Table 2

Characteristics of included trials

Study characteristicsParticipant characteristicsIntervention and settingOutcome data/results
Aasa et al (2015)43
Michaelson et al (2016)72
2 groups:
1. High-load lifting exercise
2. Low-load motor control exercises		70 patients recruited from occupational healthcare services in Sweden (mean age 42, 56% female);
inclusion criteria included: (a) adults with low back pain >3 months’ duration and (b) with or without leg pain		Physiotherapy clinic, sports centre and home setting
1. n=35; group exercises based at a sports centre (5 participants in each group), with pain up to 50 mm Visual Analogue Scale acceptable, such that the pain subsided after each set of exercises; 12 treatment sessions over an 8-week period (weeks 1–4, 2 sessions per week; weeks 5–8, 1 session per week); 60 min in duration; no home exercises
2. n=35; pain-free individual exercises at a physiotherapy centre; 12 treatment sessions over an 8-week period (weeks 1–4, 2 sessions per week; weeks 5–8, 1 session per week); 20–30 min in duration; exercises involved improving control around joint neutral positions; in supine, four-point kneeling, sitting, and/or standing positions; Plus home exercises, 10 repetitions 2–3× a day		Main outcome assessed at baseline, 2-month and 12-month follow-up was 7 day average pain on a Visual Analogue Scale (0–100 mm) and Roland-Morris Disability Questionnaire (0–24)
Group 1: mean pain at baseline 43 (SD 24), 2 months 22 (SD 21), 12 months 24 (SD 27) and 24 months 27 (SD 27)
Group 2: mean pain at baseline 47 (SD 28), 2 months 30 (SD 26), 12 months 25 (SD 22) and 24 months 30 (SD 29)
Group 1: mean disability at baseline 7.2 (SD 4.3), 2 months 3.8 (SD 4.0), 12 months 3.6 (SD 4.2) and 24 months 3.8 (SD 3.9)
Group 2: mean disability at baseline 7.1 (SD 3.9), 2 months 3.6 (SD 4.2), 12 months 3.3 (SD 3.6) and 24 months 3.6 (SD 3.7)
Both groups had significant improvements in their pain and disability levels; no significant between-group difference for pain at any follow-up (2 months p=0.71; 12 months p=0.94; 24 months p=0.89); no significant between-group difference for disability at any follow-up (2 months p=0.77; 12 months p=0.74; 24 months p=0.99)
Holmgren et al (2012)70
Hallgren et al (2014)71
2 groups:
1. Specific exercises group
2. Control exercise group
Patients were given the option at 3 months of continuing to have an arthroscopic subacromial decompression.		97 patients recruited from the waiting list for an arthroscopic subacromial decompression from a university hospital in Sweden (mean age 52, 37% female);
inclusion criteria included (a) adults with lateral shoulder pain >6 months, (b) failed 3 months of previous primary care, (c) signs of impingement symptoms and (d) positive Neer’s impingement test of a subacromial anaesthetic injection		Physiotherapy and home setting
1. n=51; eccentric rotator cuff exercises and concentric/eccentric scapula exercises; recommendation of 5/10 numerical rating scale for pain during exercises, such that the pain subsided by the next exercise session; 7 physiotherapy appointment, weekly first 2 weeks, alternative weeks thereafter; exercises to be performed at home once or twice a day for 12 weeks
2. n=46; pain-free upper limb and neck exercises; 7 physiotherapy appointment, weekly first 2 weeks, alternative weeks thereafter; exercises to be performed at home once or twice a day for 12 weeks		Main outcome of Constant-Murley score (C-M) (0–100), along with shoulder assessment scores and pain scores taken at baseline, 3 months and 12 months, including pain at rest measured on Visual Analogue Scale (0–100 mm)
Group 1: mean C-M at baseline 48 (SD 15), 3 months 72 (SD 19) and 12 months 83 (SD 14)
Group 2: mean C-M at baseline 43 (SD 15), 3 months 52 (SD 23) and 12 months 76 (SD 18)
Group 1: mean pain at rest at baseline 15 (SD 19), 3 months 10 (SD 14) and 12 months 2 (SD 6)
Group 2: mean pain at rest at baseline 20 (SD 21), 3 months 20 (SD 25) and 12 months 4 (SD 13)
Both groups had significant improvements in all outcomes at 3 months and 1 year follow-up. Significantly more patients in the control group decided to have surgery (63%) than those in the specific exercise group (24%; p<0.0001).
