The Effect of Hormone Replacement Therapy on Body Composition ...

29 downloads 2711 Views 517KB Size Report
Feb 1, 2005 - fat, sc abdominal fat, total fat, percent fat, fat-free mass, and weight did .... per fat-free mass also decreased with HRT (Fig. 3B). ..... E-mail: [email protected]. ... Dixon WJ 1988 BMDP statistical software manual, volume 2.
0021-972X/05/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 90(5):2701–2707 Copyright © 2005 by The Endocrine Society doi: 10.1210/jc.2004-1479

The Effect of Hormone Replacement Therapy on Body Composition, Body Fat Distribution, and Insulin Sensitivity in Menopausal Women: A Randomized, Double-Blind, Placebo-Controlled Trial Cynthia K. Sites, Georgia D. L’Hommedieu, Michael J. Toth, Martin Brochu, Brian C. Cooper, and Penny A. Fairhurst Departments of Obstetrics and Gynecology (C.K.S., G.D.L., B.C.C., P.A.F.) and Medicine (M.J.T., M.B.), University of Vermont, Burlington, Vermont 05405 Purpose: After menopause, women gain abdominal fat and become less sensitive to insulin. We sought to determine whether hormone replacement therapy (HRT) reduced intraabdominal and sc abdominal fat and improved insulin sensitivity in early menopausal women. Methods: Seventy-six postmenopausal women, age 51.6 ⴞ 3.9 yr with body mass index of 24.9 ⴞ 3.2 kg/m2, were randomized to conjugated estrogens (0.625 mg) plus medroxyprogesterone acetate (2.5 mg) or placebo daily. Women received a computed tomography scan at the L4 –L5 vertebral disk space, a dual x-ray absorptiometry scan, and a euglycemic hyperinsulinemic clamp at baseline, 6 months, 1 yr, and 2 yr. Results: Fifty-one women completed the trial and were analyzed (n ⴝ 26 on HRT and n ⴝ 25 on placebo). Intraabdominal

T

HE MENOPAUSE TRANSITION, with its subsequent decrease in ovarian estrogen and progesterone, is associated with an increased risk of cardiovascular disease. Part of this risk may be explained by changes in body composition, body fat distribution, and insulin sensitivity (1– 4). During menopause, women gain body fat preferentially in the abdominal region and have decreased insulin sensitivity (3, 4). These metabolic changes, along with dyslipidemia and hypertension which increase after menopause, are consistent with the metabolic syndrome and may predict type 2 diabetes mellitus (5) and coronary artery disease (6) in postmenopausal women. The population is aging, and the number of postmenopausal women will increase dramatically over the next decades. Consequently, strategies that improve the metabolic profile could also be beneficial in preventing diabetes mellitus and cardiovascular disease. We hypothesized that combined estrogen plus progestin replacement therapy would reduce the central accumulation of body fat and thus improve insulin sensitivity in postmenopausal women. Previous studies have focused on short-term treatment with hormone replacement therapy (HRT) and its effects on body First Published Online February 1, 2005 Abbreviations: CT, Computed tomography; DXA, dual x-ray absorptiometry; HRT, hormone replacement therapy. JCEM is published monthly by The Endocrine Society (http://www. endo-society.org), the foremost professional society serving the endocrine community.

fat, sc abdominal fat, total fat, percent fat, fat-free mass, and weight did not differ between treatment groups by time. Insulin sensitivity did not change in the placebo group, but decreased by 17% in the HRT group by 6 months and persisted at 2 yr (P < 0.01 for treatment by time effect). One year after the trial, insulin sensitivity increased by 25% in women who had taken HRT (P ⴝ 0.006 for treatment by time effect), to a level similar to those women in the placebo group. Conclusions: Conjugated estrogens plus medroxyprogesterone acetate reduce insulin sensitivity in menopausal women without affecting body composition or body fat distribution. The reduction in insulin sensitivity is reversible after discontinuing HRT. (J Clin Endocrinol Metab 90: 2701–2707, 2005)

