The complex segregation analyses performed in our previous studies revealed a significant major gene (MG) effect on the age-adjusted cortical and cancellous bone mineral density (BMD) in two ethnically different populations, Chuvasha and Turkmenians. The aim of the present study was to test the hypothesis of pleiotropic MG control of three components of bone aging, that is, the baseline level of BMD (mu(gs)), the age at onset of the bone mass loss (T(gs)), and the rate of this loss over the years (alpha(gs)). Nuclear and more complex pedigrees from the same two ethnic samples were assessed for hand phalangeal BMD (Chuvasha, 1208 individuals, and Turkmenians, 643 individuals), and complex segregational analysis incorporating age and sex effects directly into MG penetrance function was carried out. The results of the present analysis clearly confirmed the existence of the putative MG and showed that the proportion of BMD variation attributable to this MG effect within the sex was remarkably similar in both populations and ranged between 34.7% and 35.2%. The most parsimonious model for BMD transmission in Chuvasha pedigrees additionally indicated significant residual correlation between siblings and clear sex differences in the annual rates of bone loss alpha(gs). The latter was more than twice as high in females than that in males (0.086 SD vs. 0.033 SD per year). In Turkmenian pedigrees the most parsimonious model presented obvious evidence of the MG control of BMD baseline levels in both sexes with significantly lower baseline levels and younger age at onset (T(gs)) in females. No clear MG effects were inferred on T(gs) and/or alpha(gs) in either sample, either in males or in females. That is, the present study does not suggest MG x SEX x AGE interaction. We suppose that if the rate of age-related changes in phalangeal BMD is genetically determined, then these are not the same genes as those affecting the BMD baseline levels.