Khalili 2016.

Methods	RCT: IRCT2013100914957N1
Participants	150 women randomised Inclusion criteria: women aged 18‐35 years; 1st to 5th pregnancy, with a household health record in the participating health centres, ability to read and write, singleton pregnancy, stable phone access Exclusion criteria: bleeding during pregnancy, placenta praevia, abruption or cerclage in the present pregnancy, history of allergy to fish oil or other fish products, allergy to gelatin, history of previous underlying diseases such as heart disease, kidney disease, hyperlipidaemia, taking medication for 1 of these, consuming fish more than twice a week, smoking or drug addiction, bleeding disorders or taking anticoagulants, BMI > 30, participating in another interventional study Characteristics: very low baseline omega‐3 concentrations Setting: 27 health care centres, Tabriz, Iran
Interventions	SUPPLEMENTATION: DHA + EPA versus placebo Group 1: 1000 mg fish oil supplements (120 mg DHA and 180 mg EPA). Total number randomised: n = 75 Group 2: placebo (similar shape, size and weight); liquid paraffin. Total number randomised: n = 75 Timing of supplementation: from 20 weeks GA to 30 days after birth (women were recruited between 16‐20 weeks GA) DHA + EPA dose/day: low: 120 mg DHA + 180 mg EPA
Outcomes	Women/birth: maternal serum fatty acid profiles at 35‐37 weeks GA; adherence; adverse events (nausea, unpleasant taste, vomiting diarrhoea, stomach pain); caesarean section Babies/infants/children: neonatal death; perinatal death; birthweight; low birthweight; birth length; head circumference at birth; growth at 4 and 6 months; ASQ (2nd edition) at 4 and 6 months (scores and thresholds)
Notes	Funding: Tabriz University of Medical Sciences. follow‐up study supported by Tabriz Health Services Management Research Centre (Grant No. 5/77/5241) Declarations of interest: none declared
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computerised random number table generator
Allocation concealment (selection bias)	Low risk	Quote: “randomly assigned … by a staff member not involved in the research”
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Placebo controlled (although smell or taste of paraffin may have been able to be detected)
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Not reported, but probably done.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	8/75 women in the omega‐3 group and 7/75 did not have blood samples at 35‐37 weeks GA (omega‐3 group: 5 not interested in sampling; 2 preterm labour: placebo group: 6 not interested in sampling; 2 preterm labour); 1 women in the placebo group stopped capsules due to nausea. For other health outcomes, 0/75 in the omega‐3 group and 4/75 in the placebo group were lost to follow‐up.
Selective reporting (reporting bias)	Low risk	No apparent evidence of selective reporting (though stillbirth not explicitly reported).
Other bias	Low risk	Baseline characteristics appeared similar.