composition (7–9) and indirect measures of insulin sensitivity (10, 11). To our knowledge, no long-term, randomized, placebo-controlled trial has examined the effects of a common HRT regimen available in the United States on the temporal changes in body fat distribution and insulin sensitivity in postmenopausal women. Furthermore, no trial has examined the discontinuation of HRT to determine whether these metabolic changes persist. Subjects and Methods Participants One hundred five women were recruited by newspaper advertisement and screened for eligibility by phone from the Burlington, Vermont, community (Fig. 1). Twenty-nine women were excluded due to failure to meet the inclusion criteria or refusal to participate. Seventy-six perimenopausal and early postmenopausal women, aged 51.6 ⫾ 3.9 yr (mean ⫾ sd) and from 6 months to 5 yr since last menstrual period, were randomized by block design to oral conjugated estrogens 0.625 mg plus medroxyprogesterone acetate 2.5 mg (n ⫽ 40) or placebo (n ⫽ 36) (Prempro; donated by Wyeth Ayerst Pharmaceuticals, Philadelphia, PA). Of these subjects, 74 were Caucasian, 1 was Abenaki Indian, and 1 was Asian. Fifty-one women completed the 2-yr trial (n ⫽ 25 on HRT, n ⫽ 26 on placebo). At the conclusion of the trial, all women discontinued HRT or placebo and were asked to consider returning in 2 yr for an additional visit. Seventeen women agreed to do so (n ⫽ 9 in HRT group and n ⫽ 8 in placebo group). Postmenopausal status was defined by the absence of menses for less than 5 yr, with an FSH level more than 30 IU/liter. Other inclusion criteria included a body mass index less than 30 kg/m2, fasting glucose less than 112 mg/dl (6.22 mmol/liter), and waist circumference less than 94 cm. These inclusion criteria were chosen because we were interested

2701

2702

J Clin Endocrinol Metab, May 2005, 90(5):2701–2707

Sites et al. • HRT, Fat, and Insulin Sensitivity

where Einf is the enrichment of [6,6 D2]glucose (mole percent excess) in the glucose infusate, Ep is the enrichment of [6,6 D2]glucose in plasma, Edex is the enrichment of [6,6 D2]glucose in the dextrose infusion, and i is the infusion rate of [6,6 D2]glucose tracer (milligrams per minute) from the inf and dex infusions, respectively (16). Any negative values for glucose Ra during hyperinsulinemia were assumed to indicate complete suppression of endogenous glucose production (i.e. 0 mg/min). Ra under euglycemic hyperinsulinemic conditions was subtracted from Ra data under postabsorptive conditions as an indicator of insulin-induced suppression of endogenous glucose production. Body composition. Fat mass, percent fat, and fat-free mass were measured by dual-photon x-ray absorptiometry (DXA) using a Lunar DPX-L densitometer (Lunar Corp., Madison, WI) as reported previously (13, 15). Scans were analyzed using the Lunar version 1.3y DPX-L extended analysis program for body composition. The coefficient of variation for repeated measurements in our laboratory is 1% for fat mass. Computed tomography (CT). Intraabdominal fat and sc abdominal fat were measured at the L4 –L5 vertebral disk space at an attenuation range of ⫺190 to ⫺30 Hounsfield units by CT using a GE High Speed Advantage scanner (General Electric Medical Systems, Milwaukee, WI) (13, 15). Sagittal diameter was measured as the anterior-to-posterior distance in millimeters at this disk level. The within-subject variation for repeated analysis of fat measurements in our laboratory is less than 1%. Biochemical analysis. Plasma glucose concentrations were measured by the glucose oxidase method with an automated glucose analyzer (YSI Instruments, Yellow Springs, OH). Serum insulin concentrations were determined with a double-antibody RIA (Diagnostic Products Corp., Los Angeles, CA). Serum FSH was measured with a chemiluminescent assay (Bayer Diagnostics, Tarrytown, NY). Inter- and intraassay coefficients of variation for insulin were 10 and 4%, respectively, and for FSH were 2.2 and 0.3%, respectively.

FIG. 1. Enrollment of women in the study. in examining the effect of HRT on disease prevention in healthy, nonobese, postmenopausal women. Waist circumferences greater than 94 cm in women have been associated with a greater likelihood of metabolic abnormalities, including diabetes mellitus. The study was approved by the Institutional Review Board at the University of Vermont, and all women signed statements giving their informed consent.

