Estimation of variance components for age at menarche in twin families

Age at menarche (AAM), time of first menstrual period, is an important developmental milestone in females. Follow-up data from 1,302 adolescent twins and their sisters were used to partition the normal variation in AAM. The proportion of censoring was 14.1%. Both a standard and a survival analysis method were used. The best fitting model from the survival analysis method was an ACE model, where 57% and 23% of the variance in AAM was explained by additive genetic and environmental effects, respectively. The best fitting model when using a standard variance decomposition method was an AE model, where 82% of the variance was explained by additive genetic effects. The lack of correspondence between the results of the two methods was an artefact of the different ascertainment of AAM reports from siblings and twins. After the removal of the sibling sample, both methods indicated that an ACE model was the most likely. Standard and survival analysis methods estimated the proportion of variance explained by additive effects to be 0.50 and 0.54, and common environmental effects to be 0.31 and 0.29, respectively. We conclude that variation in AAM can be explained by additive genetic and common environmental components.