Table 2 Data extraction for included studies (n = 25)

Foster et al. [42]

Prospective cohort. Causality implied

n = 82. Potentially high performance from club and school teams. Mean age 16.8 years, age range 15 to 22 years. Male. Australia. No control group included. 1986–87 cricket season. One cricket season duration. No reporting of injury history. No calculation of sample size

Body mass, skinfold levels, anthropometric measurements, hamstring and quadriceps torque, shoulder strength, trunk strength, abdominal muscle endurance, posture, shoulder and hamstring muscle flexibility, aerobic capacity. Bowling kinematics (shoulder and hip alignment at BFC, SCR, BR height). Bowling kinetics (GRF)

CT diagnosed LSF and soft tissue injuries that resulted in a participant missing at least one cricket match.

Stress fracture group bowlers had lower longitudinal foot arch height, greater non-dominant quadriceps torque, higher BR height than non-injured bowlers (p < 0.05) and displayed increased SCR (> 40°) between BFC and FFC (non-injured bowlers rotated 15.7°). No association between peak vertical or horizontal GRF during FFC and back injury. Bowlers who sustained a back injury had greater bowling arm shoulder depression and horizontal flexion strength and displayed a shoulder alignment > 240° at BFC (p < 0.05). Higher BR position in stress fracture bowlers as a result of more extended front hip and knee joint angles and a more upright trunk during FFC phase

Elliott et al. [54]

Cross sectional. Causality implied

n = 20. Western Australian fast bowling development squad. Mean age 17.9 ± 1.6 years. Male. Australia. No control group included. 1989–90 cricket season. No reporting of injury history. No calculation of sample size

Skinfold levels, hamstring/lower back flexibility, abdominal muscular endurance, posture, trunk extension/flexion strength. Bowling kinematics (shoulder and hip alignment at BFC, SCR, BR height). Bowling kinetics (GRF)

Disc degeneration determined by MRI. CT scan diagnosed bone abnormalities categorised as spondylolysis, spondylolisthesis, or pedicle sclerosis

Bowlers with radiological abnormalities had a more front on alignment (no injury 179°, disc abnormality 206°, LSF 197°) at BFC. (p < 0.05) and higher degrees of SCR during the delivery stride (no injury 0°, disc abnormality 25°, LSF 12°). Bowlers who adopted a side on technique had the least likelihood of abnormal radiological features, in contrast to those who employed a mixed technique who had the highest rate of radiological abnormalities. LSF bowlers had significantly higher BR height relative to standing height (114% vs 110% in non-injured bowlers) and increased minimum shoulder alignment (LSF 193°, disc 181°, no injury 179°). Age of fast bowlers in injury free group (median 16.4 years) significantly younger than LSF group (median 18.4 years). Sit and reach scores significantly lower in intervertebral disk abnormality group (4.5 cm) than the injury free group (8.0 cm)

Hardcastle et al. [55]

Cross sectional. Causality implied

n = 24. Special training in a Western Australian squad. Mean age 17.9 years, age range 16 to 18 years. Male. Australia. Control group 13 batsmen from the same age group. Study dates not described. No reporting of injury history. No calculation of sample size

Kinematics of fast bowling action relating to difference in rotation of the shoulders and the line of the back foot at the time of BFC. Classification of the fast-bowling action whether it was side on, mixed or front on based on shoulder and hip alignment

Disc degeneration determined by MRI. CT scan and radiograph diagnosed bone abnormalities categorised as spondylolysis, spondylolisthesis, or pedicle sclerosis

75% of fast bowlers with a mixed technique experienced LBP. 100% of fast bowlers who rotated > 10° demonstrated a radiological abnormality and 80% of these experienced LBP

Elliott et al. [56]

Cross sectional. Causality implied

n = 24. School/club level and showed most potential as determined by match statistics. Mean age 13.7 years. Male. Australia. No reporting of injury history. No control group included. 1991–92 cricket season. No calculation of sample size

Body height, body mass, skinfold levels, hamstring/lower back flexibility, abdominal muscular endurance. Bowling kinematics (shoulder and hip alignment at BFC, SCR and BR height). Bowling kinetics (GRF)

