SU.FOL.OM3 2010.

Methods	Supplementation en Folates et Omega 3 (SU.FOL.OM3) RCT, 2 × 2 factorial (LCn3 omega 3 vs placebo, also B vitamin comparison), 4 years Summary risk of bias: low
Participants	People with a history of MI, unstable angina or ischemic stroke N: control: 1248, intervention: 1253 Level of risk for CVD: high Men: 80.85% intervention, 78.25% control Mean age in years (SD): 61.1 (8.8) intervention, 60.8 (8.7) control Age range: 53‐68 years intervention, 54‐68 years control Smokers: 11.1% intervention, 10.4% control Hypertension: not reported Medications taken by at least 50% of those in the control group: beta‐blockers, aspirin or antiplatelets, lipid lowering, ACE inhibitors Medications taken by 20%‐49%: not reported Medications taken by some, but < 20%: calcium channel blocker, angiotensin II receptor blockers Location: France Ethnicity: not reported
Interventions	Type: supplement (capsule) Comparison: EPA + DHA vs non fat placebo Intervention: 2 gelatin capsules Pierre Fabre omega 3 (400 mg/d EPA and 200 mg/d DHA) Control: 2 gelatin capsules/d placebo (liquid paraffin with fish flavour)                    Compliance: tested by questionnaire, response rate was on average 96%. Out of this, 86% complied Duration of intervention: 4 years
Outcomes	Main study outcome: composite of myocardial infarction, cerebral vascular ischemic accident or cardiovascular deaths Dropouts: control: 145 (66 withdrew, 11 lost to follow‐up, 68 deaths), intervention: 134 (61 withdrew, 7 lost to follow‐up, 66 deaths) Available outcomes: deaths, cardiovascular death, non fatal MI, stroke, CV events, coronary events, cancer events, Geriatric Depression Scale score, authors provided additional information on outcomes and methodology Response to contact: yes (data provided)
Notes	The other factorial intervention was B‐vitamins (560 µg methyl‐terahydrofolate, 3 mg B‐6, 20 µg B12) vs placebo Study funding: French Ministry of Research, Ministry of Health, Sodexo, Candia, Unilever, Danone, Roche, Merck
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Used computerized block randomisation with stratification by sex, age, prior CVD, and city of residence". "Permuted block randomisation (with a block size randomly selected as 8) was used".
Allocation concealment (selection bias)	Low risk	Allocation of participants was programmed by the statistical coordinating centre, who sent participants sufficient treatment capsules for 1 year in an appropriately labelled package
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "All subjects and investigators were blinded to treatment allocation", and placebo capsules looked and tasted "identical to the active supplementation". Fish oil flavour was used in placebos.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcome investigators were blinded to allocation
Incomplete outcome data (attrition bias) All outcomes	Low risk	Attritions and exclusions were well described. Only 10% loss over 4 years, well balanced
Selective reporting (reporting bias)	Low risk	ISRCTN41926726 registered 2005, 2003 publication on background and rationale, recruitment started April 2003, 2008 protocol, recruitment ended June 2009, 2010 results published. Outcomes in registry entry appear to have been published.
Attention	Low risk	Not likely as capsules used
Compliance	Low risk	Quote: "Allocation to omega 3 fatty acids increased plasma concentrations of omega 3 fatty acids by 37% compared with placebo" (appears statistically significantly different, though not explicitly stated) … "The overall response rate for return of completed questionnaires was 99%, 96%, 94%, and 95% at 6, 12, and 24 months and at the end of the trial, respectively. About 86% of those who returned a questionnaire reported that they were compliant with the study treatment and compliance was similar in all four groups"
Other bias	Low risk	No further bias noted