Randomised clinical trials of fish oil supplementation in high risk pregnancies

Objective To test the postulated preventive effects of dietary n-3 fatty ac ids on pre-term delivery, intrauterine growth retardation, and pregnancy in duced hypertension. Design In six multicentre trials, women with high risk pregnancies were ran domly assigned to receive fish oil (Pikasol) or olive oil in identically-lo oking capsules from around 20 weeks (prophylactic trials) or 33 weeks (ther apeutic trials) until delivery. Setting Nineteen hospitals in Europe. Samples Four prophylactic trials enrolled 232, 280, and 386 women who had e xperienced previous preterm delivery, intrauterine growth retardation, or p regnancy induced hypertension respectively and 579 with twin pregnancies. T wo therapeutic trials enrolled 79 women with threatening pre-eclampsia and 63 with suspected intrauterine growth retardation. Interventions The fish oil provided 2.7 g and 6.1 g n-3 fatty acids/day in the prophylactic and therapeutic trials, respectively. Main outcome measures Preterm delivery, intrauterine growth retardation, pr egnancy induced hypertension. Results Fish oil reduced recurrence risk of pre-term delivery from 33% to 2 1% (odds ratio 0.54 (95% CI 0.30 to 0.98)) but did not affect recurrence ri sks for the other outcomes (OR 1.26; 0.74 to 2.12 and 0.98; 0.63 to 1.53, r espectively). In twin pregnancies, the risks for all three outcomes were si milar in the two intervention arms (95% CI for the three odds ratios were 0 .73 to 1.40, 0.90 to 1.52, and 0.83 to 2.32, respectively). The therapeutic trials detected no significant effects on pre-defined outcomes. In the com bined trials, fish oil delayed spontaneous delivery (proportional hazards r atio 1.22; 1.07 to 1.39, P = 0.002). Conclusions Fish oil supplementation reduced the recurrence risk of pre-ter m delivery, but had no effect on pre-term delivery in twin pregnancies. Fis h oil had no effect on intrauterine growth retardation and pregnancy induce d hypertension, affecting neither recurrence risk nor risk in twin pregnanc ies.