Interventions

Food intake. Study volunteers recorded food intake for 3 d at each time point to determine whether HRT affected body composition by affecting self-reported food intake. Diaries were analyzed by computer in the General Clinical Research Center using the Food Intake Analysis System program. Total calories, carbohydrate, fat, and protein intake were assessed (17, 18). Physical activity. Leisure time physical activity at each time point was measured using the Minnesota Leisure Time Physical Activity Questionnaire (19). This validated, standardized, 60-question survey quantifies various activities performed over the previous year based on intensity of the exercise and time performed, and provides reliable group means for assessing physical activity (20). Test-retest correlations of this questionnaire are 0.79 – 0.88 for total activity (21).

Statistical analysis

Insulin-stimulated glucose disposal. We performed euglycemic hyperinsulinemic clamps according to the method of DeFronzo et al. (12), with modifications as previously described by our group (13–15). After 3 d of standardized meals (55% carbohydrate, 30% fat, and 15% protein), volunteers were tested after a 12-h overnight fast. An iv catheter was placed in an antecubital vein for the infusion of insulin, [6,6 D2]glucose, and 20% dextrose. A second catheter was placed in the volunteer’s contralateral hand and kept in a hot box (60 C) for sampling of arterialized blood samples. A primed (60:1) continuous (2 mg/kg䡠h) infusion of [6,6 D2]glucose was started at 0700 h. A constant infusion of insulin (40 mU/min per m2) was started at 0900 h to approximate postprandial insulin levels. A 20% dextrose solution enriched with [6,6 D2]glucose was also started at 0900 h. Plasma glucose levels were measured every 5 min during the insulin infusion to adjust the dextrose infusion and maintain fasting glucose levels. During hyperinsulinemia, steady-state was achieved for plasma glucose concentrations and glucose enrichment. Thus, the rate of insulinstimulated glucose disposal corresponded to the removal of glucose from two sources: the mean dextrose infusion used to maintain euglycemia during the last 30 min of the clamp (milligrams per minute), and the residual endogenous glucose production (Ra). Ra was calculated as:

Ra ⫽ 关共Einf ⫺ Ep兲iinf ⫹ 共Edex ⫺ EP兲idex兴/Ep

Baseline clinical variables were analyzed using Fisher’s Exact test for categorical variables, Student’s t test for continuous variables, and the Wilcoxon test for nonparametric variables (22). Fifty-one subjects who completed the study were examined by intention to treat, and were those considered for statistical analyses. According to pharmacy records, subjects in the HRT group took 86.9 ⫾ 6.7% of their study medication, and subjects in the placebo group took 86.6 ⫾ 6.9% of their study medication (means ⫾ sd). For analyses of variables over the four time periods, we used a repeated-measures analysis with orthogonal decomposition and baseline values as covariate to determine whether effects differed over time, and whether the two treatment groups differed (23). Baseline fasting glucose levels, baseline serum insulin levels during the last 30 min of the clamp, and the serum insulin levels during the last 30 min of the clamp at the corresponding time point were included as additional covariates when analyzing insulin sensitivity. In all cases, treatment effects were considered statistically significant at P ⬍ 0.05.

Results

Demographic characteristics of study volunteers at baseline are shown in Table 1. Placebo and HRT groups were not significantly different with regard to any baseline variable

Sites et al. • HRT, Fat, and Insulin Sensitivity

J Clin Endocrinol Metab, May 2005, 90(5):2701–2707

TABLE 1. Baseline characteristics of study participants HRT group (n ⫽ 26)

Placebo group (n ⫽ 25)

Age (yr) FSH (IU/liter) Waist circumference (cm) Body mass index (kg/m2) Weight (kg) Fasting glucose (mg/dl)

50.9 (3.7) 79.1 (38.7) 79.5 (8.1) 24.8 (3.5) 64.3 (9.0) 79.4 (4.6)a

51.6 (3.6) 67.5 (24.0) 82.2 (6.7) 24.7 (2.8) 66.6 (8.7) 82.8 (5.6)a

Fasting insulin (␮U/ml) Alcohol (⬎7 drinks/wk) (%) Cigarette smoking (% yes) Family history of diabetes mellitus (%) Leisure time activity (kcal/d)

