Adriëtte Oostvogels

Sensitivity analyses Sensitivity analyses in the Dutch origin population only (N=2324) yielded similar ΔSDS for all growth measures of girls of mothers with overweight or obesity and of boys of mothers with overweight compared to the complete study population (supplementary table 3 and 4 and supplementary figure 1). Similar ΔSDS were also found for height of boys of mothers with obesity, but higher ΔSDS were found for their weight and BMI. Discussion Sex-specific growth patterns of offspring of overweight and obese mothers before pregnancy were compared with those of offspring of normal weight mothers before pregnancy. We found that growth patterns of both boys and girls from birth up to 7 years differed per maternal pBMI category: boys and girls of overweight and obese mothers grew faster in weight, height and BMI and were heavier and taller at age 7 than offspring of normal weight mothers. The effects were more pronounced in girls than in boys. Faster growth in children of obese mothers seemed to result from faster weight gain compared to children of mothers with normal weight and not the result of slower height gain, as also height gain was increased compared to children of normal weight mothers. This study used serial measurements of postnatal growth collected by trained nurses using a standardized protocol. We took information on important covariates affecting fetal growth, like gestational diabetes and hypertension into account. Moreover, our study was large enough to study growth patterns for boys and girls separately, whereas most other studies ignored possible sex differences in the associations. Furthermore, sensitivity analyses in a Dutch origin only subsample yielded similar ΔSDS when comparing children of overweight and obese women with children of normal weight women, showing the robustness of our results. This study also had some limitations. First, maternal pBMI was self-reported in the pregnancy questionnaire, this might have resulted in an underestimation and misclassification in pBMI groups, resulting in less differences in growth patterns between the groups. 26 Therefore, the associations between maternal pBMI and offspring’s growth patterns could in fact be stronger than the associations presented here. Second, we did not have information on paternal pBMI. Research in a comparable cohort showed that the effect and explained variance of maternal and paternal prepregnancy anthropometrics on the child’s anthropometrics were similar, but the combined explained variance was less than the sum of the maternal and paternal explained variance individually. 8 Thus, despite the fact that paternal pBMI is missing, the 158 Chapter 6