Lumbar intervertebral disc abnormalities (degeneration and bulging) identified by MRI scan

Fast bowlers who counter rotated their shoulders by 30° or more were more likely to sustain a disc abnormality then bowlers who counter rotated their shoulders by 18.9° with p = 0.088. Significance level set at 0.1. All abnormal scans recorded from bowlers who used a mixed bowling technique and no bowlers who used side-on, or front-on technique recorded scans with abnormal discs in the lumbosacral region

Burnett et al. [43]

Prospective cohort. Causality implied

n = 19. School/club level and showed and showed potential as genuine fast bowlers. Mean age at session 1; 13.6 ± 0.6 years, at session 2; 16.3 ± 0.6 years. Male. Australia. No control group included. Session 1; commencement of the 1991–92 cricket season. Session 2; conclusion of the 1993–94 cricket season. Study duration 3 years. No reporting of injury history. No calculation of sample size

Kinematics of fast bowling action relating to classification as either side on, mixed or front on based on shoulder and hip alignment at BFC. Degree of SCR

Presence of disc degeneration as measured by MRI

Significant increase in occurrence of disc degeneration (21% and 58%, p = 0.008) and LBP (5% and 53%, p = 0.002) between sessions 1 and 2. The progress of disc degeneration was found to be related to those bowlers who used a mixed bowling action and was seen in 80% of bowlers who used a mixed action during both sessions as opposed to 14% using this technique during one of these sessions only (p = 0.015).

Elliott & Khangure [65]

Intervention. Causality implied

n = 41. Western Australian Cricket Association fast bowling development squad. Group 1 mean age 13.4. Group 2 mean age 13.2. Male. Australia. No control group included. Group 1; 1997 to 2000. Group 2; 1998 to 2000. Study duration 3 years. No reporting of injury history. No calculation of sample size

Maximum knee flexion angle. Shoulder alignment at BFC, FFC, and at the position of maximum SCR. Classification of fast bowling action whether it was side on, mixed or front on

Abnormal radiological appearance of the intervertebral disk, such as disk degeneration or bulging assessed by MRI. Intervertebral disk degeneration

Bowlers who used a front-on or side-on action recorded significantly lower levels of lumbar disk degeneration when compared with mixed action bowlers (chi-square value 9.5126; p = 0.002). In year 1, bowlers who used a front-on or side-on action recorded significantly lower levels of lumbar disk degeneration (12.5%) when compared with mixed action (27%) bowlers. The incidence and progression of lumbar disk degeneration were significantly reduced in parallel with decreased SCR from 35.4° at the commencement of the study to 21.3° in year 4. Progression of the number and/or severity of degeneration only occurred with bowlers who used the mixed action

Portus et al. [60]

Combined retrospective cohort and prospective cohort. Causality implied

n = 42. National/international level. Mean age 22.4 ± 3.5 years. Male. Australia. No control group included. Outcomes gathered from 1992 to 2000. No calculation of sample size

Fast bowling technique-related factors; hip-shoulder separation angle at BFC, SCR, hip counter rotation. Classification into side-on, semi-open, front-on and mixed techniques, as well as styles of front lower limb actions during FFC phase

Radiologically diagnosed injuries in the presence of pain classified as either stress fractures of the lumbar spine pars interarticularis or back sprain injuries, including injury to disc, facet joint or ligament

SCR significantly higher in LSF group (41°) than the no trunk injury group (19°) (F = 4.5, p = 0.01). Hip counter rotation was highest in the LSF group and lowest in the no injury group. The LSF group had the most front-on shoulder orientation at BFC. At BFC, soft tissue injury group exhibited a larger hip-shoulder separation angle than the no injury group (p = 0.03). The LSF group was characterised by a more upright hip joint angle at FFC and at BR, whereas a more flexed knee characterised the non-trunk injured group, but this was statistically non-significant

Engstrom et al. [44]

Prospective cohort. Causality not implied

n = 56 at commencement and 51 at conclusion. Club to national level fast bowlers within respective age groups. Age range 13 to 17 years. Male. Australia. Control group of 20 swimmers. Study dates not described. Study duration for fast bowlers 4 years. No reporting of injury history. No calculation of sample size

QL CSA and muscle volume through the lumbar spine and subsequent percent QL asymmetry score for individual bowlers relative to the bowling arm side and swimmers relative to the dominant arm side

Symptomatic pars interarticularis lesions of the lumbar spine diagnosed by MRI.