12.0 (6.0) 13.0 7.7 69.2

9.4 (2.6) 13.0 8.0 56.0

390.6 (244.4)

360 (289.3)

Characteristic

Data are presented as means (SD). Conversion factors (metric to SI units): glucose ⫽ 0.05551, insulin ⫽ 7.175. a Differences between HRT and placebo groups (P ⫽ 0.02).

except for fasting glucose (79.4 mg/dl or 4.4 mmol/liter in the HRT group vs. 82.8 mg/dl or 4.6 mmol/liter in the placebo group, both within normal limits). Clamp and body composition variables at baseline are shown in Table 2. Groups were not significantly different in any of these variables at baseline (Table 2). Body composition and body fat distribution with placebo and HRT are shown in Fig. 2. There were no differences between groups during the treatment period in intraabdominal fat (P ⫽ 0.15), sc abdominal fat (P ⫽ 0.20), total fat (P ⫽ 0.54), or in fat-free mass (P ⫽ 0.13) (Fig. 2). Furthermore, we found no differences between groups during the treatment period in percent fat (P ⫽ 0.62) or sagittal diameter (P ⫽ 0.46) (data not shown). Waist circumference did not differ between groups during the treatment period (82.8 ⫾ 6.8 to 84.9 ⫾ 7.9 cm in the placebo group, and 79.5 ⫾ 8.1 to 80.6 ⫾ 8.4 cm in the HRT group, P ⫽ 0.42). Similarly, there were no differences in weight between groups during the treatment period (66.9 ⫾ 8.8 to 67.4 ⫾ 9.5 kg in the placebo group, and from 64.6 ⫾ 9.1 to 64.9 ⫾ 8.9 kg in the HRT group, P ⫽ 0.79). The effect of HRT on insulin sensitivity is shown in Fig. 3. Insulin sensitivity declined by 17% in the HRT group over 2 yr, with no significant change in the placebo group (P ⬍ 0.01 for differences between groups; Fig. 3A). Insulin sensitivity per fat-free mass also decreased with HRT (Fig. 3B). The decrease in insulin sensitivity was present at 6 months and remained significant at 24 months. HRT reduced insulin sensitivity compared with placebo by 14% more at 6 months, 15% more at 12 months, and 17% more at 24 months (Fig. 3C). Fasting endogenous glucose production before hyperinsuTABLE 2. Clamp and body composition variables at baseline Variable

HRT group (n ⫽ 26)

Placebo group (n ⫽ 25)

Glucose disposal rate (mg/min) Intraabdominal fat (cm2) SC abdominal fat (cm2) Sagittal diameter (mm) Total fat (kg) Percent fat Lean body mass (kg)

418 (131) 94.1 (38.6) 261.7 (90.2) 215.7 (25.9) 22.5 (7.2) 34.4 (6.9) 39.3 (3.7)

403 (103) 84.8 (31.7) 288.9 (73.2) 214.5 (20.9) 23.4 (6.0) 34.8 (5.1) 40.4 (3.2)

Data are presented as means (SD). There were no significant differences between groups.

