Fig. 9From: Using linear and natural cubic splines, SITAR, and latent trajectory models to characterise nonlinear longitudinal growth trajectories in cohort studiesEffect of increasing the number of knots in the random effects spline on the individual growth velocity curves from the natural cubic spline LME model. Figure shows velocity curves for 5 randomly selected individuals, obtained from natural cubic spline LME models in PBMAS males with 2, 4, and 6 knots for the random effects spline curve. All models included 6 knots in the fixed effects spline curve. All knots were placed at quantiles of age distributionBack to article page