Significant association between increasing QL asymmetry and the development of symptomatic L4 pars lesions in adolescent fast bowlers with greater QL volumes on the bowling arm side. Increasing QL asymmetry was associated with increased risk. 105% QL asymmetry = 4% risk, 125% QL asymmetry = 58% risk, and 130% QL asymmetry = 78% risk. The area under the nonparametric ROC curve was 0.89 (p ≤ 0.001), indicating excellent discrimination between fast bowlers with and without symptomatic L4 pars lesions, according to percent QL asymmetry

Stuelcken et al. [57]

Retrospective cohort. Causality not implied

n = 34; 26 females and 8 males. Females—elite level, Males—Australian Capital Territory representative squad Mean age of females 22.5 ± 4.5 years. Mean age of males 21.5 ± 3 years. Australia. Control group not specifically mentioned but results in females compared to males. Study dates not described. Reporting of injury history in addition to assessment of LBP and study commencement. No calculation of sample size

Range of motion of lumbar flexion, extension, and lateral flexion in standing. Range of motion of hip flexion and hip extension. Trunk extensor endurance

History of LBP if self-reported pain had occurred (1) in the previous 12 months, or (2) at any stage in the bowler’s career and could be verified by records kept by team medical support staff. Episodes of self-reported LBP that did not meet these criteria were disregarded to minimise recall bias

Female fast bowlers with a history of LBP (16.9 ± 3.1°) exhibited a restriction in lumbar lateral flexion range of motion to the bowling arm side compared to those female bowlers with no history of LBP (20.6 ± 7.5°) (p = 0.05)

Ranson et al. [45]

Prospective cohort. Causality implied

n = 28. ECB Elite Fast Bowling Group, England U19 or senior men’s team. Mean age 19 years, age range 16 to 24 years. Male. United Kingdom. No control group included. 2005—2008. Study duration 2 years. No reporting of injury history. No calculation of sample size

Appearance of lumbar intervertebral discs and BMO in the posterior bony elements on lumbar MRI scans

If bowler missed 7 or more consecutive days of cricket because of LBP. Lumbar stress fracture was recorded if history was corroborated by MRI and/or subsequent CT which confirmed acute bone stress changes associated with partial or complete fracture of the posterior elements of the lumbar spine

Principal finding was that acute stress changes such as oedema and periostitis in the posterior bony elements are related to the subsequent development of a stress fracture. Significant association between appearance of acute bone stress on MRI and subsequent LSF. Of the 15 players who had MRI scans which included presence of acute bone stress on an MRI scan, 11 developed LSF within a mean of 10 weeks from scan time (p < 0.001). Of these 15, 9 had no fracture line at scan and they developed LSF at mean of 11 weeks, but 6 did have a fracture line on initial scan and they all went on to develop a LSF within the next 4 weeks. No association found between intervertebral disc degeneration and MRI changes of bone stress (p = 0.544) or subsequent LSF (p = 0.124)

Stuelcken et al. [58]

Retrospective cohort. Causality implied

n = 26. Elite; national/international level. Mean age 22.5 ± 4.5 years. Female. Australia. No control group included. Study dates not described. Reporting of injury history. No calculation of sample size

Pelvis-shoulder separation at BFC. SCR during the delivery stride. Angle of thorax relative to pelvis during the delivery stride. Classification of bowling technique as side-on, front-on, or mixed

LBP history if self-reported pain had occurred in the previous 12 months, or at any stage in the bowler’s career and could be verified by records kept by team medical support staff

The mean maximum lateral flexion range of the thorax relative to the pelvis during the delivery stride was significantly greater for the LBP subgroup compared with the no LBP subgroup (p = 0.004), with a large effect size (g = 1.25). The thorax of the LBP subgroup was more laterally flexed away from the bowling arm relative to the pelvis between 73–79% of the delivery stride with upper and lower confidence limits surrounding the difference between the group means being < 0

