Table 3 Characteristics of the 7 cohort studies on risk factors according to the PICO-framework

Asker et al. (2020) [5]

471 female (54%) and male adolescent elite handball players from handball-profiled schools in Sweden, 15–19 years old, free of shoulder injuries at baseline

Baseline shoulder examination and questionnaire followed by weekly online monitoring of shoulder injuries over one (2014/15) or two seasons (2015/16)

Gender differences in relationship between shoulder injury incidence (dominant side) during handball play and shoulder ROM, strength, scapular dyskinesia, and joint position sense

Shoulder incidence was 0.92/1000 hrs in females and 0.71/1000 hrs in males; positive relationship between isometric shoulder external/internal rotation strength deficit and injury risk in females (hazard rate ratio: < 2.37) but not in males (< 1.02), positive relationship between scapular dyskinesia during abduction and injury risk in males (3.43) but not in females (1.53), no association with internal/external/total rotational ROM (< 1.56) and joint position sense (< 1.14) in both genders

Asker et al. (2018) [4]

471 female (54%) and male adolescent elite handball players from handball-profiled schools in Sweden, 15–19 years old, free of shoulder injuries at baseline

Baseline questionnaire followed by weekly online monitoring of shoulder injuries over one (2014/15) or two seasons (2015/16)

Gender, school grade, playing position, and playing level differences in shoulder injury prevalence

Shoulder prevalence was higher in females (< 48%, prevalence ratio: < 1.46) than in males (< 39%, 1.00) and higher in backcourt players (< 51%, < 1.58) than in other positions (< 40%, < 1.00), no differences in school grade and playing level

Hams et al. (2019a) [20]

218 female (59%) and male adult sub-elite water polo players from Australia, mean age of 19.3 and 20.6 years

Self- (2009–2013) and physiotherapist-report (2014–2016) on injury data of several body areas over up to 5 years

Body area, gender, and training/match differences in injury incidence rate, mechanism of injury, and injury burden

Shoulder incidence rate was 0.65/1000 training days, 25% (self-report) and 16% (physiotherapist-report) of all injuries being shoulder injuries with no gender differences (p = 0.33), proportion of shoulder injuries (16–25%) was higher than for all other body areas (11–17%) (p < 0.01), 67% of shoulder injuries were due to overuse and 33% due to trauma, more injuries in training (48%) than match (24%), each shoulder injury resulted in 6 days of training lost and 47 days in modified training

Matsuura et al. (2017) [26]

900 youth baseball players from Japan participating in regional summer championship, 7–11 years old

Baseline questionnaire and at one year follow-up (2012–2013)

Multivariate relationships between shoulder/elbow pain and age, playing position, length of baseball experience, training hours per week, and history of shoulder/elbow pain

Shoulder and elbow pain was evident in 18% and 35% of players; shoulder pain was only related to pitcher/catcher position, training hours per week, and history of shoulder/elbow pain (all p < 0.05). No relationship with age (p > 0.42) and length of experience (p > 0.52)

Oliver et al. (2019) [35]

Softball players from 100 high schools in USA, ~ 16 years old on average

Surveillance system on shoulder and elbow injury data (2005–06 to 2016–17)

Differences in injury rate, mechanism of injury, match/training occurrence, playing position distribution, injury burden, and further characteristics of shoulder and elbow injuries

Shoulder injury rate (1.14/1000 athlete-exposure) was higher than for elbow (0.41/1000), shoulder injury rate was higher in match (1.33/1000) than training (1.04/1000), 50% of shoulder injuries were due to overuse and most common diagnoses were muscle strains (31%) followed by tendinitis (24%), 17% of shoulder injuries were sustained by pitchers, 87% of all shoulder injured players returned to play within 21 days and remaining 13% did not return

Saper et al. (2018) [40]

Baseball players from 100 high schools in USA, ~ 16 years old on average

Surveillance system on shoulder and elbow injury data (2005–06 to 2016–17)

Differences in injury rate, mechanism of injury, match/training occurrence, playing position distribution, injury burden, and further characteristics of shoulder and elbow injuries

Shoulder injury rate (1.39/1000 athlete-exposure) was higher than for elbow (0.86/1000), shoulder injury rate was higher in match (1.73/1000) than training (1.20/1000), 71% of shoulder injuries were due to overuse and most common diagnoses were muscle strains (31%) followed by tendinitis (19%), 40% of shoulder injuries were sustained by pitchers, 90% of shoulder injuries were managed non-surgically, 87% of shoulder injured players returned to play within 21 days

Wilk et al. (2014) [54]

296 professional baseball pitchers from the USA participating in the major and minor league, 24.7 years old on average, free of shoulder injuries at baseline

Baseline shoulder examination and questionnaire over eight seasons (2005–2012)

Differences in shoulder ROM between dominant/non-dominant side and relationship with shoulder injuries and surgeries

17% of all pitchers suffered a shoulder injury, 7% required a surgery whereby most were labral and rotator cuff debridements (35%) followed by labral repairs (30%) and debridements (20%), pitchers showed less shoulder internal/total rotation ROM but higher external ROM in their dominant than non-dominant shoulder (all p < 0.01), shoulder internal/total rotation ROM deficits were not related to shoulder injury or surgery (p > 0.21) but shoulder external ROM deficit increased the likelihood to sustain shoulder injury (2.2 times higher) and surgery (4.0 times)