HERO 2009

Methods	Healthy Eating to Reduce Overweight in people with type 2 diabetes (HERO) RCT, parallel, (n‐3 ALA vs low n‐3), 12 months Summary risk of bias: moderate or high
Participants	Overweight adults with non‐insulin treated diabetes N: 26 intervention, 24 control (analysed, intervention: 18 control: 17) Level of risk for CVD: moderate Male %: not reported Mean age in years (SD): 54 (8.7), not reported by arm Age range: 33‐70 years Smokers: not reported Hypertension: not reported Medications taken by at least 50% of those in the control group: lipid lowering drugs, oral hypoglycemics Medications taken by 20%‐49% of those in the control group: not reported Medications taken by some, but less than 20% of the control group: not reported Location: Australia Ethnicity: not reported
Interventions	Type: food supplement (walnuts) Comparison: ALA vs nil Intervention: 30 g/d snack portions of walnuts (provided 10% MUFA, 10% E PUFA, and a P/S ratio of 1.0) and advised not to take fish oil supplements. ALA dose not reported. Dose: ˜3 g/d ALA based on 30 g/d intake of walnuts Control: no supplements Both groups were given low‐fat isocaloric dietary advice (30% E fat (10% E SFA, 15% E MUFA; 5% E PUFA, P/S ratio of 0.5), 20% E protein and 50% E CHO) plus advice to brisk walk 30 min × 3 times/week Compliance: measured by erythrocyte membrane fatty acid levels which were similar in both groups Duration of intervention: 12 months
Outcomes	Main study outcome: change in body weight and % body fat Dropouts: 8 intervention, 5 control Available outcomes: all cause mortality (nil deaths), weight, visceral adipose tissue, lipids, glucose, insulin, HbA1c (body fat % and subcutaneous adipose tissue measured but too different at baseline to use) Response to contact: not yet attempted
Notes	Body fat % was too different between groups at baseline hence data not used Study funding: California Walnuts Commission
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation was conducted using a computerised random number generator by a researcher independent of the subject interface.
Allocation concealment (selection bias)	Unclear risk	No further details
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: "Subjects, but not dietitians, were blinded to the type of overall diet (a prepackaged 30 g snack portion of walnuts was given to the walnut group unbeknown to the controls)". However, there was no placebo supplement, so blinding easily broken
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Paper states "code was concealed from the researchers collecting data, as well as from subjects." However as participants could not be blinded outcome assessors may not have been (problem for measures of adiposity, not for biochemical measures)
Incomplete outcome data (attrition bias) All outcomes	High risk	High dropout rate 35 of 50 analysed (30% attrition rate)
Selective reporting (reporting bias)	Unclear risk	Trial was registered postanalysis
Attention	Low risk	Both groups appear to have had same level of attention
Compliance	High risk	ALA levels almost exactly the same in intervention and control
Other bias	Low risk	None noted