Prothro J W, Rosenbloom C A
Department of Nutrition and Dietetics, Georgia State University, Atlanta.
J Gerontol. 1993 Jan;48(1):M15-8. doi: 10.1093/geronj/48.1.m15.
This investigation examined (a) select physical measurements for age and gender differences in an elderly Black population, and (b) the relationship of stature as a dependent variable on knee height, weight, age, and gender in this population. T-test values were greater for men than for women in stature and knee height (p < .001), whereas gender differences in weight were not significant. Multiple regression was used to determine the relationship of stature to the four independent variables. When all Black subjects (n = 119) were included, three variables were significant: 3 square root knee height, gender, and 3 square root weight (R2 = .49). The equation inclusive only of the 98 Black females contained two variables: knee height2 and 3 square root weight (R2 = .29). The probability of both R2's was < .001. In decreasing order, knee height, gender, and weight were important in predicting stature; age was not significant.