- Email: [email protected]
Central abdominal fat and endogenous hormones during the menopausal transition
FERTILITY AND STERILITY威 VOL. 79, NO. 6, JUNE 2003 Copyright ©2003 American Society for Reproductive Medicine Published by Elsevier Inc. Printed on acid-free paper in U.S.A.
Central abdominal fat and endogenous hormones during the menopausal transition Received July 22, 2002; revised and accepted November 14, 2002. Presented in part at the 10th International Menopause Society Congress, Berlin, Germany, June 14, 2002. Supported by the National Health and Medical Research Council of Australia and the Victorian Health Promotion Foundation. During 2001, the research program received unrestricted educational grants from Australia and New Zealand Banking Group (ANZ) Trustees, Eli Lilly (Australia), Pharmacia, and Upjohn (Australia); Prince Henry’s Institute of Medical Research received grants from Organon for hormone assays. Reprint requests: Janet R. Guthrie, Ph.D., Office for Gender and Health, University of Melbourne, Charles Connibere Building, Royal Melbourne Hospital, Victoria 3050, Australia (FAX: 613-93474127; E-mail: janetrg@ unimelb.edu.au). a Office for Gender and Health. b Department of Psychiatry, Monash University, Clayton, Victoria, Australia. c Department of Diabetes and Endocrinology, Royal Melbourne Hospital, Victoria, Australia. d University of Michigan Health System, Ann Arbor, Michigan. e Prince Henry’s Institute of Medical Research, Clayton, Victoria, Australia. f Department of Medicine. 0015-0282/03/$30.00 doi:10.1016/S0015-0282(03) 00361-3
Janet R. Guthrie, Ph.D.,a Lorraine Dennerstein, M.B., Ph.D.,a John R. Taffe, Ph.D.,b Peter R. Ebeling, M.D.,c John F. Randolph, M.D.,d Henry G. Burger, M.D.,e and John D. Wark, M.B., Ph.D.f The University of Melbourne, Victoria, Australia
Objective: To investigate the effect of endogenous hormone levels on central abdominal fat during the menopausal transition in a population-based cohort of Australian-born women. Design: Prospective observational study. Setting: Population-based sample. Body composition was assessed in the Royal Melbourne Hospital, and interviews were conducted at the patient’s home. Subject(s): One hundred two women from the Melbourne Women’s Midlife Health Project. Data, physical measures, and blood were obtained by interview when the longitudinal study commenced (baseline) and at the time of the total body scan approximately 5 years later. Body composition was measured using dual-energy X-ray absorptiometry. Intervention(s): None. Main Outcome Measure(s): Total body fat and central abdominal fat. Result(s): The 102 women were either premenopausal or in the early menopausal transition at baseline. At the time of their dual-energy X-ray absorptiometry scan, 31 were in the early menopausal transition, 22 were in the late menopausal transition, and 49 were postmenopausal. Multiple regression analysis found that total percentage of body fat was associated with weight measures, whereas central abdominal fat was also positively associated with baseline free T index (FTI) and with the increase in FTI since baseline. Conclusion(s): The major hormonal change associated with central adiposity during the menopausal transition is the increase in the FTI. This effect is significant even after allowing for baseline and final weight. (Fertil Steril威 2003;79:1335– 40. ©2003 by American Society for Reproductive Medicine.) Key Words: Central abdominal fat, percent body fat, free testosterone index, estradiol, menopausal transition, weight
Visceral abdominal fat is independently associated with the development of dyslipidemia, hypertension, and hyperinsulinemia in women (1). These metabolic abnormalities increase the risk of developing cardiovascular disease and type 2 diabetes. Visceral abdominal fat drains directly into the portal circulation, releasing free fatty acids and causing more insulin resistance than subcutaneous abdominal fat, which drains into the systemic circulation (2). Several population-based longitudinal studies (3–5) have shown that around the time of menopause, women experience an increase in weight. Cross-sectional studies using dual-energy X-ray absorptiometry (DXA) have reported menopause-related changes in body
composition and body fat distribution (6 – 8), although not all changes were significant (7, 8). There are a number of methodological problems with these studies: some were based on convenience (6) or clinical samples (8) with inherent biases; one defined women as postmenopausal after 6 months of amenorrhea (8), and one excluded obese women (6). Menopause is associated with major hormonal changes, including a large increase in FSH and a profound fall in E2 levels (9), but there are also changes in sex hormone– binding globulin (SHBG) and the free T index (FTI) (10). The role of these hormonal changes in the changes in body fat distribution during the 1335
menopausal transition has not been investigated in an Australian-born population. The Melbourne Women’s Midlife Health Project, a 9-year prospective population-based study of Australianborn women’s experiences of the menopausal transition, provided an opportunity to study variables associated with body composition. Although only one DXA measure of body composition was made, the prospective design of this project allowed baseline measures as well as annual changes to be used to predict body fat distribution. The aim of this paper is to describe the effect of endogenous hormone levels on central abdominal fat in a populationbased cohort of women experiencing the menopausal transition.
MATERIALS AND METHODS The study was approved by the Human Research Ethics Committee of the University of Melbourne, and the procedures followed were in accordance with the ethical standards of the National Health and Medical Research Council. All subjects provided written informed consent for their participation in the study.
Subjects The Melbourne Women’s Midlife Health Project began in 1991 with population sampling by random telephone digital dialing and baseline interviews of 2,001 Australian-born women aged 45–55 years and residing in Melbourne (71% response rate) (11). All of those women at baseline who had experienced menses in the prior 3 months, and who were not taking the oral contraceptive pill or hormone therapy, were invited to participate in a longitudinal study. Of those eligible, 56% accepted (n ⫽ 438). Volunteers for the longitudinal study were more likely than nonparticipants to report better self-rated health, paid employment, more than 12 years of education, having ever had a Papanicolaou smear, exercising at least once a week, and having undergone dilatation and curettage (12). The retention rate by year 8 of follow-up was 88%. The cohort for the current study comprised 139 of these women who had previously participated in a prospective hip and lumbar spine bone mineral density study (13). These women were invited to return for a third bone densitometry visit during the sixth year of follow-up, which included a whole body bone scan.
Body Composition Body composition was measured by DXA using a Hologic QDR-1000W densitometer in the Bone Densitometry Unit, Department of Medicine, Royal Melbourne Hospital, and Hologic Software Version 5.35 (Hologic, Inc., Waltham, MA). Body composition was analyzed according to a twocompartment model, comprising fat mass and lean tissue. The method used to measure central abdominal fat content was similar to that described by Carey et al. (14). 1336 Guthrie et al.
