We commend Weissman for calling attention to the desirability of clarifying the meaning of racial categories used in health studies. Weissman's comments follow by almost 20 years similar concerns expressed by Milton Terris.1 For our study of body mass index and body girth associations with mortality,2 the racial classification used was from the 1960 examination of the Charleston Heart Study cohort and was based on interviewer's evaluation. No attempt was made to distinguish the "mixed racial heritage" Weissman refers to. Skin color was measured by photoelectric reflectance and, although not used for classification, there was little overlap in the distribution of values in the two groups.
The Charleston Heart Study investigators have long recognized the combined influence of genetic and social and physical environmental effects on health indexed by classification of race.3,4 Nowhere in our study of obesity and fat patterning in black and white women did