Increased upper body and overall adiposity is associated with decreased sex hormone binding globulin in postmenopausal women.

An unfavorable body fat distribution is associated with many metabolic abnormalities including a high prevalence and incidence of noninsulin dependent diabetes mellitus and decreased high density lipoprotein cholesterol and increased triglyceride levels. One mechanism for the effect of body fat distribution on metabolic variables may be through sex hormones. We examined the relationship of body mass index (BMI), ratio of subscapular-to-triceps skinfold ratio (centrality index) and ratio of waist-to-hip ratio (WHR) to sex hormone binding globulin (SHBG) (an in vivo measure of androgenicity) in 101 postmenopausal Mexican-American and non-Hispanic white women from the San Antonio Heart Study, a population based study of diabetes and cardiovascular disease. SHBG was significantly correlated with BMI (r = -0.440, P less than 0.001), WHR (r = -0.255, P less than 0.01) and centrality index (r = -0.210, P less than 0.05). In a multiple linear regression analysis, SHBG remained significantly associated with BMI (P less than 0.001) and WHR (P less than 0.05) but not with age, ethnicity or centrality index. This work suggests that in postmenopausal women overall adiposity and an unfavorable body fat distribution are associated with increased androgenicity as measured by a lower SHBG concentration. Our finding may help to explain the association of body fat distribution with diabetes and cardiovascular risk factors in older women.

Relationship of sex hormone binding globulin to overall adiposity and body fat distribution in a biethnic population.

Previous data have indicated that decreased sex hormone binding globulin (SHBG) is associated with increased overall and upper body adiposity and higher levels of glucose, insulin and triglyceride (TG) and decreased levels of high-density lipoprotein (HDL) cholesterol. Since Mexican Americans have greater overall and upper body adiposity, higher rates of non-insulin-dependent diabetes mellitus, higher TG and lower HDL levels than non-Hispanic whites, we postulated that they would also have lower levels of SHBG. We measured total testosterone and total estradiol using a commercial radioimmunoassy and SHBG using a dextran-coated charcoal technique in premenopausal women (61 Mexican American and 39 non-Hispanic white) as part of the San Antonio Heart Study, a population-based study of diabetes and cardiovascular risk factors. There were no significant ethnic differences in total testosterone or total estradiol. SHBG, however, was lower in Mexican American (0.285 micrograms/dl) than in non-Hispanic white women (0.429 micrograms/dl) (P = 0.009). After adjustment for body mass index (BMI), ratio of waist-to-hip circumference (WHR) and ratio of subscapular-to-triceps skinfolds (centrality index), SHBG remained lower in Mexican Americans (0.307 micrograms/dl) than in non-Hispanic whites (0.396 micrograms/dl), although this difference was no longer statistically significant (P = 0.083). BMI, WHR and centrality index were all negatively associated with SHBG (P less than 0.01). The lower levels of SHBG in premenopausal Mexican American women compared to non-Hispanic white women may reflect greater in-vivo androgenicity and may be related to a variety of metabolic abnormalities seen in this ethnic group.

The relationship of sex hormones to hyperinsulinemia and hyperglycemia.

Mexican-Americans, a high-risk population for non-insulin-dependent diabetes mellitus (NIDDM), have been previously reported to have decreased levels of sex-hormone-binding globulin (SHBG). We measured total testosterone, total estradiol and SHBG, glucose and insulin in premenopausal women (58 Mexican-Americans and 38 non-Hispanic whites) as part of the San Antonio Heart Study, a population-based study of cardiovascular risk factors. Although total estradiol and total testosterone were, in general, not correlated with metabolic variables, SHBG was negatively correlated with glucose and insulin. After adjustment for body mass index (BMI), ratio of waist-to-hip circumference (WHR) and ratio of subscapular-to-triceps skinfold (Centrality Index), SHBG was still significantly correlated with insulin concentrations (P less than .001). Since Mexican-Americans were previously reported to be more hyperinsulinemic than non-Hispanic whites, we examined the effect of adjusting for SHBG on insulin levels in this small population. While unadjusted insulin concentrations in Mexican-Americans were higher than in non-Hispanic whites (354 microU/mL v 236 microU/mL, respectively, P = .009), adjustment for BMI, WHR, and centrality index reduced the ethnic difference in insulin levels considerably (P = .014). However, only after adjusting for SHBG as well, did the ethnic difference in insulin levels became nonsignificant. Our data suggest that alterations in sex hormones and SHBG in particular may be related to the hyperinsulinemia and the high rates of NIDDM in Mexican-Americans.