Kountouris et al. [46]

Prospective cohort. Causality not implied

n = 38. Club level adolescents. Mean age 14.9 ± 1.34 years, age range 12 to 17 years. Male. Australia. No control group included. 2002–03 Australian cricket season. Study duration of one cricket season. No reporting of injury history. No calculation of sample size

QL CSA bilaterally at the L3–L4 vertebral levels. BMI

Bowlers who reported LBP and subsequently assessed and diagnosed by a sports physician with either LBSI (lumbar stress fracture or stress reaction of the posterior vertebral element) or soft tissue injury (disc, muscle, or ligament injury)

Significantly higher BMI in bowlers that sustained a LBSI. All 4 asymptomatic participants with radiological evidence of lumbar bone stress at baseline developed symptomatic LBSI. No association between average QL asymmetry for players who sustained lumbar spine injury (soft tissue (12.5%) and bone stress (15.7%) and players who were uninjured (12.4%) (p = .537). When participants were grouped as either LBSI (mean QL asymmetry 15.7%) or no LBSI (mean QL asymmetry 12.4%), there was no difference in QL CSA (p = .267).

Kountouris et al. [47]

Prospective cohort. Causality not implied

n = 23. Identified by respective selection panels as potential national and international players. Mean age 24 ± 3.6; years, age range 19 to 32 years. Male. Australia. No control group included. Study dates not described. Study duration of one cricket season. No reporting of injury history. No calculation of sample size

QL CSA bilaterally between the L2 and L4 vertebral levels. BMI

LBSI; lumbar stress fracture or stress reaction of the posterior vertebral element), which was confirmed radiologically (scintigraphy, MR, or CT), Soft tissue lumbar spine injury (any other injury other than bone stress)

Fast bowlers with no injury had significantly larger asymmetries when compared with those in the soft tissue and bone stress groups (p = 0.050). When participants were grouped as either having no injury or any (soft tissue and bone stress) lumbar injury, there was a significant difference between groups (p = 0.025). When LBSI group was compared to the two other groups combined, there was no significant difference in mean asymmetry (p = 0.949). Four participants (17%) had radiological evidence of lumbar bone stress at baseline, and all went on to develop symptomatic LSF

Olivier et al. [48]

Prospective cohort. Causality implied

n = 17. Recruitment derived from questionnaires and telephone calls, but response rate not reported. Amateur premier league club level. Age range 18 to 26 years, but mean age not reported. Male. South Africa. No control group included. Study dates not described. Study duration of one cricket season. Injury history reported. No calculation of sample size

Lumbar proprioception (lumbar reposition error) in the neutral lumbar spine position as well as lumbar positions corresponding to those at FFC and BR of the fast-bowling action

A musculoskeletal condition to the lower back that resulted in loss of at least 1 day of sporting activity or that occurred during a sporting activity that required medical attention, and which forced the bowler to quit the activity.

Of 9 position sense variables tested, lumbar reposition error on 8 associated with previous low back injury; F/E neutral (p < 0.01), average neutral (p < 0.01), F/E FFC (p < 0.02), left/right FFC (p < 0.03), average FFC (P < 0.01), F/E BR (p < 0.01), left/right BR (p < 0.03), average BR (P < 0.01). Lumbar reposition error on 2 variables associated with low back injury sustained during the period of the study: F/E neutral (p < 0.04), average neutral (p < 0.03)

Gray et al. [62]

Cross sectional. Causality implied

n = 25. Representative cricket at a provincial level. Age range 14 to 18 years. Male. South Africa. No control group included. Study dates not described. No reporting of injury history prior to study commencement. No calculation of sample size

Thickness of the TrA, IO and EO muscles assessed on the dominant and non-dominant side and resultant asymmetry between these measurements

Experiencing LBP for minimum of 6 weeks associated with fast bowling, or bowlers who were currently playing cricket but experienced LBP that caused bowler to miss a match or practice session in the previous 6 weeks

