Simulation of Huntington's disease onset

Previous investigators have noted that the type of sampling scheme used to study the natural history of Huntington's disease can affect the observed age of onset distribution. Simulated data are presented to demonstrate that bias is introduced when onset characteristics derived from prevalence samples (i.e., samples of affected individuals alive in the population) are used to evaluate cofactors of disease onset and to compute genetic risk estimates for persons "at risk" of developing the disease. This study demonstrates that prevalence sampling underestimates the proportion of Huntington's disease cases in the population with later onset ages. Using plausible values for onset time, duration, and life expectancy, simulation results suggest that the paternal transmission effect on Huntington's disease onset is overestimated when based on prevalence data.