Adriëtte Oostvogels

82 Chapter 3 In order to conclude that maternal lipids were intermediates in the association between maternal pBMI and child’s adiposity we first conducted a mediation analyses according to the causal-steps analysis. 25 According to the causal-steps analyses all three associations (a–c) should be statistically significant after adjustments, and when analysing the association between pBMI and adiposity of the child (c), the addition of the maternal lipid variables to the full adjusted models (Model 2) should attenuate the association into a statistically non-significant association between pBMI and child’s adiposity. 25 In addition we also assessed whether maternal lipids were mediators between maternal pBMI and child’s adiposity by testing the significance of the indirect or mediated effect computed as the product of regression coefficient estimates using a bootstrapping approach with a multiple mediator model (Figure 2). 26,27 All the maternal lipid variables were included in this model. The resampling rate was set at 10.000. Figure 2. Indirect effect of maternal prepregnancy BMI on adiposity of the child through the maternal lipid profile.