| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Clinical Research Centers |
Division of Gerontology, Department of Medicine, University of Maryland School of Medicine, and Geriatric Research, Education, and Clinical Center, Baltimore Veterans Affair Medical Center, Baltimore, Maryland 21201
Address all correspondence and requests for reprints to: Dora M. Berman, Ph.D., Geriatric Research, Education, and Clinical Center (BT/18/GR), Baltimore Veterans Affairs Medical Center, 10 North Greene Street, Baltimore, Maryland 21201. E-mail: dberman{at}umaryland.edu
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Sex hormone-binding globulin (SHBG) is a serum glycoprotein that binds testosterone with high affinity and estradiol with somewhat lower affinity. In addition to being regulated primarily by sex steroids, hepatic production of SHBG is stimulated by thyroid hormones and cortisol and is inhibited by insulin and PRL (12, 13). SHBG levels are inversely correlated with bioavailable or free testosterone and can be used as an indirect index of relative estrogen to androgen balance (14). In CAU women, increased androgenicity, evidenced by low circulating SHBG, is an independent risk factor for type 2 diabetes and CVD (15, 16). Several studies show a relationship between low levels of SHBG and central obesity, hyperglycemia, hyperinsulinemia, and insulin resistance (17, 18); hypertriglyceridemia and low HDL-C levels (19); and metabolic abnormalities associated with the insulin resistance syndrome (IRS) (20).
In premenopausal CAU women, the relationship between SHBG and HDL-C levels is independent of obesity; however, the relationships between SHBG and the other metabolic abnormalities associated with the IRS disappear after controlling for total body fatness or IAF accumulation (21, 22). These findings suggest that the relationship between SHBG and metabolic abnormalities associated with the IRS in CAU women are mediated by the amount of total as well as intraabdominal adipose tissue (19, 21, 22). In young nonobese women, including subjects with essential hypertension, Hughes et al. (23) showed that AA have lower SHBG levels than CAU. Conversely, in healthy obese premenopausal women, higher SHBG levels are reported in African-American compared with CAU women of similar degrees of obesity (10). Nevertheless, in a large cohort of young obese AA women, SHBG correlated negatively with indexes of overall and central obesity and positively with HDL-C levels and insulin sensitivity (24).
In CAU women menopause is associated with the development of central obesity, insulin resistance, and worsening of glucose and lipid metabolism, conditions that increase the risk for CVD (25, 26). However, similar to findings in premenopausal women, for the same degree of obesity older AA women have less intraabdominal adipose tissue than CAU women (27). This suggests that there may be racial differences in relationships between body fat distribution, sex hormone balance, and metabolic risk factors for CVD in these women. However, there are no studies comparing the relationships between body fat distribution, SHBG, and metabolic risk factors for CVD in postmenopausal AA vs. CAU women. This study tests the hypothesis that obese postmenopausal AA women will have higher levels of SHBG than CAU women, but there will still be a negative relationship between SHBG and indexes of total and central obesity, and profiles of lipid and glucose metabolism in both races.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Fifty-five CAU and 35 AA healthy, overweight and obese [body mass index (BMI), >25 kg/m2] postmenopausal women of similar age from the Baltimore/Washington, D.C., metropolitan area provided informed written consent to participate in this study according to guidelines of the University of Maryland institutional review board for human research. Race was assigned according to the woman’s characterization of herself as either AA or CAU, and all subjects were born in the United States. The women were considered postmenopausal if they had not menstruated for at least 1 yr, and their plasma FSH levels were more than 30 mIU/mL. None of the women had been taking estrogen replacement therapy or medications affecting lipid or glucose metabolism for at least 1 yr. Five CAU and 8 AA women were taking antihypertensive medications known not to affect glucose or lipid metabolism (angiotensin-converting enzyme inhibitors and calcium channel blockers). All women were sedentary (<20 min of exercise twice weekly), weight stable (<2 kg weight change in 2 months), and had not smoked cigarettes for at least 5 yr.
Women were accepted to participate in the study after completing a medical evaluation, which included a medical history, physical examination, fasting blood profile, and electrocardiogram. None of the women had evidence of CVD, diabetes (fasting glucose, >7.0 mmol/L), hyperlipidemia, liver, renal, or hematological disease. A graded exercise test according to a modified-Bruce protocol was performed to exclude subjects with an abnormal cardiovascular response to exercise (27).
