Gynoid and android fat patterning in Japanese-American men: Body build and glucose metabolism

Subcutaneous fat patterning was investigated in a sample of 229 second generation (Nisei) Japanese-American men enrolled in an epidemiological study of noninsulin-dependent diabetes mellitus. The study sample consisted of 78 men with diabetes, 72 with impaired glucose tolerance (IGT), and 79 with normal glucose tolerance, classified by a 75 gm oral glucose tolerance test(OGTT) and world Health Organization criteria. A combination of hierarchal cluster analysis and logarithmic regression analysis using skinfold thickness at seven anatomical sites yielded two distinctive fat pattern groups, viz, a gynoid fat pattern group (n = 16), with a significantly greater percentage of fat on the extremities; and an exaggerated android group (n = 34), with proportionately more fat on the trunk than predicted from total skinfold fat thickness. The two fat pattern groups and a sample of men of normal glucose tolerance (n = 53) were compared by analysis of variance with respect to 12 somatometric variables and indices encompassing length, breadth and muscle mass parameters. The three groups did not differ in age or height. In weight and body mass index, the gynoid fat pattern group was lowest, although there were no significant intergroup differences in total skinfold fat (seven sites summed). The gynoid group was also significantly less robust in body build than the android group, as judged by a lower intermembral index, lower ratios of bony chest breadth and biacromial diameter to sitting height, and reduced arm muscle mass. The prevalence of diabetes within each fat pattern group was slightly less than expected on the basis of the prevalence of IGT (44%) was found within the gynoid group; none of the differences were statistically significant. During the OGTT, the gynoid and android groups, minus diabetic men, differed significantly from men with normal glucose tolerance in serum glucose levels at 1.5 and 2 hr but not from one another. The limitations of cross-sectional data of this nature for determining a differential risk fordisorders of glucose metabolism associated with type of fat deposition are discussed.