Baptista 2009.

Study characteristics
Methods	Study design: Randomized controlled trial Study grouping: Parallel group How were participants recruited: Recruitment was conducted among those participants willing to lose BW or prevent excessive BWG after an educational program of healthy life style. Voluntary participation was requested and a written informed consent was obtained from each patient and from the medical director of the institution. Type of RCT: Open
Participants	Baseline characteristics Rosiglitazone Sex (N, % female, % male): female 6 (43%), male 8 (57%) Age (mean, SD): female 48.7 (SD 9.2), male 40.9 (SD 9.2) years Number of participants: 14 Placebo Sex (N, % female, % male): female 9 (60%), male 6 (40%) Age (mean, SD): female 48.8. (SD 6.7) and male 55.3 (SD 6.7) years Number of participants: 15 Overall Sex (N, % female, % male): female 15 (52%), male 14 (48%) Number of participants: 29 Inclusion criteria: Psychiatric inpatients with schizophrenia who switched to olanzapine 8 months before the study, older than 18, free of hormone replacement and any chronic disease besides the mental disorder and normal physical and laboratory tests. Patients had to be willing to lose weight or prevent excessive weight after an educational healthy lifestyle program. Exclusion criteria: ‐ Pretreatment: Insulin levels seem higher at baseline in rosiglitazone (5.1) than placebo (2.9) group. Men in the placebo group were older (55) than in rosiglitazone group (41).
Interventions	Intervention characteristics Rosiglitazone Class of drug: Antidiabetic Dose: 4‐8 mg Frequency: Daily Duration: 12 weeks Who delivered the intervention? Nurses trained in research Type of intervention: Pharmacological Placebo Class of drug: Placebo Dose: 8 mg Frequency: Daily Duration: 12 weeks Who delivered the intervention? Nurses trained in research Type of intervention: Control
Outcomes	Glucose Outcome type: Continuous Reporting: Fully reported Unit of measure: mg/dL Direction: Lower is better Data value: Endpoint BMI Outcome type: Continuous Reporting: Fully reported Unit of measure: kg/m2 Direction: Lower is better Data value: Endpoint Total cholesterol Outcome type: Continuous Reporting: Fully reported Unit of measure: mg/dL Direction: Lower is better Data value: Endpoint Waist circumference Outcome type: Continuous Reporting: Fully reported Unit of measure: cm Direction: Lower is better Data value: Endpoint
Identification	Sponsorship source: This study was supported by FONACIT, Caracas, Venezuela, Grant2005‐000‐384 Country: Venezuela Setting: inpatient psychiatric institute Comments: N=30 at baseline but N=1 left study; as treated analysis and no baseline information. Authors name: Prof. T. Baptista, MD, PhD Institution: Los Andes University Medical School, Venezuela Email: trinbap@yahoo.com Address: Department of Physiology, Los Andes University Medical School, P.O. Box 93, Mérida 5101‐A, Venezuela
Notes	Interventions Recruitment was conducted among those participants willing to lose BW or prevent excessive BWG after an educational program of healthy life style. Recommendations for healthy food and physical exercise to control BWG were provided at the beginning of the study. All participants engaged in healthy food and monitored physical exercise programs but their effects could not be quantified.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Subjects were randomly assigned through a computer‐based program to either rosiglitazone (Avandia, Glaxo‐Smith‐Kline, Caracas, Venezuela) or placebo." Judgement comment: Further details could have given clarity on stratification, etc.,
Allocation concealment (selection bias)	Unclear risk	Quote: "Medication and placebo were administered as a single daily dose by nurses trained in research and the number of pills was weekly quantified." Judgement comment: Not specifically mentioned whether nurses were aware of allocation
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Quote: "Rosiglitazone was started at the lower dose (4 mg), and at week 4 the full rosiglitazone (8 mg) or placebo dose was achieved and maintained thereafter in all subjects. The rosiglitazone schedule encompassed the usual dose range in clinical practice. Since rosiglitazone was devoid of significant side effects, no dose adjustment was necessary." Judgement comment: Blinding of participants is not mentioned clearly. Blinding of assessment personnel is not mentioned clearly. Risk of bias due to changing doses.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Quote: "Each subject ’ s blood pressure, BW and body mass index (BMI = weight in kg / height squared) were measured at baseline, weeks 6 and 12." Judgement comment: Outcomes are observer reported without judgement, even though assessment personnel are not identified as nurses, psychiatrists or social work professionals. Blinding has not been specifically mentioned.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "One subject in the rosiglitazone group left the study in the first week of treatment due to a change in residence. Hence, it was not possible to conduct either an “ intention to treat ” or a “ last observation carried forward ” analysis." Judgement comment: As‐treated analysis but unlikely to bias results as only one participant dropped‐out due to change in residence. For the rest, outcome data is completely reported.
Selective reporting (reporting bias)	Unclear risk	Judgement comment: No protocol referenced, although it is mentioned that a protocol was written and approved.
Other sources of bias	High risk	Judgement comment: Placebo group had higher cholesterol levels at baseline, and higher psychiatric symptoms measured by BPRS.