Group descriptive statistics and generalized linear mixed model (GLMM) were used to characterize and compare baseline and follow-up measures. Familial clustering was taken into account in all GLMMs. Furthermore, baseline age was treated as a covariate when comparing baseline characteristics in a sensitivity analysis. Effect sizes (either odds ratio or proportion estimate and corresponding 95% confidence interval) were used to assess the degree that the three hypotheses were supported by the data. To correct for multiple comparisons, only variables with False Discovery Rate (FDR)-adjusted p-values < 0.05 were considered statistically significant (Benjamini and Yekutieli, 2001). SAS 9.4 was used for all analyses.
