Table 2

Characteristics of included studies. Mean unless otherwise stated

AuthorRecruitedDropout rate
(% of total recruited), reason and number included
Population characteristics of total sampleMethod of measuring dyskinesisFollow-up duration and method of injury reportingPrevalence and severity of shoulder pain events
(% of total included in meta-analysis)
Clarsen et al 24
206 elite handball players
With and without self-reported shoulder pain but nil on testing
Participants with self-reported pain at baseline excluded from meta-analysis
Did not complete sufficient questionnaires
164 analysed in study
110 meta-analysed
Age 24 (SD 4)
All male
Height 189 cm (SD 7)
Weight 89 kg (SD 10)
Playing age 14 years (SD 5)
Elite playing age 4 years (SD 4)
Right handed (73%)
Back players (42%)
Wing players (23%)
Line players (15%)
Goalkeepers (14%)
Combination of positions (6%)
Live visual observation by one physiotherapist, with video playback if uncertain
5× reps of shoulder flexion and abduction using 5 kg weights
Categorised as subtle dyskinesis, obvious dyskinesis or normal
One season—9 months.
OSTRC overuse injury questionnaire emailed to all players every second Sunday— total 15 questionnaires
If <4 questionnaires returned, data excluded.
Each questionnaire gives a severity score (0–100).
Average severity score = sum of scores/number of questionnaires.
Injury =
cut-off in study >40
cut-off in meta-analysis>0
Average cumulative severity score:
9.91 (SD 13.34)
Kawasaki et al 36 70 elite rugby playersn=8
Retired for reasons unrelated to the shoulder
62 meta-analysed
Age 24.6 (SD 3.3)
All male
BMI 28.9 (SD 3.6)
Playing age 12.7 (SD 4.8)
Position (forward: backward) 61:42
Dominant side (right: left) 99:4
Frequent collision side (right: equal: left) 65:30:8
Live visual observation by two sports doctors, with video playback if uncertain
5× reps of shoulder flexion and abduction in scapular plane using 3 kg weights.
Categorised as type 1–3 dyskinesis or type 4 = Normal
One elite Japanese rugby season—length not specified— via questionnaire
Injury = subjective discomforts (pain, apprehension or fatigue) persisting more than 2 weeks; and shoulder trauma requiring off-game>7 days
Discomfort persisting >2 weeks
Severity not reported
McKenna et al 35
46 adolescent swimmers
43 adolescent non-swimmers (excluded)
Did not respond to questionnaire
All from non-swimmer group
46 meta-analysed.
[no pain, pain]
Female gender: 65.7%, 63.6%
Age (years) 14.5 (SD 1.4), 14.9 (SD 1.7)
Matured by start of study: 51.6%, 45.5%
Previous shoulder or arm pain: 35.3%, 50.0%
Number of weekly training sessions
6.7 (3–10), 6.6 (5–10)
Swimming age (years)
3.6 (SD 2.6) 4.7 (SD 2.3)
Freestyle swim time (seconds)
67.1 (SD9.1), 66.1 (SD6.5)
Height (m)
1.65 (SD 0.09), 1.65 (SD 0.09)
Weight (kg)
54.7 (SD 10.4), 57.9 (SD 9.7)
19.8 (SD 2.3), 21.3 (SD 2.2)
Chest width (cm)
30.1 (SD 3.1), 30.0 (SD2.9)
Anthropometric tape measures
In neutral, hands on hips, 90 degrees’ abduction + internal rotation and full flexion
Distances measured between:
T7 → inferior scapula
T3 → medial spine scapula
Humeral head position in relation to acromion using palpation, photography and digital calculation
Cut-off for dyskinesis determined by post hoc receiver operated curve analysis to find best fit sensitivity/specificity.
12 months later via questionnaire
Injury = positive answer to the question ‘Have you had any pain in your shoulder in the last year?’
Severity not reported.
Shitara et al 34
132 high school baseball pitchers
With and without self-reported shoulder pain, but nil on testing
Participants with self-reported pain at baseline excluded from meta-analysis.
Failed to provide consent or did not complete questionnaire
105 analysed in study.
88 meta-analysed.
Median age 16.3 (SD 0.6)
All male
Baseball experience (years)
Non-injured 8.1 (SD 2.0)
Injured mean 8.4 (SD 2.2)
Past shoulder pain
Non-injured 34 (40.5%)
Injured 7 (33.3%)
Past elbow pain
Non-injured 44 (52.4%)
Injured 13 (61.9)
Present shoulder pain (excluded)
Non-injured 12 (14.3%)
Injured 5 (23.8%)
Present elbow pain
Non-injured 19 (22.6%)
Injured 6 (28.6%)
Live visual observation by two orthopaedic surgeons, with video playback if uncertain
3–5× reps of shoulder flexion and abduction in scapular plane without weights
Categorised as yes/no
One high school baseball season—length not specified
Self-reported at medical check-ups and via end of season questionnaire
Injury = unable to participate in training/games for ≥8 days
Injuries from other mechanisms than throwing excluded.
Required >7 days of missed participation in training/game
Severity not reported.
Struyf et al 37 196 recreational overhead athletes:
Volley ball 37
Tennis 26
Baseball 5
Badminton 35
Handball 10
Not willing to be contacted, cessation of overhead activity, no reason, emigration or impossible to get in contact with
113 meta-analysed.
Age 34 (SD 12)
Men (48%)
Right handed (89%)
Must play >1 hour per week of preferred sport
Live visual observation by one physiotherapist
At rest, hands on hips, hands at 90° abduction
3× reps of shoulder abduction without weights
Categorised as winging, forward tilt or normal
24 months
Contacted by phone at 12 and 24 m
Injury = any physical complaint recalled lasting at least 1 day in the previous month, irrespective of disability induced
Severity measured with Shoulder Disability Questionnaire.
Mean SDQ =34.8 (SD 17.4)
  • Modified data excluded the results of participants with pain at baseline in two studies30 39 or dichotomisation of continuous measurements of scapular position in one study.40

  • **Indicates a significant difference between groups.

  • BMI, body mass index; OSTRC, Oslo Sports Trauma Research Centre; Reps, repetitions; SDQ, Shoulder Disability Questionnaire.