

• Ardern C. L., Glasgow P., Schneiders A., Witvrouw E., Clarsen B., Cools A., et al.. (2016). 2016 Consensus statement on return to sport from the First World Congress in Sports Physical Therapy, Bern . Br. J. Sport. Med . 50 , 853–864. 10.1136/bjsports-2016-096278 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Avery J. G. (1995). Accident prevention-injury control-injury prevention-or whatever? Inj. Prev. 1 , 10–11. 10.1136/ip.1.1.10 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Bahr R., Clarsen B., Derman W., Dvorak J., Emery C. A., Finch C. F., et al. (2020). International Olympic Committee consensus statement: methods for recording and reporting of epidemiological data on injury and illness in sport 2020 (including STROBE Extension for Sport Injury and Illness Surveillance (STROBE-SIIS).) Br. J. Sports Med. 54 , 372–389. 10.1136/bjsports-2019-101969 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Bahr R., Krosshaug T. (2005). Understanding injury mechanisms: a key component of preventing injuries in sport . Br. J. Sport. Med . 39 , 324–329. 10.1136/bjsm.2005.018341 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Bittencourt N. F. N., Meeuwisse W. H., Mendonça L. D., Nettel-Aguirre A., Ocarino J. M., Fonseca S. T. (2016). Complex systems approach for sports injuries: moving from risk factor identification to injury pattern recognition—narrative review and new concept . Br. J. Sports Med. 50 , 1309–1314. 10.1136/bjsports-2015-095850 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Bolling C., Barboza S. D., Mechelen W., Van Pasman H. R. (2019a). Letting the cat out of the bag : athletes, coaches and physiotherapists share their perspectives on injury prevention in elite sports . Br. J. Sports Med. 9 :bjsports-2019-100773. 10.1136/bjsports-2019-100773 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Bolling C., Delfino Barboza S., van Mechelen W., Pasman H. R. (2019b). How elite athletes, coaches, and physiotherapists perceive a sports injury . Transl. Sport. Med. 2 , 17–23. 10.1002/tsm2.53 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Bolling C., van Mechelen W., Pasman H. R., Verhagen E. (2018). Context matters: revisiting the first step of the ‘sequence of prevention' of sports injuries . Sport. Med. 48 , 2227–2234. 10.1007/s40279-018-0953-x [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Buckthorpe M., Wright S., Bruce-Low S., Nanni G., Sturdy T., Gross A. S., et al.. (2019). Recommendations for hamstring injury prevention in elite football: translating research into practice . Br. J. Sports Med. 53 , 449–456. 10.1136/bjsports-2018-099616 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Casals M., Finch C. F. (2018). Sports biostatistician: a critical member of all sports science and medicine teams for injury prevention . Br. J. Sports Med. 52 , 1457–1461. 10.1136/bjsports-2016-042211rep [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Commission on Chronic Illness (1957). Chronic Illness in the United States, Vol. 1: Prevention of Chronic Illness . Cambridge: Harvard University Press; 10.4159/harvard.9780674497474 [ CrossRef ] [ Google Scholar ]
• Cools A. M., Maenhout A. G., Vanderstukken F., Declève P., Johansson F. R., Borms D. (2020). The challenge of the sporting shoulder: from injury prevention through sport-specific rehabilitation toward return to play . Ann. Phys. Rehabil. Med. 10.1016/j.rehab.2020.03.009. [Epub ahead of print]. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Edouard P., Alonso J. M., Jacobsson J., Depiesse F., Branco P., Timpka T. (2015). Injury prevention in athletics: the race has started and we are on track! New Stud. Athl. 30 , 69–78. [ Google Scholar ]
• Engebretsen L., Bahr R. (2005). An ounce of prevention? Br. J. Sports Med. 39 , 312–313. 10.1136/bjsm.2005.018333 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Engebretsen L., Bahr R., Cook J. L., Derman W., Emery C. A., Finch C. F., et al.. (2014). The IOC Centres of Excellence bring prevention to Sports Medicine . Br. J. Sports Med. 48 , 1270–1275. 10.1136/bjsports-2014-093992 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Finch C. (2006). A new framework for research leading to sports injury prevention . J. Sci. Med. Sport 9 , 3–9. 10.1016/j.jsams.2006.02.009 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Fuller C., Ekstrand J., Junge A., Andersen T. E., Bahr R., Dvorak J., et al. (2006). Consensus statement on injury definitions and data collection procedures in studies of football (soccer) injuries . Scand. J. Med. Sci. Sport. 16 , 83–92. 10.1111/j.1600-0838.2006.00528.x [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Green B., Bourne M. N., van Dyk N, Pizzari, T. (2020). Recalibrating the risk of hamstring strain injury (HSI) - A 2020 systematic review and meta- - analysis of risk factors for index and recurrent HSI in sport . Br. J. Sports Med. 10.1136/bjsports-2019-100983. [Epub ahead of print]. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hagglund M., Walden M., Bahr R., Ekstrand J. (2005). Methods for epidemiological study of injuries to professional football players: developing the UEFA model . Br. J. Sports Med. 39 , 340–346. 10.1136/bjsm.2005.018267 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hespanhol Junior L. C., van Mechelen W., Verhagen E. (2017). Health and economic burden of running-related injuries in dutch trailrunners: a prospective cohort study . Sport. Med. 47 , 367–377. 10.1007/s40279-016-0551-8 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hewett T. E., Bates N. A. (2017). Preventive biomechanics: a paradigm shift with a translational approach to injury prevention . Am. J. Sports Med. 45 , 2654–2664. 10.1177/0363546516686080 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hewett T. E., Myer G. D., Ford K. R., Paterno M. V., Quatman C. E. (2016). Mechanisms, prediction, and prevention of ACL injuries: cut risk with three sharpened and validated tools . J. Orthop. Res . 34 , 1843–1855. 10.1002/jor.23414 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hickey J., Shield A. J., Williams M. D., Opar D. A. (2014). The financial cost of hamstring strain injuries in the Australian Football League . Br. J. Sports Med. 48 , 729–730. 