Doubling the aspirin dose did not prevent more preeclampsia in a 365-woman trial

Low-dose aspirin in pregnancy is one of the few genuinely preventive interventions in obstetrics. What "low dose" should mean has been contested for years, because the trial evidence supporting it was not generated at the dose most widely used in the United States. A randomised trial published online in Obstetrics & Gynecology on October 24 tested the question head-on and did not find what many expected [s1].

Why the dose question exists

The strongest evidence for aspirin comes from a 2018 systematic review and meta-analysis of 16 trials covering 18,907 participants [s2]. Aspirin was associated with a reduction in preterm preeclampsia (relative risk 0.62, 95% CI 0.45-0.87), with no significant effect on term preeclampsia (0.92, 0.70-1.21) [s2].

Crucially, the reduction in preterm preeclampsia was confined to the subgroup in which aspirin was started at or before 16 weeks of gestation at a daily dose of 100 mg or more, where the relative risk was 0.33 [s2]. Heterogeneity across studies was high overall, but zero within that subgroup [s2].

The comparator arm in the new trial, 81 mg daily [s1], sits below the 100 mg threshold at which that meta-analysis found the effect [s2]. That is the gap the trial was designed to close.

The trial

This was a pragmatic, randomised, open-label, blinded-endpoint trial [s1]. Pregnant people at high risk of preeclampsia were randomised to 162 mg or 81 mg of aspirin daily, beginning before 16 weeks of gestation and continuing until term, with follow-up to six weeks postpartum [s1].

The primary composite outcome was preterm preeclampsia — before 37 weeks — or preeclampsia with severe features, adjudicated by independent researchers blinded to treatment group [s1]. Secondary outcomes were the components of the composite, adherence, and maternal and neonatal complications [s1].

The power calculation assumed an 8.6% incidence of the primary composite in the 81 mg group, and estimated that 394 participants — 197 per group — would give 80% power to detect a 7.1% absolute reduction with 162 mg, at a two-sided alpha of 0.05 [s1].

Of 400 randomised, 365 had delivery data and were included in the intention-to-treat analysis: 184 on 162 mg and 181 on 81 mg [s1].

The result

The primary composite occurred in 26 of 184 (14.1%) on 162 mg and 31 of 181 (17.1%) on 81 mg — relative risk 0.83 (95% CI 0.51-1.33), p = 0.4 [s1].

The individual components, preterm preeclampsia and term preeclampsia with severe features, were similar between groups [s1]. Adherence was high and comparable: 88% to 91% in the 162 mg group and 89% to 92% in the 81 mg group across study visits [s1].

The authors conclude that among people at increased risk of preeclampsia, rates of preterm preeclampsia or preeclampsia with severe features were similar with either dose started before 16 weeks [s1].

Two things the trial found that were not the primary question

Singleton birth weight was slightly lower in the 162 mg group: 2.9 kg against 3.2 kg (p = 0.005) [s1].

And there were eight cases of placental abruption in the 162 mg group against none in the 81 mg group (p = 0.013) [s1].

That second finding requires care in both directions. Eight events against zero in a trial of 365 people is a small number, and secondary safety outcomes in a trial powered for something else are exactly where chance findings appear. The comparison was not pre-specified as a primary safety endpoint, and no multiplicity adjustment is described.

But abruption is a plausible harm for an antiplatelet agent at higher dose, the p value is not marginal, and zero events in the comparator group is a stark contrast. A signal that is biologically coherent and statistically nominal is the kind that gets carried into larger datasets rather than dismissed.

The power problem underneath the null

The trial assumed an 8.6% event rate in the 81 mg group [s1]. The observed rate was 17.1% [s1] — roughly double. It was powered to detect a 7.1 percentage point absolute reduction [s1], a large effect.

The observed difference was 3.0 percentage points, with a confidence interval on the relative risk running from 0.51 to 1.33 [s1]. That interval is compatible with a 49% reduction and with a 33% increase [s1]. This is a trial that did not detect a difference, not a trial that established the doses are equivalent — a distinction that matters more here than usual, because the higher dose carried the abruption signal.

What to watch

Whether larger trials or pooled analyses of aspirin dose report the abruption finding; whether guideline bodies revisit the 81 mg versus 100 mg-plus discrepancy in light of a trial that tested it directly; and whether the birth-weight difference appears elsewhere.

This article describes a single randomised trial and is informational only. It is not medical advice, and nothing here should be used to start, stop or change any medication in pregnancy.

Sources

• [s1] Comparison of 162 mg and 81 mg Aspirin for Prevention of Preeclampsia: A Randomized Controlled Trial, Obstetrics & Gynecology, 2026;147(1):87-96, published online 2025-10-24. ClinicalTrials.gov NCT04070573.
• [s2] Aspirin for the prevention of preterm and term preeclampsia: systematic review and metaanalysis, American Journal of Obstetrics and Gynecology, 2018;218(3):287-293.e1, published online 2017-11-11.