Supervised exercise in pregnancy cut gestational diabetes risk by a quarter across 19 trials

Pregnancy is one of the last places where “take it easy” survives as medical advice, and it has been wrong for about two decades.

What has been missing is a clean answer on how much difference structured exercise makes to the things that actually go wrong. Writing in the British Journal of Sports Medicine, researchers pooled 19 studies covering 6,213 pregnancies and found four outcomes moved at once. Supervised exercise reduced the risk of GDM, gestational diabetes, by around a quarter.

Gestational hypertension fell further, and it fell with almost no disagreement between trials.

Four outcomes, and the one to watch

The headline is gestational diabetes, down about a quarter. The most striking number is the one next to it.

The same programs reduced the risk of gestational hypertension by 45%, with an I-squared of 1%. That second figure is the interesting part: I-squared measures how much the trials disagreed, and 1% means they essentially did not. Nineteen trials run in different countries by different teams, all landing in the same place, is a much stronger signal than a bigger effect with wide scatter.

The other two outcomes are downstream of the first two. Excessive gestational weight gain fell to 0.75, and macrosomia, a baby born large enough to complicate delivery, to 0.61. A chain where the metabolic outcomes, the vascular outcome and the birth outcome all move together is coherent in a way that a single isolated finding is not.

The dose, which is the useful part

Most exercise-in-pregnancy advice stops at “stay active”. This review gets more specific.

Subgroup analyses suggested greater effects with earlier intervention, initiated in early pregnancy or lasting more than 20 weeks; higher adherence, at 80% or above; combined aerobic and resistance training; and weekly exercise duration of 150 minutes or more.

Two of those four are about turning up rather than about exercise. Adherence above 80% and a duration past twenty weeks describe a person who kept going, and the review found significant subgroup differences in gestational diabetes risk for intervention initiation timing and for intervention duration. Starting early and continuing appears to matter as much as what the sessions contained.

The 150-minute figure also aligns with general guidance. WHO recommends that adults do at least 150 minutes of moderate-intensity physical activity per week, and this analysis suggests pregnancy is not an exception to that number so much as a period when hitting it pays a specific dividend.

Why supervised is in the title

Every trial pooled here delivered supervised exercise, and that word constrains what the finding supports.

A supervised program means someone scheduled the sessions, someone was present, and attendance was recorded. That is a different intervention from a leaflet recommending activity, and the adherence subgroup finding suggests the difference is not cosmetic: the benefit concentrated where people actually completed the program.

For a health service, that is the expensive reading. For an individual, it points at whatever supervision is available, which in practice is an antenatal exercise class rather than a resolution.

What moderate certainty means here

The authors grade the evidence moderate certainty, which in this literature counts as good.

Two things hold it back from higher. Exercise trials cannot blind participants; everyone knows which group they are in. That matters less for gestational diabetes, which is diagnosed by a glucose test rather than reported, and more for anything subjective, which is part of why the objective outcomes here are the ones worth citing.

The second is scatter. Exercise reduced the risk of excessive gestational weight gain as well, but the trials disagreed far more about that outcome than about the others, with I-squared at 70%. The gestational diabetes and hypertension results are the more solid pair.

What to do about it

Gestational diabetes is a type of diabetes that develops during pregnancy in women who don’t already have diabetes, and it usually resolves after the birth. Usually, but the mother’s later risk of type 2 diabetes stays elevated, which is why preventing it has value beyond the pregnancy itself.

The practical shape of this evidence is: start early rather than at thirty weeks, aim for around 150 minutes a week, mix cardio with resistance work rather than doing only one, and find something with a schedule and other people in it.

The condition on all of that is a conversation first. Most uncomplicated pregnancies can exercise safely and guidelines encourage it, but bleeding, placental and cervical complications change the answer entirely, and a meta-analysis knows nothing about any individual pregnancy. That check belongs before the first session, not after.