Abdominal fat and endogenous hormones
The central abdominal fat region was manually determined and included the area between the upper surface of the second lumbar vertebra, the lower surface of the fourth lumbar vertebra, and the lateral margins of the outer rib cage. This area has been shown by magnetic resonance imaging to contain a relatively high visceral and low subcutaneous fat content (15). The coefficient of variation for measuring percentage of body fat using the Hologic QDR-1000W densitometer is 1.3% (16). Weight and height were measured as described previously (3). Three measures of lifestyle related behavior were recorded annually: exercise, which was recorded as frequency of engaging in exercise for fitness or recreational purposes (17), current smoking (Yes/No), and alcohol intake (number of drinks per week).
Hormone Measures Fasting morning blood samples for hormone immunoassays were collected at baseline and at the time of the DXA measurements. Blood was taken between days 4 and 8 of the menstrual cycle if the women were cycling. If not cycling, blood was taken after 3 months or more of amenorrhea. E2 was measured using the double-antibody RIA kit purchased from Diagnostic Products Corporation (DPC), Los Angeles. Assay sensitivity (as detection limit) was 20 pmol/L. The interassay coefficient of variation was 6.6% at 400 pmol/L, and the intra-assay coefficient of variation was 11% at 110 pmol/L, 13% at 470 pmol/L, and 13% at 1,160 pmol/L. Serum T was measured by double-antibody RIA after sample extraction and polyethylene glycol– enhanced separation of bound from free ligand, using 125I iodinated T as a tracer. The between-assay coefficient of variation at a T level of 2 nmol/L was 12%. The “normal” female range for serum T using this assay is quoted as 1.0 –2.8 nmol/L. SHBG was measured by an automated chemiluminescent enzyme immunoassay (DPC) using the Immulite Automated Analyser. The intra- and inter-assay coefficients of variation were 4.9% and 5.8%, respectively. FTI was calculated as the ratio of measured T to measured SHBG ⫻ 100.
Menopausal Status Data on menstrual history were collected at the time of interview, and the date of final menstrual period (FMP) was verified from prospectively kept menstrual diaries. Menopausal status was determined at these two time points by menstrual status. Women were defined as premenopausal if they reported no change in menstrual frequency in the prior 12 months. Women who had menstruated in the previous 3 months with changes in menstrual frequency were considered to be in early menopausal transition. Women who reported at least 3 months but less than 12 months of amenorrhea were considered to be in late menopausal transition. Women were considered postmenopausal if they had had amenorrhea for at least 12 months (18). Vol. 79, No. 6, June 2003
TABLE 1 Body composition variables at baseline and at time of dual-energy X-ray absorptiometry (DXA) scan in the whole cohort (n ⫽ 102). Continuous variables are reported as mean (SD). Variables Weight (kg) BMI ⱖ 30kg/m2 (%) Total fat mass (kg) Total lean mass (kg) Total body fat (%) Central abdominal fat mass (kg) Fat in central abdomen (%)
Baseline
Time of DXA scan
68.6 (14.4) 16.7 – – – – –
69.1 (13.8) 17.6 24.3 (10.9) 44.1 (5.0) 33.0 (7.7) 1.39 (0.81) 33.7 (10.6)
Note: BMI ⫽ Body mass index (weight [kg]/height [m]2). Guthrie. Abdominal fat and endogenous hormones. Fertil Steril 2003.
Statistical Analysis Because of skewed distributions, hormone values were log transformed before analysis. Multiple linear regression analysis was used to determine which variables independently predicted the body composition measures. Analysis of variance was used to examine differences in body composition among menopausal status groups. Independent t-tests were used to examine the differences in body composition depending on exercise, smoking, or alcohol intake habits. Unless otherwise noted, a nominal level of significance of P⬍.05 was used and all statistical tests were two tailed. SPSS software (19) was used for all analyses.
RESULTS At baseline, all 139 women who participated in the study were premenopausal (n ⫽ 48) or in the early menopausal
transition (n ⫽ 54). At the time of their DXA scans, 31 of these women were in the early menopausal transition, 22 were in the late menopausal transition, and 49 were naturally postmenopausal and were on average 34.0 months (SD ⫾ 12.2 months) past FMP. Twenty-eight women were taking or had taken hormone therapy, and nine had experienced surgical menopause. Hormone users and those with a surgical menopause were excluded from the analyses, leaving a cohort of 102 women with a mean age of 48.9 years (SD ⫾ 2.4 years) at baseline. Although only one DXA measure of body composition was obtained, we were able to use the women’s prospectively acquired data since baseline as independent variables. At baseline, 20% of the cohort were smokers, 17% consumed eight or more alcoholic drinks per week, and 61% participated in exercise for fitness or recreational purposes more than once a week. At the time of their DXA scans, 5 years later, these percentages were 15%, 17%, and 75% respectively. Table 1 shows the DXA scan measures and weight and body mass index (BMI) information of the cohort at baseline and at the time of their DXA scans. Table 2 shows the levels of FSH and E2 and the FTI at baseline and at the time of the DXA scan by menopausal status. At the time of the DXA scan, women in the early menopausal transition had significantly higher E2 levels compared with women who were in the late menopausal transition (P⬍.05) or postmenopausal (P⬍.0005), and FSH levels were significantly lower in the early menopausal transition compared with the late menopausal transition and postmenopause (P⬍.001 in both cases). There was no significant difference among the menopausal transition groups at the time of the DXA scan with respect to FTI values. Overall, the FTI levels at the time of the scan
TABLE 2 Hormone levels (geometric mean, 95% confidence interval) at baseline and at time of dual-energy X-ray absorptiometry (DXA) scan by menopausal status. Baseline
Time of DXA
Hormone
Menopausal status
Hormone levels
FSH (IU/L)
Premenopausal Early MT
E2 (pmol/L)
Premenopausal Early MT
259.1 (200.6–334.8) 245.9 (171.4–352.8)
Free T Index
Premenopausal Early MT
2.03 (1.73–2.37) 1.58 (1.34–1.87)
7.7 (6.4–9.1) 13.2 (9.9–17.5)
Menopausal status Early MT Late MT Postmenopausal Early MT Late MT Postmenopausal Early MT Late MT Postmenopausal
Hormone levels 25.3 (16.6–38.6) 93.7 (78.1–112.4) 111.8 (100.9–123.9) 116.4 (71.6–189.3) 58.1 (34.3–98.2) 23.5 (21.1–26.1) 2.78 (2.23–3.46) 2.75 (2.14–3.54) 2.87 (2.45–3.37)
Note: The cohort was comprised of 102 women: at baseline 48 were premenopausal and 54 were in the early menopausal transition; at the time of the DXA scan, 31 were in the early menopausal transition, 22 in the late menopausal transition, and 49 in postmenopause. MT ⫽ menopausal transition. Guthrie. Abdominal fat and endogenous hormones. Fertil Steril 2003.
