RESUMO
OBJECTIVE: Accurate assessment is needed to identify the nutritional status of elderly persons. Anthropometric data were collected to describe body composition of the sample, including blacks and whites aged 55 to 89 years, and to explore the usefulness of several methods of body composition measurement. DESIGN: Baseline measurements were made as part of a longitudinal study. Body composition variables, particularly lean and lean-to-height measures, were used. SUBJECTS/SETTING: One hundred twenty-nine free-living medically stable elderly at senior community centers were self-selected into the study. STATISTICAL ANALYSIS: Descriptive statistics were generated for all variable by gender, race, and age (< 65, 65 to 74, and > or = 75 years). Actual height was correlated with estimated height using published equations. Analysis of variance revealed the effect of gender, race, or age on outcome variables. RESULTS: White women, black women, and white men made up 54%, 23%, and 23% of the sample, respectively. Most gender differences were expected. Black women had greater weight, body mass index, arm muscle circumference, and ratio of lean to height but lower percent lean body mass than white women. Subjects over 75 years old were shorter, lighter, and had lower ratio of lean to height. Gender and age had the greatest effect on ratio of lean to height. CONCLUSIONS/APPLICATIONS: A lean-to-height index appears to be a useful tool for tracking the status of lean mass in the elderly. Knee height may be especially useful because it, unlike stature, does not decrease with age. Furthermore, some published equations for estimating stature from knee height need adjustment, specifically for elderly black women.