Diet record
Women were counseled on the principles of an American Heart Association Step 1 diet (28) for 6 weeks and kept a 7-day food record before research testing.
Body composition, fat distribution, and maximal aerobic capacity (VO2max)
Waist to hip ratio (WHR) was measured as the ratio of the minimal waist circumference to the hip circumference at the maximal gluteal protuberance, and the mean of three values within 2 mm of each other was used as the measurement. Percent body fat, fat-free mass, and lean mass were measured using dual energy x-ray absorptiometry (model DPX-L, Lunar Corp., Madison, WI). Intraabdominal and sc fat areas were measured using a single slice computed tomography (CT) scan taken midway between L4 and L5 performed on a PQ6000 Scanner (Marconi Medical Systems, Cleveland, OH). The VO2max was measured on a motor-driven treadmill during a progressive exercise test to voluntary exhaustion (27). A valid VO2max was obtained when at least two of these three criteria were met: 1) maximal heart rate greater than 90% of age-predicted maximal heart rate (220 beats/min - age), 2) respiratory exchange ratio of at least 1.10, and 3) plateau in VO2 with increasing work rate. If this was not achieved, the test was repeated. The VO2max is expressed as milliliters per kg/min.
Glucose tolerance, lipid metabolism, and hormonal assays
Venous blood samples for measurement of lipoprotein lipid, glucose, and hormone levels were drawn after a 12-h fast on 2 separate days, and the mean of the two measurements of lipoprotein lipids is reported. Plasma leptin (Linco Research, Inc., St. Louis, MO) and serum free testosterone concentrations (measured in 32 CAU and 25 AA women) were measured in duplicate by RIA (Diagnostics Systems Laboratories, Inc., Webster, TX). The intra- and interassay coefficients of variation were, respectively, 5.2% and 3.5% for leptin and 5.0% and 8.3% for free testosterone. Serum SHBG was measured using an immunoradiometric assay (Diagnostics Systems Laboratories, Inc.) with intra- and interassay coefficients of variation of 2.0% and 8.3%, respectively. Plasma free fatty acid concentrations were measured in 31 CAU and 32 AA using a colorimetric assay (Wako Chemicals USA, Inc., Richmond, VA). Total cholesterol, HDL-C, low density lipoprotein cholesterol (LDL-C), and TG levels were measured as previously described (29). A 2-h 75-g oral glucose tolerance test with sampling every 30 min was performed after a 12-h overnight fast in 38 CAU and 25 AA women. Glucose and insulin areas (AUC) above baseline levels during the oral glucose tolerance test (OGTT) were calculated by the trapezoidal method. The plasma glucose concentration was measured in duplicate using the glucose oxidase method (glucose analyzer Beckman Coulter, Inc., Fullerton, CA). Plasma insulin (Linco Research, Inc.) was measured in duplicate by RIAs with intra- and interassay coefficients of variation of 5.0% and 9.0%, respectively.
Statistics
Data are reported as the mean ± SD. Plasma TG and insulin and serum SHBG values were logarithmically transformed to yield normal distributions before parametric analyses. Group differences were assessed using Student’s t test. Analysis of covariance was used to test for relationships between SHBG and risk factors after controlling for obesity, and Pearson’s product-moment correlation coefficients were calculated to examine the relationship of SHBG to regional fat distribution and metabolic variables within and across races. All analyses were performed using Jump Software (SAS Institute, Inc., Cary, NC). Differences were considered significant when P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The AA women had higher BMI, fat mass, and fat-free mass and lower
VO2max compared with the CAU women (Table 1
).