FERTILITY & STERILITY威
1337
TABLE 3 Multiple linear regression with dual-energy X-ray absorptiometry measure of central abdominal fat as dependent variable. Independent variables
Dependent variable % Fat in central abdomen
Baseline weight Change in weight Baseline FTI Change in FTI E2 Change in E2 Menopausal status Age

P
⫹0.5 ⫹0.7 ⫹4.7 ⫹3.7 ⫹0.5 ⫺0.7 ⫹1.7 ⫹0.2
⬍.001 .001 .02 .04 .6 .4 .5 .7
Note: FTI ⫽ free T index. Guthrie. Abdominal fat and endogenous hormones. Fertil Steril 2003.
were significantly higher than baseline FTI values, as were FSH levels, and E2 levels were significantly lower (P⬍.001). Women who were in the early menopausal transition had a lower percentage of abdominal fat compared with women in the late menopausal transition or in postmenopause. The values were 31.4% (SD ⫾ 1.8), 37.3% (SD ⫾ 2.4), and 33.5% (SD ⫾ 1.5) for the early menopausal transition, late menopausal transition, and postmenopause, respectively. The differences between the early and late menopausal transition groups were just significantly different (P ⫽ .05). Multiple regression analysis found that total body fat and central abdominal fat measures were both positively associated with weight measured at baseline (P⬍.0001), which was on average 5 years before the body composition measurements when all members of the cohort were still menstruating, and with change in weight over this period (P ⫽ .001) after controlling for age, E2 levels, and menopausal status. Central abdominal fat was also positively associated with baseline FTI and with a change in FTI since baseline (P⬍.05) (Table 3). Women in the sample who currently participated in physical activities for fitness or recreational purposes more than once a week tended to have lower central abdominal and total percent body fat than women who exercised once a week or less (32.9%, SD ⫾ 11.1 vs. 36.5%, SD ⫾ 8.7 for central abdominal fat; 32.5% SD ⫾ 7.4, vs. 34.8%, SD ⫾ 8.7 for total percent body fat), but the differences were not statistically significant. There was no significant association between fat measures and age, alcohol intake, or smoking in this cohort.
age and E2 levels. Although serum total T levels appear to remain unchanged during the menopausal transition (14), SHBG has been shown to decrease during this period (14). The time of maximum change is estimated at 2 years before the FMP, and consequently the FTI increases. SHBG levels are also negatively associated with BMI and positively correlated with E2 levels. About one-third of the decline in SHBG was explicable by BMI and changes in estrogen levels (14). This decrease in SHBG and subsequent increase in free T may act directly on fat stores or may modify expression of genes controlling adipocyte differentiation and metabolism. Changes at the cellular level, such as a decrease in lipoprotein lipase activity in femoral adipose tissue but not in abdominal adipose tissue, have been reported after menopause (20, 21). This could explain a shift of adipose accumulation toward the abdominal area. The exact mechanism underlying the association of FTI and other hormones with the distribution of body fat is yet to be elucidated. However, this androgen-influenced storage of fat in the intra-abdominal area may well predispose women to a greater risk for cardiovascular disease and diabetes (1) and has important health implications for postmenopausal women. Total percent body fat and waist-to-hip ratios have previously been reported to be positively associated with T concentrations in pre- and perimenopausal women by Sowers and colleagues (22). The measure used in this latter study was total rather than free, bioavailable T. In our cohort, total percent body fat was positively associated with baseline weight and change in weight during follow-up. This observation is to be expected, as weight is primarily related to fat and lean body mass and in midlife women, the increase in weight is likely to be an increase in fat mass rather than an increase in muscle and bone mass (23). It was only the percent fat in the abdominal region that was associated with free T levels, after allowing for weight and weight changes. Our cohort was predominately postmenopausal at the time of the DXA measures, and this may explain the differences in the findings for total percent body fat between the Sowers study (22) and ours.
DISCUSSION
Our study and others (7, 8) have used DXA to measure abdominal fat content. DXA has been shown to be a highprecision method for measuring body composition and in particular has demonstrated a smaller variability in percentage fat values compared with other methods (16). Some studies (6) have used computed tomography (CT). DXA measurements have been shown to compare well with those from CT. We positioned our measurement of central abdominal fat on an area shown by magnetic resonance imaging to contain a relatively high content of visceral fat and lowsubcutaneous fat (15).
Our results suggest that sex hormone concentrations affect fat distribution. We found an association between a measure of central body fatness and the FTI, independent of
CT measures do allow the independent measurement of abdominal subcutaneous and intra-abdominal fat. Toth et al. (6) found, using CT, that intra-abdominal fat but not abdom-
1338 Guthrie et al.
Abdominal fat and endogenous hormones
Vol. 79, No. 6, June 2003
inal subcutaneous fat was greater in early postmenopausal compared with premenopausal women, the age range of the sample being comparable with ours. The mean values for fat mass were lower in Toth’s study of American women (6) compared with our Australian population, and they showed a large difference in body fat between pre- and postmenopausal women. This may be explained by the exclusion from the American study of women with a BMI ⱖ30 kg/m2 rather than a difference in methods of measurement. Our cohort was population based, and 17% of the women had a BMI ⱖ30 kg/m2. The increase in fat mass in middle-aged women is most probably due to a combination of behavioral and physiological mechanisms, including a decrease in energy expenditure, an increase in energy intake, and a decrease in metabolic rate (23). We found a trend for lower values of total body fat and abdominal fat in women who participated in regular exercise. There was no evidence of preferential reduction in abdominal fat in response to regular exercise. Although a number of studies have reported that exercise is associated with reductions in abdominal fat, others have not found an effect (15). The review by Ross and Janssen (15) found an absence in the literature of randomized controlled trials that employed imaging techniques to determine changes in abdominal fat. Abdominal obesity was measured by waist circumference in most trials (15). A recent paper (24) using DXA and CT scans reported that visceral adipose tissue decreased in obese postmenopausal women who increased their maximal oxygen uptake during a 6-month walking program. A genetic influence on central abdominal fat levels has been reported in a study of nonobese Caucasian monozygotic and dizygotic female twins (14). The investigators reported that the majority of intersubject variance in central abdominal fat was due to genetic factors. This genetic influence on abdominal fat was present after adjustment for genetic effects due to total fat, suggesting the existence of genes specific for abdominal obesity. We did not have the relevant information to identify women in our study whose abdominal fat levels may have been influenced by genetic factors. Our finding of the association between body composition and baseline weight is similar to findings in other studies on predictors of developing carotid atherosclerosis (25) and impaired fasting glucose (26). These studies all indicate that premenopausal women should be counseled on lifestyle changes that might prevent adverse health outcomes in postmenopause. In conclusion, results from this study indicate that the major hormonal change associated with central adiposity during the menopausal transition is an increase in the FTI. FERTILITY & STERILITY威
This effect is significant even after allowing for baseline and final weight.