Total combined thickness of TrA, IO and EO on non-dominant side > dominant side in fast bowlers without LBP (post hoc p = 0.01) but did not differ between sides for bowlers with LBP (post hoc p = 1.0). Greater total thickness in bowlers without pain (3.0 ± 0.4 cm) than with pain (2.4 ± 0.4 cm) on non-dominant side (post hoc—p = 0.03), but similar for both groups on dominant side (no pain: 2.5 ± 0.4 cm; pain: 2.5 ± 0.4 cm) (post hoc—p = 1.0). Thickness of each abdominal muscle > on non-dominant side in no pain group (post hoc—p < 0.001) but did not differ between sides in bowlers with pain (post hoc—p = 0.01). Thickness of OI < in pain group than no pain group (post hoc—all p = 0.02)

Bayne et al. [49]

Prospective cohort. Causality implied

n = 46 who volunteered and underwent initial MRI assessment. 25 included in final analysis. District and/or state junior level. Mean age 15.8 years, age range 14 to 19 years. Male. Australia. No control group included. Study dates not described. Study duration of a 6-month cricket season. No reporting of injury history. No calculation of sample size

Ranges of motion of ankle dorsiflexion, hip internal and external rotation. Foot arch ratio. Trunk extensor endurance, hold times for prone plank and side plank. Tests for lumbopelvic stability, calf endurance, bridge capacity and lower limb movement control. Kinematics of the trunk, pelvis, and lower limbs during the fast-bowling action. Kinetics of the lower limbs and lumbar spine during the fast-bowling action

LBP that affected bowler’s ability to perform in a match, consistent with the consensus of cricket injury. Definition of injury was later expanded to include asymptomatic participants with radiological evidence of lumbar bone stress

Injured bowlers had: lower front hip angle at FFC (46 ± 6° vs 51 ± 6°, t = 2.076, p = 0.049), greater thorax lateral flexion at FFC (20 ± 6° vs 15 ± 5°, t = 2.187, p = 0.039) and at BR (50 ± 6° vs 40 ± 8°, t = 3.396, p = 0.002), increased pelvis rotation beyond front-on at BR (287 ± 11° vs 277 ± 11°, t = 2.408, p = 0.024), increased normalised peak flexion (10.5 ± 4.9 Nm/kg/m vs 6.9 ± 2.5 Nm/kg/m,t = 2.292, p = 0.036), and lateral flexion lumbar moments (12.5 ± 2.6 Nm/kg/m vs 10.6 ± 1.9 Nm/kg/m,t = 2.079, p = 0.049) and peak lateral flexion power (25.8 ± 16.2 W/kg/m vs 14.4 ± 7.7 W/kg/m, t = 2.203,p = 0.043), reduced Biering-Sorensen test hold time (103 ± 33 s vs 132 ± 33 s, t = 2.220, p = 0.037), increased knee valgus angle during single leg decline squat on dominant (9 ± 3° vs 5 ± 4°, t = 2.299, p = 0.031) and non-dominant leg (9 ± 4° vs 6 ± 3°, t = 2.362, p = 0.027). A score of 0 on lumbo–pelvic stability test associated with increased risk of low back injury (RR = 1.7, CI 0.78–4.10)

Olivier et al. [50]

Prospective cohort. Causality implied

n = 26. From 14 different cricket clubs in Premier league. Mean age 21.8 ± 1.8 years, age range 18 to 26 years. Male. South Africa. No control group included. Study dates not described. Study duration of an 8-month cricket season. Injury history reported. No calculation of sample size

CSA of lumbar MF at the L3, L4, and L5 vertebral levels

A musculoskeletal condition of the lower back that resulted in loss of at least one day of sporting activity or that occurred during a sporting activity that required medical attention, and which forced the bowler to quit the activity

In bowlers who sustained a lower back injury, the nondominant CSA of MF at L3 (p = 0.04) and L5 (p = .0.04) were smaller than the dominant side MF, however, the percentage difference of the low back injured groups was similar to the non-injured group. No statistically significant differences were found in bowlers with/without asymmetry and bowlers who did/did not sustain lower back injuries (L3 p = 0.28; L4 p = 0.60; L5 p = 1.00)