2703

linemia was induced during the clamp increased in both HRT and placebo groups over the 2 yr study (P ⫽ 0.02), but was not different between groups (P ⫽ 0.31, data not shown). The suppression of endogenous glucose production by hyperinsulinemia during the clamp was not different between groups over time (Fig. 3D). Means of serum glucose (⫾sd) during the last 30 min of the clamp at each time point were: 79.2 ⫾ 3.5, 78.1 ⫾ 4.1, 78.0 ⫾ 5.1, and 78.8 ⫾ 5.4 mg/dl for the HRT group (4.4 ⫾ 0.2, 4.3 ⫾ 0.2, 4.3 ⫾ 0.3, 4.3 ⫾ 0.3 mmol/liter). Means of serum glucose (⫾sd) during the last 30 min of the clamp for the placebo group at each time point were: 83.7 ⫾ 6.2, 81.4 ⫾ 5.6, 79.8 ⫾ 4.5, and 80.1 ⫾ 5.8 mg/dl (4.6 ⫾ 0.3, 4.5 ⫾ 0.3, 4.4 ⫾ 0.2, 4.4 ⫾ 0.3 mmol/liter). Means of serum insulin during the last 30 min of the clamp for the HRT group were: 86.4 ⫾ 24.1, 74.1 ⫾ 23.7, 70.0 ⫾ 17.9, and 65.0 ⫾ 8.1 ␮IU/ml (619.9 ⫾ 172.9, 531.7 ⫾ 170.0, 502.3 ⫾ 128.4, and 466.4 ⫾ 58.1 pmol/liter). For the placebo group, means of serum insulin during the last 30 min of the clamp were: 80.6 ⫾ 26.2, 76.6 ⫾ 19.0, 72.2 ⫾ 11.2, and 69.5 ⫾ 11.1 ␮IU/ml (578.3 ⫾ 188.0, 549.6 ⫾ 136.3, 518.0 ⫾ 80.4, and 498.7 ⫾ 79.6 pmol/liter). Self-reported food intake during the study was limited to 28 subjects who returned both baseline and 2-yr diaries (14 on HRT and 14 on placebo). During the study, women on HRT reported a 12% decrease in caloric intake (212 kcal), with no change reported in the placebo group (P ⫽ 0.42 for differences between groups). Specifically, self-reported carbohydrate intake was decreased by 16% (38.4 ⫾ 61.4 g/d) in the HRT group at 2 yr compared with a 1% increase (2.7 ⫾ 72.7 g/d) in the placebo group (P ⫽ 0.07 for differences between groups). Self-reported protein intake decreased by 7.6 ⫾ 16.6 g/d in the HRT group compared with a 2.2 ⫾ 21.7 g/d increase in the placebo group at 2 yr (P ⫽ 0.09 for difference between groups). No changes in caloric intake between groups reached statistical significance. Leisure time physical activity during the study did not differ between groups (n ⫽ 25 on HRT and n ⫽ 23 on placebo; decreases of 57 kcal/d in HRT and 91 kcal/d in placebo groups, P ⫽ 0.69). At the conclusion of the 2-yr study, all women discontinued HRT or placebo, and were asked to consider returning in 1 yr for an additional euglycemic hyperinsulinemic clamp and DXA scan. Seventeen women agreed to do so (n ⫽ 9 in HRT group and n ⫽ 8 in placebo group). In these women, the decrease in insulin sensitivity with HRT (at 24 months) completely reversed at 1 yr after HRT discontinuation (at 36 months, P ⫽ 0.006 for treatment by time effect) (Fig. 4). Discussion

We report that continuous combined oral estrogen plus progestin therapy decreases insulin sensitivity in early menopausal women without affecting body composition or body fat distribution. The reduction in insulin sensitivity occurs within 6 months, persists during treatment at 24 months, but is reversed at 1 yr after discontinuing treatment. To our knowledge, this is the first report of a long-term randomized trial of the effect of a common HRT regimen used in the United States on body composition, body fat distribution, and insulin sensitivity using criterion measures (CT scan, DXA, and euglycemic clamp). Furthermore, ours is

2704

J Clin Endocrinol Metab, May 2005, 90(5):2701–2707

Sites et al. • HRT, Fat, and Insulin Sensitivity

FIG. 2. Effect of HRT on body composition and body fat distribution in women randomized to placebo (f) or HRT (Œ). Values are means ⫾ SEM, with baseline values adjusted for repeated measures ANOVA with baseline as covariate. There were no significant differences between groups over time in intraabdominal fat (A, n ⫽ 23 in placebo group and n ⫽ 21 in HRT group), sc abdominal fat (B, n ⫽ 23 in placebo group and n ⫽ 21 in HRT group), total fat (C, n ⫽ 22 in placebo group and n ⫽ 25 in HRT group), or fat-free mass (D, n ⫽ 22 in placebo group and n ⫽ 25 in HRT group).

