Adriëtte Oostvogels

9 Postnatal growth In this thesis, postnatal growth was studied as an underlying mechanism in the association between pBMI and cardiometabolic profile in childhood. Growth can be measured and defined in many different ways. The way in which growth was investigated in this thesis offers several advantages. First, when assessing the mediating effect of postnatal growth in the association between pBMI and offspring’s cardiometabolic profile in childhood, not only early infant weight gain but also weight-for-length gain (which reflects adiposity better than weight gain alone) were included in the analyses. 10 Second, studies on growth have shown that both relative weight gain and linear growth are independent from BMI, and are positively associated with body composition and systolic blood pressure in children and adults. 11-13 Therefore, when separately investigating the effect of maternal pBMI on postnatal growth up to 7 years, the weight, height and BMI were evaluated. Third, to compare the influence of maternal weight status on these three growth measures, standard deviation scores (SDS) were computed using population specific reference curves. The benefits of using these curves over other reference curves (e.g. the WHO growth standard) is that the distribution of the SDS is better and comparison between the SDS for weight, height and BMI are more reliable because Dutch children are (on average) taller than the children on which the WHO growth standard is based. In addition, some limitations need to be addressed. First, because birth length was not available in the ABCD-study, in our studies on growth in length, the measurements did not start until one month after birth. Although it was found that maternal pBMI is associated with cardiometabolic health independently of growth, we cannot conclude that this is due to an intra-uterine effect as no information was available on growth in length in the first month after birth. Second, our decision to focus the mediation analyses on the time period from 1-3 months was based on previous studies; 14 however, choice for another (or longer) growth period might have led to different conclusions. Third, although in one of the studies growth was studied as the difference in weight- or weight-for-length gain between two time points, in all other studies in this thesis, growth was defined as the weight, height or BMI at a certain time point, which is a relatively crude measure of the complex biological process of growth per se. However, the actual sizes will be instrumental in translating the findings to practice, where clinicians are also unable to map the full process of growth, but are able to measure size. Fourth, when assessing growth to overweight, only BMI growth pattern was included, whereas growth patterns of weight and height could have shown whether a higher BMI was the result of faster growth in weight, or slower growth in height. 225 General discussion