Alway et al. [51]

Prospective cohort. Causality implied

n = 368. English County first or second X1 fast bowlers. Mean age 24.87 ± 6.01 years. Male. United Kingdom. No control group included. 2010 to 2016 England cricket seasons. Study duration of 6 years. No reporting of injury history prior to study commencement. No calculation of sample size

Match bowling workload across all formats of cricket. Age at time of LSF of fast bowlers within this cohort

LSF with diagnosis based on symptomatic presentation and radiological evidence (MRI, CT, or SPECT), which resulted in a player being unavailable for match selection,

74% of LSF occurred in bowlers aged under 25 years, and 56% occurred between the ages 18 and 22. Risk of LSF greatest in bowlers aged 18 to 22, with match incidence at 0.32 per 10 000 deliveries, annual incidence of 4.90 per 100 fast bowlers, and prevalence of 3.21% of squad days. This compared to match incidence of 0.13 per 10 000 deliveries, annual incidence of 2.46 per 100, and prevalence of 1.37% of squad days in the entire cohort of fast bowlers

Alway et al. [63]

Cross sectional. Causality implied

n = 23. Selected from existing national senior or developmental squads. Mean age 24.58 ± 3.93 years. Male. United Kingdom. Controls consisted of 14 other cricketers (11 batters, 2 wicketkeepers, 1 spin bowler), 22 rugby players and 20 inactive control participants. Study dates not described. Injury history reported. No calculation of sample size

DEXA measured BMC and BMD of the lumbar spine as a whole (assessed from L1 to L4) and the posterior elements of L3

LSF history determined from England and Wales Cricket Board medical records with any diagnosis confirmed by MRI, CT, or SPECT CT scans

Fast bowlers who never suffered LSF had 3.6% greater BMD in the dominant side of lumbar vertebrae and 1.7% greater BMD in the non-dominant side of lumbar vertebrae compared with those who did suffer LSF, but this was not statistically significant (p = 0.08). No significant interaction found between side, vertebra and LSF history on BMC (p = 0.34). No significant interaction between side, vertebra, and disc injury history on BMD (p = 0.61) or BMC (p = 0.77)

Kountouris et al. [52]

Prospective cohort. Causality implied

n = 65. Australian junior elite fast bowlers who were selected in an under 17 or under 19 state or territory squad. Mean age 17.3 years, age range 14.7 to 18.8 years. Male. Australia. No control group included. July 2014 to March 2015. Study duration of 8 months. No reporting of injury history. No calculation of sample size

BMO being present or absent through the left and right posterior vertebral arch at each vertebral level from L1 to L5 on sagittal and coronal T2 fat suppressed or short-tau inversion recovery MRI sequences

Development of LBP that caused a participant to be unable to bowl for a period in the study period. Diagnosis of LBSI made with use of clinical judgement and imaging modalities

BMO detected on one or more scans during season associated with 39% of these bowlers suffering a symptomatic LBSI, 37% having persistent BMO but no symptoms, and 24% experiencing a reversal of detected BMO. All bowlers with a symptomatic LBSI had BMO detected at corresponding site of the vertebra in the scan immediately prior to diagnosis. The number of days between the first appearance of BMO on MRI and the player reporting LBSI related symptoms was a mean of 96 days and a median of 112 days. Participants who had BMO detected at any scan at increased risk of BSI (RR = 22.3 (95% CI 1.4—356.6), OR = 36.3 (95% CI 2.1—639.5), with positive predictive value 39%, and negative predictive value 100%

Senington et al. [61]

Combined retrospective cohort and prospective cohort. Causality implied

n = 35; 14 seniors and 21 juniors. Participants recruited through coaches from professional county cricket clubs. Mean age senior group 24.1 ± 4.3 years, junior group 16.9 ± 0.7 years. Male. United Kingdom. No control group included. 2015 England cricket season. Duration of prospective portion of study one cricket season. Injury history reported. Sample size was derived from data from [48] with an alpha of 0.05, beta 0.8, effect size of 2 and allocation ratio of 0.75