Acknowledgments: The authors thank Sue Kantor, who performed the DXA measurements for this study, and Mr. N. Balazs and his staff in the Department of Biochemistry at the Monash Medical Centre for the hormone assays. We also thank all the women who participated in this study.
References 1. Despres J-P, Moorjani S, Lupien PJ, Tremblay A, Nadeau A, Bouchard C. Regional distribution of body fat, plasma lipoproteins, and cardiovascular disease. Atherosclerosis 1990;10:497–511. 2. Yamashita S, Nakamura T, Shimimura I, Nishida M, Yoshida S, Kotani K, et al. Insulin resistance and body fat distribution. Diabetes Care 1996;19:287–91. 3. Guthrie JR, Dennerstein L, Dudley EC. Weight gain and the menopause: a 5-year prospective study. Climacteric 1999;2:205–11. 4. Blumel JE, Casrelo-Branco C, Rocangliolo ME, Bifa L, Tacla X, Mamani L. Changes in body mass index around menopause: a population study of Chilean women. Menopause 2001;8:239 –44. 5. Wing RR, Matthews KA, Kuller LH, Meilahn EN. Weight gain at the time of the menopause. Arch Intern Med 1991;151:97–102. 6. Toth MJ, Tchernof A, Sites CK, Poehlman ET. Effect of menopausal status on body composition and abdominal fat distribution. Int J Obes 2000;24:226 –31. 7. Svendsen OL, Hassager C, Christiansen C. Age- and menopauseassociated variations in body composition and fat distribution in healthy women as measured by dual-energy x-ray absorptiometry. Metabolism 1995;44:369 –73. 8. Tremollieres FA, Pouilles J-M, Ribot CA. Relative influence of age and menopause on total and regional body composition changes in postmenopausal women. Am J Obstet Gynecol 1996;175:1594 –600. 9. Burger HG, Dudley EC, Hopper JL, Groome N, Guthrie JG, Green A, et al. Prospectively measured levels of serum follicle stimulating hormone, estradiol, and the dimeric inhibins during the menopausal transition in a population-based cohort of women. J Clin Endocrinol Metab 1999;84:4025–30. 10. Burger HG, Dudley EC, Cui J, Dennerstein L, Hopper JL. A prospective longitudinal study of serum testosterone, dehydroepiandrosterone sulfate, and sex hormone– binding globulin levels through the menopause transition. J Clin Endocrinol Metab 2000;85:2832–8. 11. Dennerstein L, Smith AMA, Morse C, Burger HG, Green A, Hopper JL, et al. Menopausal symptoms in Australian women. Med J Austral 1993;159:232–6. 12. Burger HG, Dudley EC, Hopper JL, Shelley J, Green A, Smith AMA, et al. The endocrinology of the menopausal transition: a cross-sectional study of a population-based sample. J Clin Endocrinol Metab 1995;80: 3537–45. 13. Guthrie JR, Ebeling PR, Hopper JL, Barrett-Connor E, Dennerstein L, Dudley EC, et al. A prospective study of bone loss in menopausal Australian-born women. Osteoporos Int 1998;8:282–90. 14. Carey DGP, Nguyen TV, Campbell DJ, Chisholm DJ, Kelly P. Genetic influences on central abdominal fat: a twin study. Int J Obes 1996;20: 722–6. 15. Ross R, Shaw KD, Martel Y, de Giuse J, Avruch L. Adipose tissue distribution measured by magnetic resonance imaging in obese women. Am J Clin Nutr 1993;57:470 –5. 16. Pritchard JE, Nowson CA, Strauss BJ, Carlson JS, Kaymakci B, Wark JD. Evaluation of dual energy X-ray absorptiometry as a method of measurement of body fat. Eur J Clin Nutr 1993;47:216 –28. 17. Guthrie JR. Physical activity: measurement in mid-life women. Acta Obstet Gynecol Scand 2002;81:595– 602. 18. Soules MR, Sherman S, Parrott E, Rebar R, Santoro N, Utian W, et al. Executive summary: stages of reproductive aging workshop (STRAW). Fertil Steril 2001;76:874 –8. 19. SPSS for Windows. 1999 Statistical package for social sciences. Version 9.0. Chicago: SPSS Inc. 20. MacDougal OA, Lane MD. Transcriptional regulation of gene expression during adipocyte differentiation. Ann Rev Biochem 1995;64:345– 73. 21. Rebuffe-Scrive M, Andersson B, Olbe L, Bjorntorp P. Metabolism of adipose tissue in intra-abdominal depots of non-obese men and women. Metabolism 1989;38:453–8.
1339
22. Sowers MF, Beebe JL, McConnell D, Randolph J, Jannausch M. Testosterone concentrations in women aged 25–50 years: associations with lifestyle, body composition, and ovarian status. Am J Epidemiol 2001;153:256 –64. 23. Poehlman ET, Toth MJ, Gardner AW. Changes in energy balance and body composition at menopause. Ann Int Med 1995;123:673–5. 24. Lynch NA, Nicklas BJ, Berman DM, Dennis KE, Goldberg AP. Reductions in visceral fat during weight loss and walking are associated
1340 Guthrie et al.
Abdominal fat and endogenous hormones
with improvements in VO2max. J Appl Physiol 2001;90:99 –104. 25. Matthews KA, Kuller LH, Sutton-Tyrrell K, Chang YF. Changes in cardiovascular risk factors during the perimenopause and postmenopause and carotid artery atherosclerosis in healthy women. Stroke 2001;32:1104 –11. 26. Guthrie JR, Ball M, Dudley EC, Garamszegi CV, Wahlqvist ML, Dennerstein, et al. Impaired fasting glycaemia in middle-aged women: a prospective study. Int J Obes 2001;25:646 –51.
