Introduction Basic science evidence suggests that gestational weight gain (GWG) may influence offspring cognitive development. However, this relationship has not been investigated in human population studies.

Methods Data from the ALSPAC, a UK prospective pregnancy cohort were used. GWG was expressed using 2009 IOM categories of GWG and estimates from random effect linear spline models (median number of measures per woman: 10 IQR: 8–11). Outcomes were school entry assessment score (SEA, age 4, N=5832), Wisc-III assessed IQ (age 8, N=5191) and GCSE results (age 15, N=7339).

Results Offspring of women who gained less than the 2009 IOM recommended GWG had a -0.075SD lower mean SEA score (95% CI −0.127 to −0.023) compared with women who gained as recommended, even when adjusting for potential confounders including maternal education. Greater prepregnancy weight was inversely associated with all cognition measures. For example, mean difference in IQ per 1 kg increase in pre-pregnancy weight =−0.004SD (−0.006, −0.002). GWG in early pregnancy (0–18 weeks) and in mid-pregnancy (19–28 weeks) were positively associated with SEA and IQ but not with GCSE results. GWG in late pregnancy (29+ weeks) was positively associated with higher SEA scores (0.208SD; 0.716, 0.261) and GCSE results (OR=1.35; 1.26, 1.46), with the latter not fully mediated by the association with SEA.

Conclusions Findings support a positive association between GWG, particularly in late gestation, and offspring cognitive development, which has lasting effects on school attainment at age 16 years. However this could still be due to residuals confounding.