Weight gain outside pregnancy guidelines tracked with more severe maternal morbidity

Gestational weight gain is one of the most routinely tracked numbers in antenatal care, and one of the most contested. The ranges most clinics still use come from a 2009 Institute of Medicine report that set recommended gain by prepregnancy body mass index [s2]. What falling outside those ranges means for the mother's own risk of a life-threatening delivery complication has been harder to establish. A retrospective cohort study published in JAMA Network Open on October 22 takes that question to more than a million pregnancies [s1].

What the study looked at

Researchers linked birth registry records from Better Outcomes Registry & Network Ontario to hospitalisation data from the Canadian Institute for Health Information, covering singleton pregnancies delivered at 20 weeks or later in Ontario, Canada, from April 2012 to March 2022 [s1]. The analysis ran from May to November 2025 [s1].

The exposure was gestational weight gain, measured both continuously — as mean weekly gain across the second and third trimesters — and categorically as inadequate, adequate or excessive, judged against the 2009 Institute of Medicine thresholds for each prepregnancy BMI band [s1]. The outcome was severe maternal morbidity, a composite of 13 indicators defined by the Canadian Perinatal Surveillance System and recorded during the delivery hospitalisation [s1].

The numbers

Among 1,035,927 pregnancies, 192,731 (18.6%) had inadequate weight gain, 296,448 (28.6%) had adequate gain, and 546,748 (52.8%) had excessive gain [s1]. In other words, more than half the pregnancies in this cohort gained more than the guidelines recommend. Severe maternal morbidity occurred in 15,119 deliveries, a rate of 14.59 per 1,000 births [s1].

Compared with adequate weight gain, excessive gain was associated with higher severe maternal morbidity in every prepregnancy BMI category [s1]. The adjusted rate ratios were 1.33 for people who were underweight (95% CI, 1.15-1.53), 1.24 for normal weight (1.17-1.31), 1.16 for overweight (1.06-1.27), 1.21 for obesity class I (1.08-1.35), 1.28 for obesity class II (1.07-1.54) and 1.24 for obesity class III (1.03-1.48) [s1]. None of those confidence intervals crossed 1, so the association held across the weight spectrum.

Inadequate weight gain was a less consistent signal. It was associated with increased severe maternal morbidity mainly among people of normal weight (adjusted rate ratio 1.13; 95% CI, 1.06-1.21) and overweight (1.19; 1.05-1.35) [s1].

A U-shape, not a straight line

The continuous analysis is where the picture becomes more nuanced. Across all BMI categories the association between weight gain and severe maternal morbidity was nonlinear [s1]. For people who were underweight, normal weight or overweight, risk was elevated at both low and high weekly gain — a U-shaped relationship [s1]. For people in obesity classes I through III, the risk was concentrated at the high end only [s1].

The estimated weekly gain associated with the lowest severe maternal morbidity also shifted with BMI: roughly 0.4 kg per week for the underweight-through-overweight categories, declining across the obesity classes to about 0.1 kg per week for obesity class III [s1]. That gradient broadly tracks the direction of the existing Institute of Medicine ranges, which recommend less total gain at higher prepregnancy weight [s2].

How much weight to put on it

This is an association, not a demonstration of cause. A retrospective cohort cannot rule out that some third factor — an underlying condition, say — drives both unusual weight gain and a complicated delivery. The authors adjusted for measured confounders, but residual confounding is always possible in a design like this, and the analysis cannot say that changing a person's weight gain would change their risk.

The scale is the study's main strength. Over a million linked pregnancies, with morbidity captured from hospital records rather than self-report, gives the estimates a precision that smaller studies cannot match, and the confidence intervals are correspondingly tight. The consistency of the excessive-gain signal across every BMI band is harder to explain away than a single subgroup finding would be.

What it does not do is rewrite any guideline on its own. The authors frame the work as complementing, not replacing, the evidence behind current weight-gain advice [s1]. The thresholds themselves were not tested here; what was tested is whether gain outside them travels with a worse delivery, and in this population it did.

What to watch

Whether the U-shaped relationship replicates in other health systems; whether the raised risk at low gain in normal-weight pregnancies prompts a closer look at the "inadequate" end of the ranges; and whether any of this feeds into the next revision of weight-gain guidance.

This article describes a single observational study and is informational only. It is not medical advice, and nothing here should be used to guide weight gain in any individual pregnancy.