Littlewood et al (2015)52
2 groups:
1. Self-managed exercises
2. Usual physiotherapy		86 patients recruited from UK, NHS physiotherapy waiting list (mean age 55, 50% female);
inclusion criteria included (a) adults with shoulder pain >3 months, (b) maintained shoulder ROM and (c) pain with resisted movements		Physiotherapy and home setting
1. n=42; single shoulder exercise guided by the symptomatic response, requiring pain to be produced during exercise, such that the pain subsided after the exercises; typically involving a weighted shoulder abduction exercise of 3 sets of 10–15 repetitions; pragmatic approach to number of follow-ups, timings of appointments and point of discharge; that is, the treating physiotherapist and patient will determine these factors
2. n=44; usual physiotherapy,* including advice, stretching, exercise, manual therapy, massage, strapping, acupuncture, electrotherapy, corticosteroid injection at the discretion of the treating physiotherapist; pragmatic approach to number of follow-ups, timings of appointments and point of discharge; that is, the treating physiotherapist and patient will determine these factors		Main outcome of the Shoulder Pain and Disability Index (SPADI) (0–100) at baseline, 3, 6 and 12 months
Group 1: mean at baseline 49.1 (SD 18.3), 3 months 32.4 (SD 20.2), 6 months 16.6 (SD 19.7) and 12 months 14.2 (SD 20.0)
Group 2: mean at baseline 49.0 (SD 18.0), 3 months 30.7 (SD 19.7), 6 months 24.0 (SD 19.7) and 12 months 21.4 (SD 25.4)
Statistically significant and clinically important within group changes for SPADI from baseline to all three follow-up points.
There were no statistically significant differences between the groups across all the outcomes at 3, 6 or 12 months, (p=0.75, 0.19 and 0.32, respectively).
Maenhout et al (2013)47
2 groups:
1. Traditional rotator cuff training with heavy load eccentric training
2. Traditional rotator cuff training		61 patients recruited from a shoulder surgeon’s clinic in Belgium (mean age 39.8, 41% female);
inclusion criteria included (a) adults with >3 months of shoulder pain, (b) painful arc, (c) 2 out of 3 impingement tests, (d) pain on palpation of rotator cuff tendons		Physiotherapy and home setting
1. n=31; the same exercises as group 2, plus a heavy loaded eccentric exercise of abduction within the scapular plane; 3 sets of 15 repetitions, such that the patient experiences pain on the last set, up to 5/10 Visual Analogue Scale, such that the pain subsided by the following morning.
2. n=30; pain-free, traditional rotator cuff exercises of internal and external rotation with a resisted rubber band; performed once a day, with 3 sets of 10 repetitions;
both groups had exercise prescription and monitoring through 9 physiotherapy appointments over 12 weeks		Main outcome of the SPADI (0–100) at baseline, 6 weeks and 12 weeks
Group 1: mean at baseline 44.3 (SD 11.5), 6 weeks 17.7 (SD 12.0) and 12 weeks 14.5 (SD 11.7).
Group 2: mean at baseline 42.0 (SD 11.0), 6 weeks 25.4 (SD 11.9) and 12 weeks 17.0 (SD 11.4)
In both groups pain and function, measured with the SPADI score, improved significantly over time (p>0.001). When comparing between groups, improvement of the SPADI score was not significantly different.
Nørregaard et al (2007)73
2 groups:
1. Eccentric exercises
2. Stretching exercises		45 patients recruited from a clinic of sports medicine in Denmark (mean age 42, 49% female);
inclusion criteria included (a) adults with Achilles pain >3 months, (b) local thickening >2 mm on ultrasound, (c) diffuse posterior ankle pain		Sports medicine clinic and home setting
1. n=21; information leaflet with home exercise programme on; to be performed twice a day, for 12 weeks; 1 follow-up appointment at 3 months; 3 sets of 15 repetitions of eccentric calf exercises, with knee straight and semi-flexed; patients told to expect pain during the exercises, but to avoid increasing daily pain or morning stiffness
2. n=24; information leaflet with home exercise programme on; to be performed twice a day, for 12 weeks; 1 follow-up appointment at 3 months; pain-free standing stretches for gastrocnemius and soleus; 5 repetitions of 30 s each		Outcome measures were tenderness on palpation, ultrasound and pain, as measured by the Knee Injury and Osteoarthritis Outcome Score (KOOS) (0–4) and patient’s global assessment; follow-up was at baseline, 3, 6, 9, 12 weeks and 1 year
Group 1: mean pain domain from KOOS at baseline 1.6 (SD 0.6), 3 weeks 0.1 (SD 0.1), 6 weeks 0.3 (SD 0.1), 9 weeks 0.4 (SD 0.2), 12 weeks 0.4 (SD 0.2) and 1 year score was 1.0 (SD 0.2)
Group 2: mean pain domain from KOOS at baseline 1.6 (SD 0.6), 3 weeks 0.2 (SD 0.1), 6 weeks 0.3 (SD 0.1), 9 weeks 0.3 (SD 0.2), 12 weeks 0.4 (SD 0.2) and 1 year score was 0.7 (SD 0.2)
There were significant improvements in all dimensions of the KOOS compared with baseline, with no differences between group differences.