Vol. 79, No. 6, June 2003
Central abdominal fat and endogenous hormones during the menopausal transition Received July 22, 2002; revised and accepted November 14, 2002. Presented in part at the 10th International Menopause Society Congress, Berlin, Germany, June 14, 2002. Supported by the National Health and Medical Research Council of Australia and the Victorian Health Promotion Foundation. During 2001, the research program received unrestricted educational grants from Australia and New Zealand Banking Group (ANZ) Trustees, Eli Lilly (Australia), Pharmacia, and Upjohn (Australia); Prince Henry’s Institute of Medical Research received grants from Organon for hormone assays. Reprint requests: Janet R. Guthrie, Ph.D., Office for Gender and Health, University of Melbourne, Charles Connibere Building, Royal Melbourne Hospital, Victoria 3050, Australia (FAX: 613-93474127; E-mail: janetrg@ unimelb.edu.au). a Office for Gender and Health. b Department of Psychiatry, Monash University, Clayton, Victoria, Australia. c Department of Diabetes and Endocrinology, Royal Melbourne Hospital, Victoria, Australia. d University of Michigan Health System, Ann Arbor, Michigan. e Prince Henry’s Institute of Medical Research, Clayton, Victoria, Australia. f Department of Medicine. 0015-0282/03/$30.00 doi:10.1016/S0015-0282(03) 00361-3
Janet R. Guthrie, Ph.D.,a Lorraine Dennerstein, M.B., Ph.D.,a John R. Taffe, Ph.D.,b Peter R. Ebeling, M.D.,c John F. Randolph, M.D.,d Henry G. Burger, M.D.,e and John D. Wark, M.B., Ph.D.f The University of Melbourne, Victoria, Australia
Objective: To investigate the effect of endogenous hormone levels on central abdominal fat during the menopausal transition in a population-based cohort of Australian-born women. Design: Prospective observational study. Setting: Population-based sample. Body composition was assessed in the Royal Melbourne Hospital, and interviews were conducted at the patient’s home. Subject(s): One hundred two women from the Melbourne Women’s Midlife Health Project. Data, physical measures, and blood were obtained by interview when the longitudinal study commenced (baseline) and at the time of the total body scan approximately 5 years later. Body composition was measured using dual-energy X-ray absorptiometry. Intervention(s): None. Main Outcome Measure(s): Total body fat and central abdominal fat. Result(s): The 102 women were either premenopausal or in the early menopausal transition at baseline. At the time of their dual-energy X-ray absorptiometry scan, 31 were in the early menopausal transition, 22 were in the late menopausal transition, and 49 were postmenopausal. Multiple regression analysis found that total percentage of body fat was associated with weight measures, whereas central abdominal fat was also positively associated with baseline free T index (FTI) and with the increase in FTI since baseline. Conclusion(s): The major hormonal change associated with central adiposity during the menopausal transition is the increase in the FTI. This effect is significant even after allowing for baseline and final weight. (Fertil Steril威 2003;79:1335– 40. ©2003 by American Society for Reproductive Medicine.) Key Words: Central abdominal fat, percent body fat, free testosterone index, estradiol, menopausal transition, weight
Visceral abdominal fat is independently associated with the development of dyslipidemia, hypertension, and hyperinsulinemia in women (1). These metabolic abnormalities increase the risk of developing cardiovascular disease and type 2 diabetes. Visceral abdominal fat drains directly into the portal circulation, releasing free fatty acids and causing more insulin resistance than subcutaneous abdominal fat, which drains into the systemic circulation (2). Several population-based longitudinal studies (3–5) have shown that around the time of menopause, women experience an increase in weight. Cross-sectional studies using dual-energy X-ray absorptiometry (DXA) have reported menopause-related changes in body
composition and body fat distribution (6 – 8), although not all changes were significant (7, 8). There are a number of methodological problems with these studies: some were based on convenience (6) or clinical samples (8) with inherent biases; one defined women as postmenopausal after 6 months of amenorrhea (8), and one excluded obese women (6). Menopause is associated with major hormonal changes, including a large increase in FSH and a profound fall in E2 levels (9), but there are also changes in sex hormone– binding globulin (SHBG) and the free T index (FTI) (10). The role of these hormonal changes in the changes in body fat distribution during the 1335
menopausal transition has not been investigated in an Australian-born population. The Melbourne Women’s Midlife Health Project, a 9-year prospective population-based study of Australianborn women’s experiences of the menopausal transition, provided an opportunity to study variables associated with body composition. Although only one DXA measure of body composition was made, the prospective design of this project allowed baseline measures as well as annual changes to be used to predict body fat distribution. The aim of this paper is to describe the effect of endogenous hormone levels on central abdominal fat in a populationbased cohort of women experiencing the menopausal transition.
MATERIALS AND METHODS The study was approved by the Human Research Ethics Committee of the University of Melbourne, and the procedures followed were in accordance with the ethical standards of the National Health and Medical Research Council. All subjects provided written informed consent for their participation in the study.
Subjects The Melbourne Women’s Midlife Health Project began in 1991 with population sampling by random telephone digital dialing and baseline interviews of 2,001 Australian-born women aged 45–55 years and residing in Melbourne (71% response rate) (11). All of those women at baseline who had experienced menses in the prior 3 months, and who were not taking the oral contraceptive pill or hormone therapy, were invited to participate in a longitudinal study. Of those eligible, 56% accepted (n ⫽ 438). Volunteers for the longitudinal study were more likely than nonparticipants to report better self-rated health, paid employment, more than 12 years of education, having ever had a Papanicolaou smear, exercising at least once a week, and having undergone dilatation and curettage (12). The retention rate by year 8 of follow-up was 88%. The cohort for the current study comprised 139 of these women who had previously participated in a prospective hip and lumbar spine bone mineral density study (13). These women were invited to return for a third bone densitometry visit during the sixth year of follow-up, which included a whole body bone scan.
Body Composition Body composition was measured by DXA using a Hologic QDR-1000W densitometer in the Bone Densitometry Unit, Department of Medicine, Royal Melbourne Hospital, and Hologic Software Version 5.35 (Hologic, Inc., Waltham, MA). Body composition was analyzed according to a twocompartment model, comprising fat mass and lean tissue. The method used to measure central abdominal fat content was similar to that described by Carey et al. (14). 1336 Guthrie et al.
