Design: Systematic review and meta-analysis of 19 randomised trials
Population: 6,213 pregnant women
Primary outcome: Gestational diabetes
Effect: RR 0.74 (95% CI 0.57 to 0.96); gestational hypertension RR 0.55 (0.40 to 0.77)

A systematic review pooled 19 randomised trials of supervised exercise against usual care in 6,213 pregnant women, with gestational diabetes as the prespecified primary outcome. Certainty of evidence was rated moderate.

Supervised exercise reduced gestational diabetes (RR 0.74, 95% CI 0.57 to 0.96), gestational hypertension (RR 0.55, 0.40 to 0.77), excessive gestational weight gain (RR 0.75, 0.61 to 0.92) and macrosomia (RR 0.61, 0.46 to 0.81). Effects were larger when exercise started early in pregnancy or ran for more than 20 weeks, when adherence was at least 80%, when aerobic and resistance training were combined, and at 150 minutes a week or more.

The practical point is that the benefit came from supervised programmes, not advice alone. Advice to 'stay active' is where most antenatal care stops. Where a structured group or physiotherapist-led class exists, referral at booking is likely to do more than a leaflet.

Ask about current activity at the booking visit, not at the glucose tolerance test.
Aim for 150 minutes a week, combining aerobic and resistance work.
Refer to a supervised class where one exists; start early rather than after 20 weeks.
Screen for contraindications: placenta praevia after 26 weeks, cervical insufficiency, significant heart or lung disease.

Why it matters

The benefit depends on starting early and supervising it, which moves exercise from a counselling line into a referral.

The statistics, in plain English

A risk ratio of 0.74 means about a quarter fewer cases of gestational diabetes. The upper limit of 0.96 is close to 1, so the true effect could be small. Heterogeneity for diabetes was moderate (I² 47%), meaning trials disagreed somewhat on size. The timing and dose findings come from subgroup comparisons, which are weaker than the main result.