However, the percent body fat was comparable in the two groups. The AA
were 2 yr younger than the CAU women, but the average duration since
menopause (10.0 ± 8.3 vs. 11.7 ± 8.1 yr) was
comparable.
|
). The abdominal fat
distribution measured by CT showed that the AA women had 18% larger sc
fat area (P < 0.01) and a 12% lower
(P = NS) IAF area than the CAU women. Thus, the ratio
of IAF/sc fat was 26% lower (P < 0.01) in AA compared
with CAU women.
|
) of
macronutrients and total calories were similar between the AA and CAU
women, but AA women reported ingestion of diets containing a higher
percentage of unsaturated fats and greater amounts of cholesterol than
CAU.
|
The AA women had similar fasting glucose and glucose AUC during an
OGTT, but higher fasting insulin, insulin AUC (P <
0.05), and leptin levels (P < 0.01) compared with CAU
women (Table 4 and Fig. 1). Despite their
larger waist circumference, lower relative IAF area, and higher fasting
insulin levels, AA had similar SHBG and free testosterone levels as CAU
women. The free testosterone values of these postmenopausal women were
lower than those reported in the literature (30) for obese
premenopausal women (mean ± SE, 5.99
± 0.51 pmol/L). The similarity in SHBG values among women from both
races persisted after adjusting for fasting insulin levels (data not
shown).
|
|
0.80) groups according to their regional
fat distribution. The SHBG levels were similar in the LBO AA
(114.7 ± 63.1 nmol/L; n = 13) and the LBO CAU women
(139.5 ± 70.4 nmol/L, n = 22), as well as in the UBO AA
(119.0 ± 66.6 nmol/L, n = 22) and UBO CAU women (113.1
± 66.6 nmol/L, n = 33). There were no differences in SHBG levels
between LBO and UBO women within each race. Similar results were
observed when comparing free testosterone levels, i.e. LBO
and UBO women from each race had similar levels, and there were no
racial differences in free testosterone levels between LBO or UBO CAU
and AA women (data not shown). Despite the higher insulin levels and similar plasma free fatty acids levels, AA women had 23% lower TG levels than CAU women (P < 0.01). Plasma total cholesterol, LDL-C, HDL-C, and HDL2-C levels did not differ between the races.
However, the racial differences in fasting insulin and insulin AUC could be related to the greater body fatness of AA women compared with CAU women. Analysis of covariance showed that were no significant interactions between race and fat mass for insulin, glucose, and lipid metabolic parameters, and the racial differences in fasting and insulin AUC during an OGTT (P = 0.04) and TG (P < 0.01) levels persisted after adjusting for fat mass, but not after adjusted for other differences in body composition.
Relationships among SHBG, free testosterone, body composition, and metabolism by race
There was an inverse relationship between indexes of total and
central obesity and IAF with SHBG levels in CAU, but not in AA, women
(Fig. 2). The relationship between WHR or
BMI and SHBG was significant in CAU (r = -0.28; P
< 0.05), but not in AA (r = 0.03 to -0.10; P =
NS), women. There was no relationship between weight, fat mass, sc fat
area, or percent body fat and SHBG in either race.
|
). However, at the average fat mass for the entire population
(42.0 kg), AA had higher leptin levels than CAU women (33.0
vs. 26.7 ng/mL, respectively). Leptin levels correlated with
insulin in the CAU (r = 0.38; P < 0.01), but not
in the AA (r = 0.26; P = NS), women. When the
leptin levels were adjusted for insulin, the racial differences in
leptin levels disappeared (P = NS).
In CAU (but not in AA) women, the insulin AUC during an OGTT correlated
negatively with SHBG (r = -0.37; P < 0.05), and
the relationship remained significant after controlling for fat mass
(P = 0.05) or sc fat area (P = 0.04),
but not after controlling for IAF (P = 0.112). There
was a positive relationship between HDL-C and SHBG levels in CAU, but
not in AA, women (Fig. 3), which
persisted after controlling for IAF (P = 0.032), but
not after controlling for fat mass (P = 0.076). There
were no relationships between other lipid metabolic risk factors and
SHBG levels in either race. There also was no relationship between any
measurement of body composition, SHBG, lipids, leptin, or glucose
metabolism and free testosterone in either the AA or CAU women (not
shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SHBG is used as an indirect marker of androgenicity. However, the concept that sex hormones are the main physiological regulators of SHBG levels, and that the estrogen to androgen balance determines SHBG levels is controversial (31). It is possible that altered SHBG levels may reflect a pharmacological, rather than a physiological, influence of sex steroids on liver metabolism (31). In a study comparing SHBG levels among young healthy obese AA and CAU women of similar upper body fat distribution, there were higher SHBG levels in the AA women (10). However, those results differed from findings by Hughes et al. (23), who reported lower SHBG levels in premenopausal AA compared with CAU women, probably because the women in the latter study were not obese, and some had essential hypertension. In contrast, despite larger waist circumference and lower IAF, the older obese AA and CAU women in this study had similar SHBG levels even after adjusting for body fat and fasting insulin levels. This suggests that in older obese AA women, SHBG levels are not related to regional distribution of body fat. The cross-sectional nature of our study precludes any mechanistic conclusions, but the lack of relationship between indexes of regional or total adiposity and SHBG levels in AA women suggests that after menopause sex steroid metabolism may affect regional fat distribution, lipid, and glucose metabolism differently in postmenopausal AA compared with CAU women (7, 8, 9, 10, 32).