Abdominal fat and endogenous hormones
The central abdominal fat region was manually determined and included the area between the upper surface of the second lumbar vertebra, the lower surface of the fourth lumbar vertebra, and the lateral margins of the outer rib cage. This area has been shown by magnetic resonance imaging to contain a relatively high visceral and low subcutaneous fat content (15). The coefficient of variation for measuring percentage of body fat using the Hologic QDR-1000W densitometer is 1.3% (16). Weight and height were measured as described previously (3). Three measures of lifestyle related behavior were recorded annually: exercise, which was recorded as frequency of engaging in exercise for fitness or recreational purposes (17), current smoking (Yes/No), and alcohol intake (number of drinks per week).
Hormone Measures Fasting morning blood samples for hormone immunoassays were collected at baseline and at the time of the DXA measurements. Blood was taken between days 4 and 8 of the menstrual cycle if the women were cycling. If not cycling, blood was taken after 3 months or more of amenorrhea. E2 was measured using the double-antibody RIA kit purchased from Diagnostic Products Corporation (DPC), Los Angeles. Assay sensitivity (as detection limit) was 20 pmol/L. The interassay coefficient of variation was 6.6% at 400 pmol/L, and the intra-assay coefficient of variation was 11% at 110 pmol/L, 13% at 470 pmol/L, and 13% at 1,160 pmol/L. Serum T was measured by double-antibody RIA after sample extraction and polyethylene glycol– enhanced separation of bound from free ligand, using 125I iodinated T as a tracer. The between-assay coefficient of variation at a T level of 2 nmol/L was 12%. The “normal” female range for serum T using this assay is quoted as 1.0 –2.8 nmol/L. SHBG was measured by an automated chemiluminescent enzyme immunoassay (DPC) using the Immulite Automated Analyser. The intra- and inter-assay coefficients of variation were 4.9% and 5.8%, respectively. FTI was calculated as the ratio of measured T to measured SHBG ⫻ 100.
Menopausal Status Data on menstrual history were collected at the time of interview, and the date of final menstrual period (FMP) was verified from prospectively kept menstrual diaries. Menopausal status was determined at these two time points by menstrual status. Women were defined as premenopausal if they reported no change in menstrual frequency in the prior 12 months. Women who had menstruated in the previous 3 months with changes in menstrual frequency were considered to be in early menopausal transition. Women who reported at least 3 months but less than 12 months of amenorrhea were considered to be in late menopausal transition. Women were considered postmenopausal if they had had amenorrhea for at least 12 months (18). Vol. 79, No. 6, June 2003
TABLE 1 Body composition variables at baseline and at time of dual-energy X-ray absorptiometry (DXA) scan in the whole cohort (n ⫽ 102). Continuous variables are reported as mean (SD). Variables Weight (kg) BMI ⱖ 30kg/m2 (%) Total fat mass (kg) Total lean mass (kg) Total body fat (%) Central abdominal fat mass (kg) Fat in central abdomen (%)
Baseline
Time of DXA scan
68.6 (14.4) 16.7 – – – – –
69.1 (13.8) 17.6 24.3 (10.9) 44.1 (5.0) 33.0 (7.7) 1.39 (0.81) 33.7 (10.6)
Note: BMI ⫽ Body mass index (weight [kg]/height [m]2). Guthrie. Abdominal fat and endogenous hormones. Fertil Steril 2003.
Statistical Analysis Because of skewed distributions, hormone values were log transformed before analysis. Multiple linear regression analysis was used to determine which variables independently predicted the body composition measures. Analysis of variance was used to examine differences in body composition among menopausal status groups. Independent t-tests were used to examine the differences in body composition depending on exercise, smoking, or alcohol intake habits. Unless otherwise noted, a nominal level of significance of P⬍.05 was used and all statistical tests were two tailed. SPSS software (19) was used for all analyses.
RESULTS At baseline, all 139 women who participated in the study were premenopausal (n ⫽ 48) or in the early menopausal
transition (n ⫽ 54). At the time of their DXA scans, 31 of these women were in the early menopausal transition, 22 were in the late menopausal transition, and 49 were naturally postmenopausal and were on average 34.0 months (SD ⫾ 12.2 months) past FMP. Twenty-eight women were taking or had taken hormone therapy, and nine had experienced surgical menopause. Hormone users and those with a surgical menopause were excluded from the analyses, leaving a cohort of 102 women with a mean age of 48.9 years (SD ⫾ 2.4 years) at baseline. Although only one DXA measure of body composition was obtained, we were able to use the women’s prospectively acquired data since baseline as independent variables. At baseline, 20% of the cohort were smokers, 17% consumed eight or more alcoholic drinks per week, and 61% participated in exercise for fitness or recreational purposes more than once a week. At the time of their DXA scans, 5 years later, these percentages were 15%, 17%, and 75% respectively. Table 1 shows the DXA scan measures and weight and body mass index (BMI) information of the cohort at baseline and at the time of their DXA scans. Table 2 shows the levels of FSH and E2 and the FTI at baseline and at the time of the DXA scan by menopausal status. At the time of the DXA scan, women in the early menopausal transition had significantly higher E2 levels compared with women who were in the late menopausal transition (P⬍.05) or postmenopausal (P⬍.0005), and FSH levels were significantly lower in the early menopausal transition compared with the late menopausal transition and postmenopause (P⬍.001 in both cases). There was no significant difference among the menopausal transition groups at the time of the DXA scan with respect to FTI values. Overall, the FTI levels at the time of the scan
TABLE 2 Hormone levels (geometric mean, 95% confidence interval) at baseline and at time of dual-energy X-ray absorptiometry (DXA) scan by menopausal status. Baseline
Time of DXA
Hormone
Menopausal status
Hormone levels
FSH (IU/L)
Premenopausal Early MT
E2 (pmol/L)
Premenopausal Early MT
259.1 (200.6–334.8) 245.9 (171.4–352.8)
Free T Index
Premenopausal Early MT
2.03 (1.73–2.37) 1.58 (1.34–1.87)
7.7 (6.4–9.1) 13.2 (9.9–17.5)
Menopausal status Early MT Late MT Postmenopausal Early MT Late MT Postmenopausal Early MT Late MT Postmenopausal
Hormone levels 25.3 (16.6–38.6) 93.7 (78.1–112.4) 111.8 (100.9–123.9) 116.4 (71.6–189.3) 58.1 (34.3–98.2) 23.5 (21.1–26.1) 2.78 (2.23–3.46) 2.75 (2.14–3.54) 2.87 (2.45–3.37)
Note: The cohort was comprised of 102 women: at baseline 48 were premenopausal and 54 were in the early menopausal transition; at the time of the DXA scan, 31 were in the early menopausal transition, 22 in the late menopausal transition, and 49 in postmenopause. MT ⫽ menopausal transition. Guthrie. Abdominal fat and endogenous hormones. Fertil Steril 2003.
