Munters 2013.

Methods	Evaluator‐blind, parallel‐group RCT
Participants	Sample size Intervention group: 5 adults with PM, 6 adults with DM, control group: 4 adults with PM, 6 adults with DM Inclusion criteria Diagnosis of definite or probable DM and PM Age 18 years Duration since diagnosis 6 months Exercising ≤ once a week Receiving stable medication for at least 1 month Exclusion criteria Severe heart or lung conditions Severe osteoporosis Not being able to exercise Baseline demographics 23 people with PM or DM (12 in the exercise group and 11 in the control group) were included in the 12‐week endurance exercise programme. All participants were in a stable disease phase with no to high damage and with unchanged medication for at least 1 month before inclusion in the study, as well as during the 12‐week intervention.
Interventions	Aerobic exercise and strength training vs no training Type of training and exercise Cycling exercises and muscular endurance exercises Intensity Aerobic exercise: during the first weeks, the exercise intensity was gradually increased from 50% up to 70% of the participants’ individual VO2 max. Strength training: 30%‐40% of 1RM Frequency 3 times/week, twice a week at a physical therapy department, once a week at home Setting Twice a week at the department of physical therapy at each of the 3 participating centres (Karolinska University Hospital, Sahlgrenska University Hospital, Uppsala University Hospital, Sweden) Duration Session: 1 h. Programme: 12 weeks Muscle groups Knee extensors Supervision Physical therapist at each participating centre
Outcomes	Primary: primary outcome changed during the study from the FI2 to VO2 max Secondary FI2, SF‐36,The Myositis Activities Profile for limitations in ADL, 1RM measurements, disease activity was performed using the core set measures developed by the International Myositis Assessment and Clinical Studies, the maximum load a patient can lift in a full ROM in 5 repetitions (5 VRM) of knee extensors left and right. Secondary Power performed (in W) at the time of VO2 max was recorded, time cycling to exhaustion performed with a constant power at 65% of baseline VO2 max, manual muscle testing of 8 muscle groups was performed (maximal isometric strength of neck flexors, middle deltoid, gluteus maximus, gluteus medius, biceps brachii, wrist extensors, wrist flexors, ankle dorsiflexors) Time‐points measured Before and after 12 weeks of control or training period
Dates	Participants were recruited from 2007‐2011
Funding/ declarations of interest	Supported by grants from the Myositis Association, the Swedish Research Council, the Swedish Rheumatism Association, funds at the Karolinska Institutet, Promobilia, and through “The regional agreement on medical training and clinical research (ALF) between Stockholm County Council and Karolinska Institutet.”
Notes	Outcomes were not presented for DM and PM separately.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "Using a randomisation list, patients were randomised by an independent nurse".
Allocation concealment (selection bias)	Low risk	Quote: "Using a randomisation list, patients were randomised by an independent nurse".
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not blinded as it is impossible to blind exercise training compared to no exercise training.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "assessments (...) were performed by a physical therapist at each respective centre, blinded to the type of intervention"
Incomplete outcome data (attrition bias) All outcomes	High risk	Quote: "One patient in the exercise group was not able to perform the exercise programme and was excluded from the analysis". Follow‐up was therefore incomplete and analysis was not by ITT. Quote: "We aimed for nine patients in the exercise group, but some analyses were performed with N = 7 (VO2 max measurements) or N = 3 (mitochondrial enzyme activities). (...) All measurements were not successfully performed both before and after training in each subject".
Selective reporting (reporting bias)	Unclear risk	Quote: "Because PM [polymyositis] and DM [dermatomyositis] are rare conditions, the research group decided at the study onset to perform an interim analysis in case patient recruitment did not proceed as quickly as projected. VO2 max was determined to be more sensitive to change than the FI‐2 according to the type of exercise performed, and thereby was selected instead as the primary outcome and used for the interim analysis". Comment: primary outcome changed at the study onset from the FI2 to VO2 max
Other bias	Low risk	Comment:no risk of bias from other sources detected