Spinal orientation at BFC and FFC, spinal range of motion, SCR and hip separation angle. Sacral accelerations. Peak tibia and sacral accelerations along three orthogonal axes with resultant acceleration. Normalised accelerations (to body weight), time-to-peak acceleration, loading rate and time taken to reach peak acceleration

Retrospective: history of LBP. Prospective; any LBP experienced during season. LBP defined as any pain affecting the area of the back inferior to the lower ribs, superior to the inferior gluteal folds and medial to the midaxillary line that impacted on ability to bowl for a minimum of 3 days

No statistically significant results found, but large effect sizes observed. In juniors without LBP history; more thoracic rotation away from direction of delivery (d = 1.3), and a larger range of thoracic rotation between BFC and FFC (d = 0.9). In seniors with LBP history; less peak acceleration around tibial z axis at BFC (d =  − 1.5), faster time-to- peak resultant tibial acceleration at FFC (d =  − 1.5), and greater thoracolumbar extension at BFC (d = 1.0) In seniors who did not develop LBP; higher time-to- peak resultant tibial acceleration at BFC (d = 1.55), higher tibial loading variables at FFC (d = 0.9), less lumbar extension (d = 1.9) at BFC, more lumbar lateral flexion away from direction of delivery at BFC (d = 1.0), less lumbar extension at FFC (d = 0.9) and less lumbar rotation at FFC (d = 1.3). In seniors who did develop LBP; greater time-to-peak vertical and resultant acceleration at the sacrum (d > 1.6)

Alway et al. [53]

Prospective cohort. Causality implied

n = 50. Elite cricket fast bowlers enrolled on an international performance pathway. Mean age 18.9 ± 1.9 years. Male. United Kingdom. No control group included. Study dates not described. Study duration two years. No reporting of injury history. No calculation of sample size

Kinematic parameters of the fast-bowling action; SCR, pelvis-shoulder separation, front leg plant angle, front and rear leg hip and knee angles, lumbopelvic angles, thoracolumbar side flexion and rotation. Kinetic parameters of the fast-bowling action; peak forces, average loading rates and impulse in the vertical and horizontal (braking) directions

LBSI defined as either stress reactions or stress fracture determined from radiological reports the England and Wales Cricket Board injury database. Stress reactions: evidence of bone marrow edema (without fracture line). Acute stress fractures: evidence of incomplete, complete, or multilevel stress fracture accompanied by BMO that suggested the fracture site was active.

At instance of BFC, LBSI bowlers had a more flexed rear hip (d > 0.8, p < 0.05) and knee (d > 0.5, p < 0.05), less contralateral thoracolumbar side flexion (d > 0.5, p < 0.05) and more contralateral thoracolumbar rotation (d > 0.8, p < 0.05). At FFC, LBSI bowlers had a more flexed front hip (d > 0.8, p < 0.05), more anterior pelvic tilt (d > 0.8, p < 0.05), and more extended lumbopelvic angles (d > 0.5, p < 0.05). At BR, LBSI bowlers displayed less contralateral thoracolumbar side flexion (d > 0.8, p < 0.05). LBSI bowlers had less extension of their front hip and more ipsilateral pelvic drop in transitions between BFC and BR. The best logistic model to predict LBSI included both rear hip angle at BFC and lumbopelvic angle at FFC, correctly classifying 88% of bowlers into injured or non-injured groups. For each 1° increment in rear hip angle at BFC, the odds of having a LBSI was a factor of 0.88 lower, while a 1° increment in lumbopelvic angle at FFC increased the odds of a LBSI by 1.25

Taylor et al. [64]

Retrospective case series. Causality implied

n = 38. Elite. Mean age screened group 21.2 ± 3.7 (18.5–24.5) years, self-control group 19.7 (18.5- 24.5) years, matched-control group 21.0 (18.4—24.2) years. Male. Australia. Collected data compared to self-control group (data for same bowler from closest eligible season) and matched-control group (data for a fast bowler closest in age to screened bowler). No reporting of injury history. Study dates not described. No calculation of sample size

Clinically relevant BMO detected at the left and right posterior vertebral arch at each vertebral level (L1-L5)