After adjusting for fat mass, AA women still had higher fasting- and glucose-induced postprandial hyperinsulinemia than CAU women. These findings together with the reported greater degree of insulin resistance found in AA compared with CAU women (8, 11) are consistent with the higher prevalence of type 2 diabetes in AA women (33). Our finding of a negative relationship between the insulin response to oral glucose and SHBG levels only in CAU women is similar to previous reports in premenopausal CAU and AA women (21, 22, 24), but the effect of obesity (total and abdominal) on the relationship between insulin and SHBG is not clear. In premenopausal CAU women the relationship between insulin responses during an OGTT and SHBG was influenced by total or intraabdominal adiposity (21, 22), whereas in a large mixed cohort of Mexican-American and CAU postmenopausal women, plasma 2-h insulin correlated with SHBG even after adjusting for BMI and WHR (34). In the present study there was a relationship between insulin area and SHBG in CAU women (but not in AA women), and this became nonsignificant after adjusting for visceral fat area, but not for total fat mass. This suggests that increased visceral obesity affects the relationship between the insulin response to oral glucose and SHBG levels in postmenopausal CAU women.
Leptin levels were higher in AA women than in CAU women, and there was a racial difference in the slopes of the relationships between leptin and fat mass. We previously reported lower leptin levels and similar slopes for the relationship between leptin and fat mass in AA and CAU postmenopausal women (35). However, in that study the women were older (64 yr) and leaner (80–85 kg), and there were no racial differences in fat mass. Although there were similar leptin levels in one study comparing AA and CAU women ranging from 18–89 yr, there were racial differences in the slope of the relationships between leptin and fat mass (36). The fact that the racial differences in leptin levels between older AA and CAU women disappeared after adjusting for fasting insulin levels supports the recently reported association of high leptin levels with hyperinsulinemia independent of obesity (37). However, the cause-effect nature of this association cannot be established from these cross-sectional studies. Dietary weight loss interventions will be needed to determine whether hyperinsulinemia and insulin resistance increase leptin or vice versa in AA women, and their relationship to total and regional obesity, metabolism, and race.
There is controversy regarding the effect of obesity on the positive relationship between HDL-C and SHBG in CAU women. In young CAU women, the relationship between HDL-C and SHBG persists after adjustment for differences in intraabdominal or total body fatness (22). In postmenopausal CAU women, Soler et al. (38) reported that the association between HDL-C and SHBG was largely mediated by abdominal obesity, whereas the same association was reportedly independent of obesity and body fat distribution in a mixed group of older CAU and Mexican-American women (34). In the present study the relationship between HDL-C and SHBG in older CAU women was independent of visceral obesity, but was mediated in part by total adiposity. Further studies are needed to elucidate the mechanisms by which obesity affects the relationship between HDL-C and SHBG. The lower TG levels are a consistent finding in older (39) as well as in younger AA compared with CAU women (7, 8, 10, 40). Although the metabolic mechanisms underlying these differences in TG levels are not certain, Friday et al. (41) and Després et al. (40) reported higher levels of postheparin lipoprotein lipase activity and lower levels of hepatic triglyceride lipase activity in young and older AA compared with CAU men and women. These results suggest that AA women clear TG from the blood more effectively than CAU women.
It is also possible that the racial differences we observed in adiposity and metabolism could be influenced by racial differences in the amount and/or quality of nutrients in their respective diets. This is possible over the life span, but at the time of the study these women’s reported dietary records showed the amount of kilocalories ingested was similar between races, but the AA women had a slightly, but significantly, higher unsaturated fat and cholesterol intake. It is unlikely that these small differences affected our findings.
