- Design
- Systematic review and random-effects meta-analysis of randomised trials
- Population
- 29 RCTs, 17,937 pregnant women
- Primary outcome
- Incidence of gestational diabetes
- Effect
- RR 0.88 (95% CI 0.79 to 0.98); 16.01% vs 17.06%; I² 51.1%
A systematic review and meta-analysis pooled 29 randomised trials of prenatal nutrition education involving 17,937 pregnant women, searching to January 2026 and assessing bias with the Cochrane RoB 2 tool.
Education programmes were associated with a lower incidence of gestational diabetes: pooled risk ratio 0.88 (95% CI 0.79 to 0.98). In absolute terms, incidence was 16.0% with the intervention and 17.1% without, about one case prevented for every 95 women taught. Heterogeneity was moderate (I² 51%), driven by differences in dietary content and how faithfully programmes were delivered.
The largest reductions came from programmes built on nutritional theory, delivered in hospital settings and run with high fidelity. Delivery by health professionals and individualised plans trended in the same direction but did not reach significance.
For Indian practice, where gestational diabetes is common and many women are seen first in busy public clinics, the message is that a structured, consistently delivered programme is what helped, not a single leaflet. The absolute benefit is small, so this is a reason to organise dietary teaching well rather than to expect it to replace screening.
- Offer structured dietary teaching from booking, not only after a gestational diabetes diagnosis.
- Base the content on a defined dietary approach rather than general advice to eat healthily.
- Deliver it consistently: the programmes that kept to their plan showed the largest effect.
- Continue universal or risk-based screening; education lowers the risk but does not remove it.
- Plan for about 1 case prevented per 95 women taught when weighing staff time.
Why it matters
It turns dietary advice from a courtesy into an intervention with a measurable, if modest, effect on a common complication.
Don't overread it
The absolute reduction is about one percentage point, and the findings on which programme features work best come from subgroup comparisons.
The statistics, in plain English
A risk ratio of 0.88 means about 12% fewer cases in relative terms; the interval from 0.79 to 0.98 stays below 1, so the benefit is unlikely to be chance, but its upper end is close to no effect. An I² of 51% means the trials disagreed moderately, so the size of benefit in any one clinic may differ. The component analyses compare subgroups of trials and are weaker evidence than the headline result.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for obstetrics & gynaecology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free