10.1136/bjsports-2013-092884 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Jacobsson J., Timpka T. (2015). Classification of prevention in sports medicine and epidemiology . Sport. Med. 45 , 1483–1487. 10.1007/s40279-015-0368-x [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Junge A., Engebretsen L., Alonso J. M., Renström P., Mountjoy M., Aubry M., et al.. (2008). Injury surveillance in multi-sport events: the International Olympic Committee approach . Br. J. Sports Med. 42 , 413–421. 10.1136/bjsm.2008.046631 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kerkhoffs G. M., van den Bekerom M., Elders L. A., van Beek P. A., Hullegie W. A., Bloemers G. M., et al.. (2012). Diagnosis, treatment and prevention of ankle sprains: an evidence-based clinical guideline . Br. J. Sport. Med . 46 , 854–860. 10.1136/bjsports-2011-090490 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Krist M. R., Van Beijsterveldt A. M. C., Backx F. J. G., Ardine de Wit G. (2013). Preventive exercises reduced injury-related costs among adult male amateur soccer players: a cluster-randomised trial . J. Physiother. 59 , 15–23. 10.1016/S1836-9553(13)70142-5 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Ljungqvist A. (2008). Sports injury prevention : a key mandate for the IOC . Br. J. Sports Med. 42 :391. [ PubMed ] [ Google Scholar ]
• Malm C., Jakobsson J., Isaksson A. (2019). Physical activity and sports—real health benefits: a review with insight into the Public Health of Sweden . Sports 7 :127. 10.3390/sports7050127 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Meeuwisse W. (1994). Assessing causation in sport injury: a multifactorial model . Clin. J. Sport Med . 4 , 166–170. 10.1097/00042752-199407000-00004 [ CrossRef ] [ Google Scholar ]
• Mendiguchia J., Brughelli M. (2011). A return-to-sport algorithm for acute hamstring injuries . Phys. Ther. Sport 12 , 2–14. 10.1016/j.ptsp.2010.07.003 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Mendiguchia J., Martinez-Ruiz E., Edouard P., Morin J.-B., Martinez-Martinez F., Idoate F., et al.. (2017). A multifactorial, criteria-based progressive algorithm for hamstring injury treatment . Med. Sci. Sports Exerc. 49 , 1482–1492. 10.1249/MSS.0000000000001241 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Merkel D. L. (2013). OAJSM-33556-youth-sports—risky-business-or-a-necessary-evil- . Open Access J. Sport. Med. 4 , 151–160. 10.2147/OAJSM.S33556 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Nielsen R. O., Bertelsen M. L., Ramskov D., Damsted C., Verhagen E., Bredeweg S. W., et al.. (2020a). Randomised controlled trials (RCTs) in sports injury research: Authors - Please report the compliance with the intervention . Br. J. Sports Med. 54 , 51–57. 10.1136/bjsports-2019-100858 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Nielsen R. O., Shrier I., Casals M., Aguirre A. N., Møller M., Bolling C., et al.. (2020b). Statement on methods in sport injury research from the 1st METHODS MATTER Meeting, Copenhagen, 2019 . Br. J. Sports Med. 10.1136/bjsports-2019-101323 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Philipps B., Ball C., Sackett D., Badenoch D., Strauss S., Haynes B., et al. (2009). Oxford Centre for Evidence-based Medicine Levels of Evidence . Available online at: https://www.cebm.net/2009/06/oxford-centre-evidence-based-medicine-levels-evidence-march-2009/
• Pol R., Hristovski R., Medina D., Balague N. (2019). From microscopic to macroscopic sports injuries. Applying the complex dynamic systems approach to sports medicine: a narrative review . Br. J. Sports Med. 53 , 1214–1220. 10.1136/bjsports-2016-097395 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Roe M., Malone S., Blake C., Collins K., Gissane C., Büttner F., et al.. (2017). A six stage operational framework for individualising injury risk management in sport . Inj. Epidemiol. 4 :26. 10.1186/s40621-017-0123-x [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Ruddy J. D., Cormack S. J., Whiteley R., Williams M. D., Timmins R. G., Opar D. A. (2019). Modeling the risk of team sport injuries: a narrative review of different statistical approaches . Front. Physiol. 10 :829. 10.3389/fphys.2019.00829 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Tee J. C., Mclaren S. J., Jones B. (2020). Sports injury prevention is complex : we need to invest in better processes, not singular solutions . Sport. Med. 50 , 689–702. 10.1007/s40279-019-01232-4 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Timpka T., Alonso J.-M., Jacobsson J., Junge A., Branco P., Clarsen B., et al.. (2014a). Injury and illness definitions and data collection procedures for use in epidemiological studies in Athletics (track and field): consensus statement . Br. J. Sports Med. 48 , 483–490. 10.1136/bjsports-2013-093241 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Timpka T., Jacobsson J., Bickenbach J., Finch C. F., Ekberg J., Nordenfelt L. (2014b). What is a sports injury? Sports Med. 44 , 423–428. 10.1007/s40279-014-0143-4 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• van Mechelen W., Hlobil H., Kemper H. C. G. (1992). Incidence, severity, aetiology and prevention of sports injuries. A review of concepts . Sports Med. 14 , 82–99. 10.2165/00007256-199214020-00002 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Van Melick N., Van Cingel R. E. H., Brooijmans F., Neeter C., Van Tienen T., Hullegie W., et al.. (2016). Evidence-based clinical practice update: practice guidelines for anterior cruciate ligament rehabilitation based on a systematic review and multidisciplinary consensus . Br. J. Sports Med. 50 , 1506–1515. 10.1136/bjsports-2015-095898 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Verhagen E. A. L. M., Stralen M. M., Van Mechelen W., Van (2010). Behaviour, the key factor for sports injury prevention . Sports Med. 40 , 899–906. 10.2165/11536890-000000000-00000 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Webster K. E., Hewett T. E. (2018). Meta-analysis of meta-analyses of anterior cruciate ligament injury reduction training programs . J. Orthop. Res. 36 , 2696–2708. 10.1002/jor.24043 [ PubMed ] [ CrossRef ] [ Google Scholar ]