In contrast to findings in CAU women, we did not find a relationship between total or regional adiposity or other metabolic risk factors and SHBG in older AA women. These differences could be mediated at least in part by the racial differences in fat cell (32, 41) as well as sex steroid metabolism after menopause. In a cross-sectional comparison of pre- and postmenopausal CAU women with different androgen to estrogen ratios in the Virgilio-Menopause Health Project (42), there were similar SHBG levels in young and older women. However, no such studies have been performed in AA women.
In conclusion, we found that AA women have comparable SHBG levels to CAU women despite having higher total adiposity and lower relative IAF. However, in contrast to CAU women there is no relationship between total and central obesity, insulin, or HDL-C and SHBG in older obese AA women. Furthermore, AA women have higher insulin and leptin levels and lower TG levels than CAU women. This suggests there are fundamental differences in the effects of sex hormones on metabolism and body composition in obese AA compared with CAU postmenopausal women. Understanding the mechanisms by which sex hormones affect total and regional body fat distribution and the risk factors for the IRS and CVD could have important therapeutic and preventive health implications for AA women.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received June 12, 2000.
Revised September 25, 2000.
Accepted October 2, 2000.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
evac I, Sucic M, Juric
M. 1998 Hyperinsulinemia and sex hormones in healthy premenopausal
women: relative contribution of obesity, obesity type and duration of
obesity. Metabolism. 47:13–19.[CrossRef][Medline]
-Hydroxyprogesterone responses to leuprolide and serum androgens
in obese women with and without polycystic ovary syndrome after dietary
weight loss. J Clin Endocrinol Metab. 82:556–560.This article has been cited by other articles:
 |
|
 |
|
  B. Sternfeld, K. Liu, C. P. Quesenberry Jr., H. Wang, S.-F. Jiang, M. Daviglus, M. Fornage, C. E. Lewis, J. Mahan, P. J. Schreiner, et al. Changes over 14 Years in Androgenicity and Body Mass Index in a Biracial Cohort of Reproductive-Age Women J. Clin. Endocrinol. Metab., June 1, 2008; 93(6): 2158 - 2165. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. P. Rose, S. M. Haffner, and J. Baillargeon Adiposity, the Metabolic Syndrome, and Breast Cancer in African-American and White American Women Endocr. Rev., December 1, 2007; 28(7): 763 - 777. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. You, A. S. Ryan, and B. J. Nicklas The Metabolic Syndrome in Obese Postmenopausal Women: Relationship to Body Composition, Visceral Fat, and Inflammation J. Clin. Endocrinol. Metab., November 1, 2004; 89(11): 5517 - 5522. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Yoo, T. Nicklas, T. Baranowski, I. F Zakeri, S.-J. Yang, S. R Srinivasan, and G. S Berenson Comparison of dietary intakes associated with metabolic syndrome risk factors in young adults: the Bogalusa Heart Study Am. J. Clinical Nutrition, October 1, 2004; 80(4): 841 - 848. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. E Ruhl, J. E Everhart, J. Ding, B. H Goodpaster, A. M Kanaya, E. M Simonsick, F. A Tylavsky, and T. B Harris Serum leptin concentrations and body adipose measures in older black and white adults Am. J. Clinical Nutrition, September 1, 2004; 80(3): 576 - 583. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Rennie, N McCarthy, S Yazdgerdi, M Marmot, and E Brunner Association of the metabolic syndrome with both vigorous and moderate physical activity Int. J. Epidemiol., August 1, 2003; 32(4): 600 - 606. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Kim, S. Nam, C. Ahn, K. Kim, S. Yoon, J. Kim, B. Cha, S. Lim, K. Kim, H. Lee, et al. Correlation Between Midthigh Low- Density Muscle and Insulin Resistance in Obese Nondiabetic Patients in Korea Diabetes Care, June 1, 2003; 26(6): 1825 - 1830. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. R. Fontaine, D. T. Redden, C. Wang, A. O. Westfall, and D. B. Allison Years of Life Lost Due to Obesity JAMA, January 8, 2003; 289(2): 187 - 193. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) and AA (•) women.