Aging and health: a longitudinal study.

A set of double jeopardy hypotheses predicting there is a double health disadvantage in growing old and being a member of an ethnic minority, female, of lower socioeconomic status, and unmarried is tested with data from a 4-yr longitudinal study of older Mexican Americans and Anglos. With the exception of socioeconomic status, little support is found for the double jeopardy predictions in a multivariate analysis of change in health. When subjects who died during the study interval were included in the analysis by assigning them the lowest score on health, the results changed somewhat: the relationship between age and decline in health was strengthened and the effect of sex changed from a direction suggesting that women show greater declines in health with age to a direction suggesting that men experience greater declines in health with age. Implications of the effect of decreased dropouts for studies on aging and health are discussed.