Rathleff et al (2015)51
2 groups:
1. High-load strengthening exercises
2. Stretching exercises		48 patients recruited from a university hospital, regional hospital and private clinic in Denmark (mean age 46, 66% female);
inclusion criteria included (a) adults with plantar fasciitis >3 months, (b) pain on palpation, (c) local thickening >4 mm on ultrasound		Home based exercises
1. n=24; information leaflet, heel inserts and a prescription of a high-load strength programme; consisting of single calf raises with a towel rolled up under the toes for maximum toe extension, activating the windlass mechanism; each calf raises was 3 s up, 2 s pause, 3 s down; weight was added in rucksacks, starting at 12 repetition maximum for three sets, and slowly progressed over 3 months; patients were advised to perform the exercise every other day; exercises were allowed to be painful, with no postincrease in pain.
2. n=24; information leaflet, heel inserts and a prescription of pain-free* plantar-specific stretches; patients were asked to stretch the plantar fascia in a cross-legged position by extending their toes, hold for 10 s, 10 times, 3× a day for 3 months		Primary outcome was Foot Function Index at 1, 3, 6 and 12 months, including pain at worse and pain on first step on a numerical rating scale (0–10).
Mean scores for group 1 pain at worse at baseline was 7.9 (SD 1.7), 1 month 6.1 (95% CI 5.1 to 7.2), 3 months 3.5 (95% CI 2.3 to 4.7), 6 months 2.5 (95% CI 1.4 to 3.6) and 12 months 2.9 (95% CI 1.7 to 4.0). Mean scores for group 2 pain at worse at baseline was 7.5 (SD 1.6), 1 month 6.1 (95% CI 5.2 to 7.1), 3 months 6.1 (95% CI 4.4 to 7.7), 6 months 3.4 (95% CI 2.0 to 4.7) and 12 months 1.8 (95% CI 0.7 to 3.0).
At 3 months group 1 had significantly lower pain scores than group 2 (p<0.05). At months 1, 6 and 12, there was no significant difference between groups.
Silbernagel et al (2001)74
2 groups:
1. Eccentric exercises
2. Regular concentric/eccentric exercises		40 patients recruited from mailings to hospitals, clinics and sports clubs in Sweden (mean age 45, 23% female); inclusion criteria included (a) adults with Achilles pain >3 monthsClinic and home setting
1. n=22; progressive eccentric exercise programme to be performed 2× a day, plus three sets of six different stretching exercises, 20 s each, as well as balance, toe/heel walking exercises; weekly physiotherapy contact for 12 weeks; pain was allowed during the exercises up to 5/10 Visual Analogue Scale, such that the pain subsided by the following morning with no morning stiffness
2. n=18. 3 x a day of regular concentric and eccentric calf strengthening, plus two sets of the stretching exercises from group 1. Physiotherapy contacts 3–5 x during the 12 weeks. Exercises must be pain-free.		Outcomes of pain on palpation (Visual Analogue Scale) (0–100 mm) taken at baseline, 6 weeks, 3 and 6 months. Other outcomes included pain on walking and pain on stairs (yes/no), various objective measures, plus a non-validated functional questionnaire
Median±IQR scores for pain on palpation for group 1 at baseline was 49±26.2, 6 weeks 40±27.5, 3 months 35±24.8 and 6 months 21±20.
Median±IQR scores for group 2 at baseline was 27±21.5, 6 weeks 20±20, 3 months 31±26 and 6 months 9±17.5.
There was a significant decrease in pain on palpation in both treatment groups; no significant differences between groups were seen.

  • *Information not within publication, authors contacted for clarification.

  • ROM, range of motion.