Design: Retrospective population-based cohort, US vital statistics 2022–2023
Population: 6,923,146 singleton births at 20–43 weeks
Primary outcome: Gestational age-specific stillbirth
Effect: At 31 weeks, BMI 40 vs 20: aHR 1.22 (1.13–1.33) without diabetes; 0.68 (0.54–0.85) with diabetes

This BJOG cohort, published in April, used US birth and fetal death certificates from 2022 and 2023 for about 6.9 million singleton births. Stillbirth rates rose with BMI and were higher with pre-pregnancy diabetes (16.6 per 1,000 births) than without (4.4 per 1,000).

The week-by-week pattern differed between the two groups. At 31 weeks, for example, a BMI of 40 compared with 20 was associated with higher hazard in women without diabetes (aHR 1.22) but lower relative hazard in women with diabetes (aHR 0.68), because the baseline risk with diabetes is already high. Absolute risk was highest in women with both diabetes and class III obesity.

The authors conclude that the best week for delivery in women with obesity depends on other risk factors such as diabetes. This supports individual timing decisions rather than a single BMI threshold.

Assess stillbirth risk using BMI and diabetes status together, not BMI alone.
Women with pre-pregnancy diabetes and class III obesity carry the highest absolute risk.
Discuss timing of birth individually rather than applying one BMI cut-off.
Record pre-pregnancy BMI and diabetes type at booking so the combined risk is visible.

Why it matters

A relative risk that falls with BMI in diabetes does not mean lower danger; the baseline is already high.

The statistics, in plain English

A hazard ratio below 1 in the diabetes group compares heavier with lighter women who both have diabetes; it does not mean diabetes plus obesity is safe. Absolute rates, 16.6 versus 4.4 per 1,000, show the diabetes group's much higher starting risk.