Razavi 2017.

Methods	RCT (4 arms, see below)
Participants	120 women randomised Inclusion criteria: 18‐40 years; without prior diabetes; diagnosed with GDM at 24‐28 weeks’ gestation Exclusion criteria: taking omega‐3 fatty acid supplements; insulin therapy; placental abruption; PE; eclampsia; hypo‐ and hyperthyroidism; smokers Setting: Ardabil, Iran (conducted from September 2016 to March 2017)
Interventions	SUPPLEMENTATION: omega‐3 versus omega‐3 + vitamin D versus vitamin D versus placebo Group 1: omega‐3 fatty acids (2000 mg size of capsules containing 360 mg EPA and 240 mg DHA per day) as 2 capsules plus placebo vitamin D capsules; n = 30 randomised (all completed) Group 2: omega‐3 fatty acids (2000 mg size of capsules containing 360 mg EPA and 240 mg DHA per day) as 2 capsules, plus vitamin D (50 000 IU every 2 weeks): n = 30 randomised (all completed) Group 3: vitamin D (50 000 IU every 2 weeks) plus placebo omega‐3 capsules: n = 30 randomised (all completed) Group 4: no supplement (2 placebo capsules, each containing 500 mg of liquid paraffin): n = 30 randomised (all completed) Timing of supplementation: from 24‐28 weeks' gestation (for 6 weeks) All women: requested not to change their routine physical activity or usual dietary intakes throughout study and not to consume any supplements other than the one provided to them by the investigators, as well as not to take any medications that might affect findings during the 6‐week intervention DHA + EPA dose/day: mid: 240 mg DHA + 360 mg EPA
Outcomes	Women/birth: inflammatory biomarkers; biomarkers of oxidative stress; caesarean section; preterm birth < 37 weeks; polyhydramnios; PE; length of gestation Babies/infants/children: macrosomia (> 4000 g); birthweight; length at birth; head circumference at birth; Apgar score; newborn hyperbilirubinaemia; newborn hospitalisation; newborn hypoglycaemia
Notes	Funding: research grant from Research Deputy of Ardabil University of Medical Sciences (AUMS) Declarations of interest: none declared
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Assignment was conducted using computer‐generated random numbers.
Allocation concealment (selection bias)	Low risk	Computer based randomisation; and "a person who was not involved in the trial and not aware of random sequences, assigned the subjects to the numbered bottles of capsules"
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Randomisation and allocation were concealed from the researchers and participants until the final analyses were completed"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Not reported; however outcomes probably assessed by the researchers (who were blind to group assignments), and all outcomes included for analysis objective.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All 120 participants randomised included for analysis.
Selective reporting (reporting bias)	Unclear risk	Outcomes specified in published protocol reported; no obvious sign of selective reporting, however limited range of outcomes for babies/infants/children reported.
Other bias	Low risk	No clear baseline differences between the 4 groups of participants.