Predictors of compliance with higher dose omega-3 fatty acid supplementation during pregnancy and implications for the risk of prematurity: exploratory analysis of the ORIP randomised trial

Background Intention-to-treat analyses of the Omega-3 to Reduce the Incidence of Prematurity (ORIP) trial found that omega-3 (n-3) fatty acid supplementation reduces the risk of prematurity in the subgroup of women with a singleton pregnancy and low n-3 status early in pregnancy, but not overall. However, results may have been influenced by less-than-optimal compliance. Objectives To identify predictors of compliance with n-3 supplementation and determine treatment effects among compliers. Design Exploratory analyses of a multicentre-blinded randomised trial. Setting 6 tertiary care centres in Australia. Participants 5328 singleton pregnancies. Interventions Daily capsules containing 900 mg n-3 long-chain polyunsaturated fatty acids or vegetable oil, consumed from before 20 weeks gestation until 34 weeks gestation. Outcome measures Early preterm (<34 weeks gestation) and preterm birth (<37 weeks gestation). Women were considered compliant if they reported missing less than a third of their allocated capsules in the previous week during a mid-pregnancy appointment. Results Among 2654 singleton pregnancies in the n-3 intervention group, 1727 (65%) were deemed compliant with supplementation. Maternal characteristics associated with compliance included age, years of full-time education, consuming alcohol but not smoking in the 3 months leading up to pregnancy, fewer previous births and taking dietary supplements at enrolment. Based on complier average causal effects, n-3 supplementation reduced the risk of preterm birth in compliers (relative risk=0.76; 95% CI 0.60 to 0.97), but not early preterm birth (relative risk=0.80; 95% CI 0.44 to 1.46). Consistent with intention-to-treat analyses, the lack of an overall effect on early preterm birth in compliers appeared to be due to beneficial effects in women with low n-3 status at enrolment but not women with replete status. Conclusions Results in compliers were similar to those from intention-to-treat analyses, suggesting that non-compliance was not a major factor in explaining outcomes from the ORIP trial. Trial registration number ACTRN12613001142729.