Johansen 1999.

Study characteristics
Methods	Study design Parallel RCT Study dates Duration of follow‐up: 6 months Time frame: April 1996 to July 1997
Participants	Study characteristics Setting: single‐centre (San Francisco General Hospital Medical Care outpatient dialysis unit) Country: USA Inclusion criteria: undergoing HD for at least 3 months; evidence of malnutrition; poor QoL as assessed by questionnaire; patients had to have two or more of the following to be considered for the assessment of malnutrition: albumin < 40 g/L, total cholesterol < 3.88 mmol/L (150 mg/dL), transferrin < 2 g/L, protein catabolic rate < 0.8 g/kg/day, predialysis serum urea nitrogen < 21.4 mmol/L (60 mg/dL), or insulin‐like growth factor 1 < 300 ng/mL Exclusion criteria: received dialysis for fewer than 3 months or if they had other reasons for being in a catabolic state, such as HIV, knowing malignancy, corticosteroid treatment, surgery, or infection requiring IV antibiotics, within 3 months; participation in other studies; illicit drug use Baseline characteristics Number (analysed/randomised): intervention group (12/14); control group (11/15) HD/PD: intervention group (10/4); control group (10/5) Mean age ± SD (years): intervention group (44 ± 15); control group (50 ± 10) HD: not reported PD: not reported Sex (M/F): intervention group (11/3); control group (12/3) HD: not reported PD: not reported Dialysis type: HD/PD Mean dialysis vintage ± SD (years): intervention group (2.9 ± 2.7); control group (2.3 ± 2.0) HD: not reported PD: not reported Comorbidities CVC: not reported Diabetes: treatment group (5/14); control group (6/15) HD: not reported PD: not reported Hypertension: not reported Depression (clinician diagnosis): not reported
Interventions	Intervention classification Pharmacological intervention Indication: study targeting fatigue Intervention group Nandrolone decanoate (IM): 100 mg once/week Control group Placebo (IM): saline solution once/week Co‐interventions The same equipment was used for baseline, 3‐month, and 6‐month evaluation for all patients
Outcomes	Outcomes reported Fatigue outcome measures used: validation data available Change in body weight: assessed at baseline, 3 and 6 months Electronic scale Change in body composition: assessed at baseline, 3 and 6 months Electronic scale Change in lean body mass: assessed at baseline, 3 and 6 months Grip strength: assessed at baseline, 3 and 6 months Handheld dynamometer Functional capacity Walking and stair‐climbing times: assessed at baseline, 3 and 6 months Peak oxygen consumption (VO2) Treadmill performance: assessed at baseline, 3 and 6 months Laboratory results (SCr, albumin, total cholesterol, transferrin, total and free testosterone, luteinizing hormone, follicle‐stimulating hormone, IGF‐1, HCT, Hb): assessed monthly Heart rate Treadmill performance: assessed at baseline, 3 and 6 months BP Treadmill performance: assessed at baseline, 3 and 6 months Change Kt/V: assessed until the end of the study HRQoL Questionnaire: assessed at baseline, 3 and 6 months Satisfaction Index of overall satisfaction: assessed at baseline, 3 and 6 months Eating dimension Sickness impact profile: assessed at baseline, 3 and 6 months Fatigue Profile of mood state: assessed at baseline, 3 and 6 months Anger/hostility components Profile of mood state: assessed at baseline, 3 and 6 months Adverse events Questionnaire: assessed at baseline, 3 and 6 months Sudden death: assessed until the end of treatment Hospitalisation: assessed until the end of treatment Severe hypertension: assessed until the end of treatment
Notes	Additional information Funding: grant RR‐00083 from the National Center for Research Resources, Bethesda, Md, grant DK‐45833 from the National Institute of Diabetes and Digestive and Kidney Disease, Bethesda, and a grant from the Bay Area Nutrition Center, Berkeley Conflicts of interest/disclosures: not reported Trial registration identification number: not applicable A priori published protocol: protocol was approved by the Committee on Human Research at the University of California, San Francisco
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote; "Randomisation was computer‐generated in block of 4." Comment: Computer‐generation is considered as low risk of bias. No imbalance between intervention groups was apparent
Allocation concealment (selection bias)	Low risk	Quote: "Assignments were made sequentially by a research pharmacist who dispensed medication but was not otherwise involved in the study." Quote: "Esternal research pharmacist seemed to ensure allocation concealment. No imbalance between intervention groups was apparent."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Dialysis staff, patients, and investigators were blinded through the study to treatment assigned." Comment: A double‐blind trial is considered as low risk of bias
Blinding of outcome assessment (detection bias) All outcomes	High risk	Quote: "Quality of life was assessed by and instrument administered by personal interview." Comment: The outcomes were assessed with an appropriate measure, without differences between groups. However, subjective measures were used, it was not stated whether outcomes were assessed without knowledge of treatment allocation, and knowledge of treatment assignment may have influenced reporting. Participant beliefs about the superiority/inferiority of either intervention could have influenced their assessment of the outcome, but there was no evidence that this was likely. However, objective and subjective outcomes were assessed
Incomplete outcome data (attrition bias) All outcomes	High risk	Quote: "25 subjects completed the 6‐month protocol and 23 of these (12 in the nandrolone group and 11 in the placebo group) had all measurements made. Two subjects completed the study but were unable to have final measurements taken because of medical instability. Three subjects were withdrawn from the placebo group because of elevated transaminase, hematoma at the study drug injection site, ans sudden death. One subject in the nandrolone group was withdrawn after developing angina." Comment: 12/14 participants in the intervention group and 11/15 participants in the control group completed and reported all measurements of the study (> 5% lost to follow‐up, with differences between groups). Reasons for discontinuations seemed to be related to the treatment allocation
Selective reporting (reporting bias)	High risk	Protocol was approved by the Committee on Human Research at the University of California. Fatigue was reported using multiple eligible outcome measurements (scales, time points). Fatigue was reported in a format that was extractable for meta‐analysis. All outcomes that should be addressed (fatigue, cardiovascular disease, and death) were not reported
Other bias	Low risk	There was no evidence of different baseline characteristics, or different non‐randomised co‐interventions between groups. Funding was unlikely to influence the data analysis and conflicts of interest were not reported