• Entertainment

Leinster receive major injury boost ahead of URC playoffs

Leinster have been handed a major injury boost ahead of Saturday’s URC quarter-final against Ulster at Aviva Stadium, with Garry Ringrose finally back in selection contention.

The Leinster and Ireland midfielder has not featured for his province since he suffered a shoulder injury in a Champions Cup pool game against Leicester Tigers in January.

Today's top videos

Story continues below.

Ringrose struggled with the injury throughout the Six Nations and missed the entirety of Leinster’s European knockout games, including the final defeat by Toulouse at Tottenham Hotspur Stadium a fortnight ago.

But the experienced midfielder is set to to return to action this weekend along with the rest of the frontliners who sat out last Friday’s bonus-point victory (33-7) against Connacht at the RDS.

Jack Conan is an injury doubt, however, while Jason Jenkins will miss the inter-pro clash in the capital with a knee injury.

Meanwhile, there was mixed news on the injury front for Munster yesterday.

Head coach Graham Rowntree will have Diarmuid Barron, Antoine Frisch and Josh Wycherley back on duty for the URC quarter-final clash with Ospreys at Thomond Park on Friday night.

Munster will be without Rory Scannell, Joey Carbery and Tom Ahern after the key trio picked up injuries in last weekend’s win against Ulster.

Kiwi centre Alex Nankivell also remains on the sidelines with an ankle injury as the southern province attempt to plot back-to-back league titles in the weeks ahead.

Why Leinster’s quest to land URC title won’t be easy

Rog warns leinster champions cup loss will linger ‘like a bad smell’ until duck is broken, leo cullen backs his methods and hits back at criticism over champions cup final loss, must read rugby.

More: Trending Rugby