Shinto 2014

Methods	RCT, parallel (fish oil capsule vs soybean oil capsule), 12 months Summary risk of bias: moderate to high
Participants	Patients aged 55 or more with probable Alzheimer dementia diagnosis N: 13 intervention, 13 control Level of risk for CVD: low Men: 61% intervention 46% control Mean age in years (SD): 75.9 (8.1) intervention, 75.2 (10.8) control Age range: 55+ (inclusion criteria) Smokers: not reported Hypertension: not reported Medications taken by at least 50% of those in the control group: anti‐cholinesterases or memantine 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 Lipid‐lowering medications and many other drugs were not allowed Location: USA Ethnicity: 100% white
Interventions	Type: fish oil capsules Comparison: EPA + DHA vs n‐6 Intervention: 3 × 1 g capsules/day of fish oils (975 mg EPA, 675 mg DHA per day). Dose: 1.65 g/d EPA + DHA Control: 3 × 1 g capsules/day soybean oil (which contains 5% fish oil) Both groups had a placebo lipoic acid tablet and lemon‐flavoured capsules Compliance: assessed by pill counts and FA in red blood cell membranes. Results showed increased EPA + DHA levels in the intervention group. Length of intervention: 12 months
Outcomes	Main study outcome: F2‐isoprostane levels (oxidative stress measure) Dropouts: 2 intervention, 2 control Available outcomes: mortality, CVD events, adverse events, serum fatty acids, measures of cognition (ADAS Cog and MMSE), ADL, IADL (also F2 isoprostane) Response to contact: not attempted
Notes	Study funding: National Institutes of Health/National Institute of Aging (NIH/NIA) and NIH General Clinical Research
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were randomised by a computer‐generated scheme that was stratified by smoking status
Allocation concealment (selection bias)	Unclear risk	No details
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Capsules matched for taste and flavour. Blinding assessed at the end and majority of staff and participants were unaware of treatment
Blinding of outcome assessment (detection bias) All outcomes	Low risk	As above
Incomplete outcome data (attrition bias) All outcomes	Low risk	15% dropouts explained and included
Selective reporting (reporting bias)	Low risk	NCT00090402 first received: 25 August 2004, study start date April 2004. More secondary outcomes reported than included in the trial register entry
Attention	Low risk	Both arms seem to have had the same contact
Compliance	Low risk	Compliance measured and FAs levels reported. Results showed increased EPA + DHA levels in the intervention group
Other bias	Low risk	None noted