In 9,008 pregnancies with gestational diabetes, weight gain ran fastest before the diagnosis, then slowed. In women with obesity it stayed above the national target, and a new local range didn't clearly beat the old one

Gestational diabetes is usually picked up with a glucose test at 24 to 28 weeks. So what has already happened to a woman's weight by then, and what happens next? Researchers at the Maternal and Child Health Hospital of Guangxi, in Nanning, went through the records of 9,008 women with singleton pregnancies complicated by gestational diabetes who gave birth between February 2021 and December 2025. Their median glucose test fell at about 25.7 weeks.

Weight gain in the second trimester, before the test, generally ran at or above the rate China's National Health Commission recommends. Among women of normal weight whose pregnancies went without complications, the median was 0.48 kg a week, against a recommended 0.37. For women with obesity it was 0.35 kg against 0.22. After diagnosis, the pace slowed in every BMI group: to 0.33 kg a week among those women of normal weight and 0.28 kg among those with obesity. That's within the recommended range for normal weight, but still above the 0.12 to 0.23 kg recommended for obesity.

It's tempting to credit the slowdown to diagnosis and diet advice. The authors say you can't: there is no untreated group for comparison, and the records didn't consistently capture medical nutrition therapy, activity, glucose control or treatment. What the numbers do show is timing. In these records, the faster weekly gain came before the diagnosis, not after it.

Total gain mattered in both directions. Too much was linked mainly to preeclampsia, a baby large for dates and caesarean birth, especially among women of normal weight or overweight. Too little was linked with low birth weight and being small for dates; among underweight women, gaining less than the local range described below went with higher odds of preterm birth and low birth weight (adjusted odds ratios 2.82 and 3.03). The preterm links were less solid in a check that looked only at weight gained up to 28 weeks, and for good reason: a pregnancy that ends early has had less time to add kilos. These are associations, not proof of cause, but they're a good reason to distrust the idea that less is always better.

The team also built its own ranges from the middle half of total gain among women with no adverse outcome: 11.0 to 15.0 kg for underweight, 9.2 to 14.0 for normal weight, 7.5 to 13.0 for overweight and 6.0 to 11.5 for obesity. Did they work better than the national guidance? Not clearly. After correction, the two performed broadly alike, with small differences in inconsistent directions. In overweight women, those within the local range actually had more large babies than those within the national one (4.4% against 2.7% for macrosomia).

So there's no new number to tape to the fridge. The ranges were derived and tested in the same hospital's records, pre-pregnancy weight was self-reported, and they need checking in other populations. The more useful message may be about timing. These data can't prove it, but they raise the question of whether conversations about food and weight should start well before the 24-week test, and the authors suggest that the pace of gain may still need attention after diagnosis in women in higher BMI groups. Too little gain is a risk as well, so this is one for a midwife, obstetrician or dietitian, not a do-it-yourself target.