Twiss 2009.

Study characteristics
Methods	Study design: randomised controlled trial Study grouping: parallel group
Participants	Baseline characteristics 100% women Age (yr): 58.69 (SD 7.5) Inclusion criteria: ≥ 12 months postmenopausal, aged 35–75 years, history (≥ 6 months' post‐treatment) of stage 0, I or II breast cancer, osteopenia or osteoporosis as defined by a bone mineral density T score of –1.0 or less at hip, lumbar spine or forearm, reside within 100 miles of research sites at Omaha, Lincoln, Kearney, and Scottsbluff Nebraska, physicians permission to participate Exclusion criteria: recurrence of breast cancer; currently taking hormone therapy, bisphosphonates, glucocorticosteroids, or other drugs affecting bone density; currently engaging in strength exercises; body mass index ≥ 35 kg/m², serum calcium, creatinine or thyroid‐stimulating hormone outside normal; active gastrointestinal problems or other conditions that prohibited strength exercises; intake of vitamin D or risedronate Group differences: not reported
Interventions	Usual care Not stated Exercise 2 × 30–45 minutes for 32 weeks at home. Gym‐based programme for 72 weeks (session duration and number of sessions per week not reported)
Outcomes	Tandem balance Outcome type: continuous Reporting: full Scale: time Unit of measure: seconds Direction: lower was better Falls Outcome type: adverse event Reporting: full Scale: numbers Direction: lower was better
Identification	Sponsorship source: funded by "NINR" Country: USA Setting: exercises delivered at home then at fitness centre. Testing a hospitals or rehabilitation centres at 4 sites. Comments: none Author's name: Nancy Waltman Institution: University of Nebraska Medical Center College of Nursing, Omaha E‐mail: nwaltman@unmc.edu Address: University of Nebraska Medical Center, College of Nursing, Commerce. Court PO Box 880220, Lincoln, NE, USA
Notes	All participants were grouped together at baseline
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Authors stated that participants were randomly allocated; however, did not detail the sequence generation process
Allocation concealment (selection bias)	Unclear risk	No information provided regarding allocation concealment to allow a judgement to be made
Blinding of participants and personnel (performance bias) All outcomes	High risk	No attempt to blind either the participants (understandably) or the personnel assessing outcome. Those involved in the intervention appeared to be making outcome assessments. It was possible that knowledge of the intervention group could have biased outcome assessments
Blinding of outcome assessment (detection bias)	High risk	Outcome assessors appeared to be involved in delivery of the intervention
Incomplete outcome data (attrition bias) All outcomes	Low risk	26/249 women randomised into the study did not complete the 24‐month testing. The dropout rate of the intervention group was comparable to that of the control group (14 vs 12) and reasons for withdrawal did not indicate potential for bias
Selective outcome reporting (reporting bias)	Unclear risk	Protocol not available, insufficient evidence to make a decision
Size of study	Unclear risk	Between 50 and 199 participants per treatment arm
Other sources of bias	Low risk	Study appeared free of other sources of bias