FERTILITY & STERILITY威
1337
TABLE 3 Multiple linear regression with dual-energy X-ray absorptiometry measure of central abdominal fat as dependent variable. Independent variables
Dependent variable % Fat in central abdomen
Baseline weight Change in weight Baseline FTI Change in FTI E2 Change in E2 Menopausal status Age

P
⫹0.5 ⫹0.7 ⫹4.7 ⫹3.7 ⫹0.5 ⫺0.7 ⫹1.7 ⫹0.2
⬍.001 .001 .02 .04 .6 .4 .5 .7
Note: FTI ⫽ free T index. Guthrie. Abdominal fat and endogenous hormones. Fertil Steril 2003.
were significantly higher than baseline FTI values, as were FSH levels, and E2 levels were significantly lower (P⬍.001). Women who were in the early menopausal transition had a lower percentage of abdominal fat compared with women in the late menopausal transition or in postmenopause. The values were 31.4% (SD ⫾ 1.8), 37.3% (SD ⫾ 2.4), and 33.5% (SD ⫾ 1.5) for the early menopausal transition, late menopausal transition, and postmenopause, respectively. The differences between the early and late menopausal transition groups were just significantly different (P ⫽ .05). Multiple regression analysis found that total body fat and central abdominal fat measures were both positively associated with weight measured at baseline (P⬍.0001), which was on average 5 years before the body composition measurements when all members of the cohort were still menstruating, and with change in weight over this period (P ⫽ .001) after controlling for age, E2 levels, and menopausal status. Central abdominal fat was also positively associated with baseline FTI and with a change in FTI since baseline (P⬍.05) (Table 3). Women in the sample who currently participated in physical activities for fitness or recreational purposes more than once a week tended to have lower central abdominal and total percent body fat than women who exercised once a week or less (32.9%, SD ⫾ 11.1 vs. 36.5%, SD ⫾ 8.7 for central abdominal fat; 32.5% SD ⫾ 7.4, vs. 34.8%, SD ⫾ 8.7 for total percent body fat), but the differences were not statistically significant. There was no significant association between fat measures and age, alcohol intake, or smoking in this cohort.
age and E2 levels. Although serum total T levels appear to remain unchanged during the menopausal transition (14), SHBG has been shown to decrease during this period (14). The time of maximum change is estimated at 2 years before the FMP, and consequently the FTI increases. SHBG levels are also negatively associated with BMI and positively correlated with E2 levels. About one-third of the decline in SHBG was explicable by BMI and changes in estrogen levels (14). This decrease in SHBG and subsequent increase in free T may act directly on fat stores or may modify expression of genes controlling adipocyte differentiation and metabolism. Changes at the cellular level, such as a decrease in lipoprotein lipase activity in femoral adipose tissue but not in abdominal adipose tissue, have been reported after menopause (20, 21). This could explain a shift of adipose accumulation toward the abdominal area. The exact mechanism underlying the association of FTI and other hormones with the distribution of body fat is yet to be elucidated. However, this androgen-influenced storage of fat in the intra-abdominal area may well predispose women to a greater risk for cardiovascular disease and diabetes (1) and has important health implications for postmenopausal women. Total percent body fat and waist-to-hip ratios have previously been reported to be positively associated with T concentrations in pre- and perimenopausal women by Sowers and colleagues (22). The measure used in this latter study was total rather than free, bioavailable T. In our cohort, total percent body fat was positively associated with baseline weight and change in weight during follow-up. This observation is to be expected, as weight is primarily related to fat and lean body mass and in midlife women, the increase in weight is likely to be an increase in fat mass rather than an increase in muscle and bone mass (23). It was only the percent fat in the abdominal region that was associated with free T levels, after allowing for weight and weight changes. Our cohort was predominately postmenopausal at the time of the DXA measures, and this may explain the differences in the findings for total percent body fat between the Sowers study (22) and ours.
DISCUSSION
Our study and others (7, 8) have used DXA to measure abdominal fat content. DXA has been shown to be a highprecision method for measuring body composition and in particular has demonstrated a smaller variability in percentage fat values compared with other methods (16). Some studies (6) have used computed tomography (CT). DXA measurements have been shown to compare well with those from CT. We positioned our measurement of central abdominal fat on an area shown by magnetic resonance imaging to contain a relatively high content of visceral fat and lowsubcutaneous fat (15).
Our results suggest that sex hormone concentrations affect fat distribution. We found an association between a measure of central body fatness and the FTI, independent of
CT measures do allow the independent measurement of abdominal subcutaneous and intra-abdominal fat. Toth et al. (6) found, using CT, that intra-abdominal fat but not abdom-
1338 Guthrie et al.
Abdominal fat and endogenous hormones
Vol. 79, No. 6, June 2003
inal subcutaneous fat was greater in early postmenopausal compared with premenopausal women, the age range of the sample being comparable with ours. The mean values for fat mass were lower in Toth’s study of American women (6) compared with our Australian population, and they showed a large difference in body fat between pre- and postmenopausal women. This may be explained by the exclusion from the American study of women with a BMI ⱖ30 kg/m2 rather than a difference in methods of measurement. Our cohort was population based, and 17% of the women had a BMI ⱖ30 kg/m2. The increase in fat mass in middle-aged women is most probably due to a combination of behavioral and physiological mechanisms, including a decrease in energy expenditure, an increase in energy intake, and a decrease in metabolic rate (23). We found a trend for lower values of total body fat and abdominal fat in women who participated in regular exercise. There was no evidence of preferential reduction in abdominal fat in response to regular exercise. Although a number of studies have reported that exercise is associated with reductions in abdominal fat, others have not found an effect (15). The review by Ross and Janssen (15) found an absence in the literature of randomized controlled trials that employed imaging techniques to determine changes in abdominal fat. Abdominal obesity was measured by waist circumference in most trials (15). A recent paper (24) using DXA and CT scans reported that visceral adipose tissue decreased in obese postmenopausal women who increased their maximal oxygen uptake during a 6-month walking program. A genetic influence on central abdominal fat levels has been reported in a study of nonobese Caucasian monozygotic and dizygotic female twins (14). The investigators reported that the majority of intersubject variance in central abdominal fat was due to genetic factors. This genetic influence on abdominal fat was present after adjustment for genetic effects due to total fat, suggesting the existence of genes specific for abdominal obesity. We did not have the relevant information to identify women in our study whose abdominal fat levels may have been influenced by genetic factors. Our finding of the association between body composition and baseline weight is similar to findings in other studies on predictors of developing carotid atherosclerosis (25) and impaired fasting glucose (26). These studies all indicate that premenopausal women should be counseled on lifestyle changes that might prevent adverse health outcomes in postmenopause. In conclusion, results from this study indicate that the major hormonal change associated with central adiposity during the menopausal transition is an increase in the FTI. FERTILITY & STERILITY威
This effect is significant even after allowing for baseline and final weight.
