Postnatal weight and height growth modeling and prediction of body mass index as a function of time for the study of growth determinants.

Weight, height and body mass index (BMI) growth trajectories have been associated with several chronic diseases in later life. Our aim was to describe a method to model individual weight and height growth curves during infancy and to show how it can be used to study their determinants and relationships with later health outcomes as well as to predict BMI trajectories. In the EDEN mother-child cohort, we collected 17 measurements of weight and 16 of length/height per child between birth and 3 years of age in 1,900 infants from their health care booklet and during the study clinical examinations at 1 and 3 years; 1,436 (76%) had at least 1 measurement between 2 and 3 years. We fitted individual weight and height growth trajectories using the Jenss nonlinear model including random effects using the 'SAEMIX' package (R software). We studied whether individual growth model parameters were associated with gender in one- and two-step approaches. We indirectly calculated BMI increase against time from both weight and height growth models combined and compared the fit with a direct multilevel spline model. By modeling observed growth data, we homogenized the data in terms of number and age of measurements and were able to calculate other specific parameters as growth velocities.