Significant changes in body composition that have important health related effects are believed to occur in the elderly. Knowledge of these changes is particularly important for diagnoses, prognoses, and treatment of health problems. Many health problems in the elderly could be prevented or alleviated by nutritional modulation, but better understanding of the nature, extent, and underlying physiology of body compositional changes is needed for such interventions to be successful. There are currently few data for body composition in the elderly, especially for those greater than 75 y in age, partly because conventional methods of assessing body composition are difficult to apply for technical and conceptual reasons. As a result, little is known regarding the relationships of body composition to nutritional, functional or health status in non- hospitalized, free-living elderly persons. Knowledge from studies of young and middle-aged populations should not be assumed to apply to the elderly, who are more heterogeneous as a group due to a life-time of accumulated biological and historical experiences. Longitudinal studies that include an extensive array of nutritional, physiological, and behavioral measures are needed to determine the complex role of body compositional changes in the aging process. Knowledge of the "natural history" of body compositional changes and their relationships to other nutritional and health factors could lead to new insights on prevention and treatment, the reduction of morbidity and extension of the quality of life of older persons. The primary goals of the proposed study are to determine whether: (a) body composition changes with age in elderly persons; (b)rates of change accelerate with age; (c) variation among elderly individuals in body composition is associated with levels of anabolic hormones, blood pressures, and lipid/lipoprotein cholesterols; and, (d) body composition predicts falls. Annual, serial determinations of whole and regional body composition will be made in approximately 500 free-living, elderly men and women who are presently participants in an on-going longitudinal study, the New Mexico Aging Process Study, over a 5 y period. This cohort was selected because extensive nutritional and health information has been collected over the past decade for these individuals and will continue to be collected over the study period. This will make the proposed study highly cost-efficient. It will be possible to conduct complex retrospective, as well as cross-sectional and prospective analyses of factors associated with body composition. New techniques of analyzing body composition in physiologically realistic multicompartmental models will be applied that maximize accuracy and precision, and that allow estimation of multiple components of body composition that have not been studied previously. This is a revised proposal, previously reviewed in October 1992.