Skip to content

Higher dose docosahexaenoic acid supplementation during pregnancy and early preterm birth: A randomised, double-blind, adaptive-design superiority trial

EClinicalMedicine
Q1
May 2021
Citations:62
Influential Citations:3
Interventional (Human) Studies
90
S2 IconPDF Icon

Enhanced Details

Methods
Randomized, double-blind, adaptive-design superiority trial in pregnant people with singleton pregnancies enrolled at 12 to 20 weeks' gestation at three large U.S. academic medical centers. Active intervention arms randomized 576 participants to 200 mg/day DHA and 524 to 1000 mg/day DHA; the primary analyses included 492 and 540 participants, respectively.
Intervention
Oral docosahexaenoic acid (DHA) capsules were taken daily throughout pregnancy. All participants received 200 mg DHA per day from one capsule, then were assigned to two additional daily capsules containing either corn and soybean oil for a total of 200 mg/day DHA or an algal DHA source for a total of 1000 mg/day DHA.
Results
The 1000 mg/day DHA regimen was associated with a lower rate of early preterm birth than 200 mg/day, with the clearest benefit in participants with low baseline DHA status. Early preterm birth occurred in 9/540 (1.7%) in the 1000 mg group versus 12/492 (2.4%) in the 200 mg group, and the Bayesian posterior probability favored 1000 mg/day (0.81). Among participants with low DHA status (<6%), early preterm birth was 5/249 (2.0%) with 1000 mg versus 9/219 (4.1%) with 200 mg, with posterior probability 0.93. Among participants with high DHA status (>=6%), rates were similar: 4/289 (1.4%) versus 3/271 (1.1%), posterior probability 0.57. Higher-dose DHA also favored fewer maternal and neonatal adverse and serious adverse events, and maternal postpartum and cord blood DHA levels were higher with 1000 mg/day.
Limitations
The primary outcome was uncommon, so event counts were small and some estimates are imprecise. The adaptive Bayesian design and subgroup findings may be less familiar for clinical interpretation, and several secondary outcomes were null or only reported in detail in tables and supplements. Generalizability is limited to pregnant people who could read or speak English or Spanish and were recruited from three U.S. academic centers.

Abstract

No abstract available