Supplemental Methods: Conditional analysis

The standardized residuals from multiple linear regression analysis are used in conditional modelling to assess the associations of total fat mass, subcutaneous area, preperitoneal area and subcutaneous distance/preperitoneal distance ratio with cardiovascular risk factors, statistically independent from body mass index. Conditional models of total body fat mass measures and abdominal fat mass measures were built using linear regression analysis. In these analyses, body mass index is considered as starting point. The excess total body and abdominal fat mass is similar to the standardized residuals resulting from the linear regression model of total body and abdominal fat mass measures regressed on body mass index. The standard residuals obtained from this regression model are entirely uncorrelated with body mass index (1, 2, 3). This allows inclusion of both body mass index and more detailed fat mass measures in linear regression models, without problems regarding the correlations between these measures. After construction of these new independent variables, we subsequently used linear regression analysis to estimate the change in systolic and diastolic blood pressure, left ventricular mass, lipid levels and insulin per standardized residual change of total body fat mass, android/gynoid fat mass ratio, subcutaneous and preperitoneal area fat mass, independently from body mass index.

1. Keijzer-Veen MG, Euser AM, van Montfoort N, Dekker FW, Vandenbroucke JP, Van Houwelingen HC. A regression model with unexplained residuals was preferred in the analysis of the fetal origins of adult diseases hypothesis. J Clin Epidemiol. 2005 Dec;58(12):1320-4.