Document Type



During the past half-century the prevalence of obesity in developed countries has increased greatly. Such short-term changes in prevalence must be environmentally determined, but genes can mediate response to environmental change, for example, through variable gene penetrance and expression. Obesity differences between birth cohorts complicate the interpretation of phenotypic comparisons between generations that span periods of change. Genetic segregation analyses of the body mass index in 618 Pima Indian nuclear families (2 generations) identified recessive major gene inheritance, which is the same pattern of transmission that has been found in several other populations. However, within-birth-cohort analyses of siblings (one generation) uniformly supported codominant major gene inheritance with no polygenic heritability. For untransformed data transmission probabilities were Mendelian in the later-born cohort and in the combined sample of siblings. After transformation to remove skewness, transmission probabilities were Mendelian only in the earlier-born cohort (i.e., those siblings bom before a period of marked temporal increase following World War II). A higher penetrance of codominant obesity susceptibility genes in the younger generation would result in some genetically obese individuals who have no affected parents, thus simulating recessive inheritance. Taken together, the results of these analyses suggest that the recessive pattern we identified in the Pima Indian nuclear families could be accounted for by temporal changes in penetrance of codominant obesity-predisposing genes. By implication, more modest temporal increases in white and black populations could in part account for the recessive pattern of inheritance described by several investigators.