Acknowledgments: The authors thank Sue Kantor, who performed the DXA measurements for this study, and Mr. N. Balazs and his staff in the Department of Biochemistry at the Monash Medical Centre for the hormone assays. We also thank all the women who participated in this study.
References 1. Despres J-P, Moorjani S, Lupien PJ, Tremblay A, Nadeau A, Bouchard C. Regional distribution of body fat, plasma lipoproteins, and cardiovascular disease. Atherosclerosis 1990;10:497–511. 2. Yamashita S, Nakamura T, Shimimura I, Nishida M, Yoshida S, Kotani K, et al. Insulin resistance and body fat distribution. Diabetes Care 1996;19:287–91. 3. Guthrie JR, Dennerstein L, Dudley EC. Weight gain and the menopause: a 5-year prospective study. Climacteric 1999;2:205–11. 4. Blumel JE, Casrelo-Branco C, Rocangliolo ME, Bifa L, Tacla X, Mamani L. Changes in body mass index around menopause: a population study of Chilean women. Menopause 2001;8:239 –44. 5. Wing RR, Matthews KA, Kuller LH, Meilahn EN. Weight gain at the time of the menopause. Arch Intern Med 1991;151:97–102. 6. Toth MJ, Tchernof A, Sites CK, Poehlman ET. Effect of menopausal status on body composition and abdominal fat distribution. Int J Obes 2000;24:226 –31. 7. Svendsen OL, Hassager C, Christiansen C. Age- and menopauseassociated variations in body composition and fat distribution in healthy women as measured by dual-energy x-ray absorptiometry. Metabolism 1995;44:369 –73. 8. Tremollieres FA, Pouilles J-M, Ribot CA. Relative influence of age and menopause on total and regional body composition changes in postmenopausal women. Am J Obstet Gynecol 1996;175:1594 –600. 9. Burger HG, Dudley EC, Hopper JL, Groome N, Guthrie JG, Green A, et al. Prospectively measured levels of serum follicle stimulating hormone, estradiol, and the dimeric inhibins during the menopausal transition in a population-based cohort of women. J Clin Endocrinol Metab 1999;84:4025–30. 10. Burger HG, Dudley EC, Cui J, Dennerstein L, Hopper JL. A prospective longitudinal study of serum testosterone, dehydroepiandrosterone sulfate, and sex hormone– binding globulin levels through the menopause transition. J Clin Endocrinol Metab 2000;85:2832–8. 11. Dennerstein L, Smith AMA, Morse C, Burger HG, Green A, Hopper JL, et al. Menopausal symptoms in Australian women. Med J Austral 1993;159:232–6. 12. Burger HG, Dudley EC, Hopper JL, Shelley J, Green A, Smith AMA, et al. The endocrinology of the menopausal transition: a cross-sectional study of a population-based sample. J Clin Endocrinol Metab 1995;80: 3537–45. 13. Guthrie JR, Ebeling PR, Hopper JL, Barrett-Connor E, Dennerstein L, Dudley EC, et al. A prospective study of bone loss in menopausal Australian-born women. Osteoporos Int 1998;8:282–90. 14. Carey DGP, Nguyen TV, Campbell DJ, Chisholm DJ, Kelly P. Genetic influences on central abdominal fat: a twin study. Int J Obes 1996;20: 722–6. 15. Ross R, Shaw KD, Martel Y, de Giuse J, Avruch L. Adipose tissue distribution measured by magnetic resonance imaging in obese women. Am J Clin Nutr 1993;57:470 –5. 16. Pritchard JE, Nowson CA, Strauss BJ, Carlson JS, Kaymakci B, Wark JD. Evaluation of dual energy X-ray absorptiometry as a method of measurement of body fat. Eur J Clin Nutr 1993;47:216 –28. 17. Guthrie JR. Physical activity: measurement in mid-life women. Acta Obstet Gynecol Scand 2002;81:595– 602. 18. Soules MR, Sherman S, Parrott E, Rebar R, Santoro N, Utian W, et al. Executive summary: stages of reproductive aging workshop (STRAW). Fertil Steril 2001;76:874 –8. 19. SPSS for Windows. 1999 Statistical package for social sciences. Version 9.0. Chicago: SPSS Inc. 20. MacDougal OA, Lane MD. Transcriptional regulation of gene expression during adipocyte differentiation. Ann Rev Biochem 1995;64:345– 73. 21. Rebuffe-Scrive M, Andersson B, Olbe L, Bjorntorp P. Metabolism of adipose tissue in intra-abdominal depots of non-obese men and women. Metabolism 1989;38:453–8.
1339
22. Sowers MF, Beebe JL, McConnell D, Randolph J, Jannausch M. Testosterone concentrations in women aged 25–50 years: associations with lifestyle, body composition, and ovarian status. Am J Epidemiol 2001;153:256 –64. 23. Poehlman ET, Toth MJ, Gardner AW. Changes in energy balance and body composition at menopause. Ann Int Med 1995;123:673–5. 24. Lynch NA, Nicklas BJ, Berman DM, Dennis KE, Goldberg AP. Reductions in visceral fat during weight loss and walking are associated
1340 Guthrie et al.
Abdominal fat and endogenous hormones
with improvements in VO2max. J Appl Physiol 2001;90:99 –104. 25. Matthews KA, Kuller LH, Sutton-Tyrrell K, Chang YF. Changes in cardiovascular risk factors during the perimenopause and postmenopause and carotid artery atherosclerosis in healthy women. Stroke 2001;32:1104 –11. 26. Guthrie JR, Ball M, Dudley EC, Garamszegi CV, Wahlqvist ML, Dennerstein, et al. Impaired fasting glycaemia in middle-aged women: a prospective study. Int J Obes 2001;25:646 –51.
Vol. 79, No. 6, June 2003