Symptomatic LBSI that caused missed time from bowling (time loss injury) with MRI identified BMO with or without a fracture

Bowlers with BMO intensity of 2.0 or higher on screening MRI were 1.8 times the risk of sustaining LBSI in the following 12 months compared to bowlers who did not have abnormal BMO detected on screening MRI; RR 1.8 (95% CI 0.6–5.5; p = 0.321)

Sims et al. [59]

Retrospective cohort. Causality implied

n = 222. Youth fast bowlers in elite pathway programs. Mean age 17.4 ± 1.1 years, age range 15.1–19.7 years. Male. Australia. No control group recruited, but non-injured bowlers acted as controls in multivariate analysis. July 2015 to March 2020. No reporting of injury history. No calculation of sample size

Height, weight, ranges of motion for ankle dorsiflexion, hip internal and external rotation, trunk lateral flexion. Lumbo-pelvic stability, Lumbar extension endurance hold time, single leg balance range, hip abduction and extension strength, aerobic fitness. Age at start of season. Number of balls bowled per day in training and matches. Ranges of SCR and lateral flexion during the fast-bowling action. Bowling speed.

MRI identified LBSI categorised as a stress reaction (bone oedema with no cortical breach) or stress fracture (bone oedema with cortical breach) with a bowler subsequently classified by medical staff as unavailable to train or play

49 of the 222 bowlers sustained a LBSI. Univariate analysis: 1) injured bowlers were younger (p = 0.005), taller (p = 0.007), performed less efficiently on Star Excursion Balance Test (front foot p = 0.006, back foot p = 0.005), and on average bowled more days per one week (p = 0.009), 4 weeks (p = 0.042) and 12 weeks (p = 0.008); 2) no difference between injured and non-injured groups in bowling technique variables. Multivariate analysis: risk of LBSI was 2.99 times higher for every year younger in bowlers aged between 15 and 20 years. Bowlers were 1.1 times more likely to be injured for every centimetre taller and 1.1 times more likely for every km/hr faster the ball was bowled. The multivariate model was able to explain 36% of the variance

Keylock et al. [41]

Prospective cohort. Causality implied

N = 40. Recruited from professional academies or schools and clubs with well-developed cricket programs. 15.5 ± 1.1 years. Male. United Kingdom. No control group included. Study dates not described. Study duration was 2.26 ± 0.03 years after baseline. No reporting of injury history. No calculation of sample size

Height (cm). Weight (kg). Fat-free mass (kg). Chronological age (years). Skeletal age (years). Skeletal maturity rating. DEXA measured L3 CL BMD (g/cm2), L4 CL BMD (g/cm2), L3 vertebral area (cm2), L4 vertebral area (cm2). CL ankle dorsiflexion (cm). CL ankle dorsiflexion (°). Sit and reach length (cm). Hip internal rotation (°). Hip external rotation (°). Bent knee fall out (°). Straight leg raise (°). Total balls bowled (n. balls). Bowling days per week (n. days). Peak acute workload (n. balls). Peak medium workload (n. balls)

Bowlers who developed symptoms during study with radiologically confirmed LBSI and asymptomatic bowlers who possessed LBSI at the end of the study. LBSI defined as either stress reactions (evidence of BMO without fracture line), or LSF (acute or chronic stress fracture identified by evidence of incomplete or complete stress fracture with BMO)

Chronological age significantly differed at the 0.05 alpha level between prospectively injured and uninjured bowlers (P = 0.006), with a large effect size (g = 1.396). Injured bowlers were 1.3 years older at the beginning of the season preceding injury than uninjured bowlers on average although there was little difference in average skeletal age or maturation (g = 0.274 and 0.611, respectively, P ≥ 0.278). 33% of the injured bowlers had delayed maturation, compared to 13% of the uninjured bowlers. Non-significant (P ≥ 0.090) large effect sizes of increased L3 and L4 contralateral BMD (g ≥ 0.812) in LBSI bowlers. Hip internal rotation of the contralateral leg was non significantly 7.2 degrees less in injured fast bowlers compared with uninjured bowlers (32.3 versus 39.5 degrees (g = 0.987)