the first study to report the effect of HRT discontinuation on these outcome variables. The effect of combined estrogen plus progestin replacement therapy to decrease insulin sensitivity in menopausal women might be explained by the use of a progestin. Specifically, no studies report that estrogen alone decreases insulin sensitivity (24, 25). However, studies in which estrogen plus progestin were employed either show a decrease (26) or no change (27–30) in insulin sensitivity. It appears that insulin sensitivity is decreased when estrogen and progestin are taken together daily (26), but is unaffected if progestin is taken only for a few days every 3 months along with daily estrogen (29), or if insulin sensitivity is measured during the estrogen alone phase of a sequential estrogen plus progestin regimen (30). Thus, sequential regimens that use a progestin infrequently to induce a withdrawal bleed may be preferable to continuous estrogen plus progestin regimens to minimize changes in insulin sensitivity. Results of these prior studies should be interpreted with caution, however, since some were of short-term duration (⬍7 months) (24 –26, 30), did not include a control group or use a clamp to measure insulin sensitivity (25), and included open-label (29) or singleblinded (30) study designs. The route of administration of HRT may also contribute to the decrease in insulin sensitivity. When transdermal estrogen is employed with an oral progestin, insulin sensitivity appears to be unaffected (27, 28). Furthermore, transdermal

estrogen alone has no effect on insulin sensitivity (24, 25). In contrast, a continuous regimen of oral estradiol plus oral norethisterone acetate decreased insulin sensitivity in a 3-month study (26). Prior studies of transdermal regimens are also limited by their short-term duration (24, 27, 28), lack of a control group (28), and use of methods other than the clamp to measure insulin sensitivity (25). Our study, which employed a clamp to measure insulin sensitivity in women receiving a continuous combined regimen of oral conjugated estrogens plus oral medroxyprogesterone acetate or placebo, found a reduction of 17% in insulin sensitivity with HRT over 2 yr. The effect of transdermal estradiol plus transdermal progestin on insulin sensitivity has not been reported, but should be studied as an alternative regimen which may minimize changes in insulin sensitivity. The ability of hyperinsulinemia to suppress endogenous glucose production was not affected by HRT. This suggests that the deterioration in insulin sensitivity with oral conjugated estrogens plus medroxyprogesterone acetate occurs at sites of insulin-stimulated glucose uptake (i.e. skeletal muscle and fat) rather than at sites of glucose production (i.e. liver and kidney). Effects on insulin sensitivity with other medications such as thiazolidinediones also occurs at the peripheral level (muscle) (31). The mechanism for the decrease in peripheral insulin sensitivity with HRT is not clear, but could involve a reduction in glucose uptake by skeletal muscle with a reduction in muscle glycogen synthase (32).

Sites et al. • HRT, Fat, and Insulin Sensitivity

J Clin Endocrinol Metab, May 2005, 90(5):2701–2707

2705

FIG. 3. Effect of HRT on insulin sensitivity and endogenous glucose production in women randomized to placebo (f) or HRT (Œ). Values are means ⫾ SEM. A, HRT decreased absolute insulin sensitivity compared with placebo based on time by treatment analysis (repeated measures ANOVA with baseline as a covariate, P ⬍ 0.01, n ⫽ 22 in placebo group and n ⫽ 25 in HRT group). B, HRT decreased insulin sensitivity per kilogram of fat-free mass (repeated measures ANOVA with baseline as a covariate, P ⫽ 0.02, n ⫽ 23 in placebo group and n ⫽ 25 in HRT group). C, Percent change in insulin sensitivity from baseline at each time point in women randomized to placebo or HRT. D, Suppression of endogenous glucose production by hyperinsulinemia during the clamp. Residual endogenous glucose production under euglycemic hyperinsulinemic conditions was subtracted from residual endogenous glucose production under postabsorptive conditions as an indicator of insulin-induced suppression of endogenous glucose production. There were no significant differences between groups over time.

The reversibility of the decrease in insulin sensitivity with HRT may be reassuring for postmenopausal women considering using HRT for a relatively short period of time. The American College of Obstetricians and Gynecologists and the U.S. Food and Drug Administration now recommend that HRT be used for the shortest time possible and at the lowest dose in symptomatic individuals (33, 34). Clinical practice has responded to these recent guidelines, with many

patients discontinuing HRT or tapering to lower doses since the publication of the Women’s Health Initiative findings (35, 36). Despite these recent changes in clinical practice, little information is available with regard to insulin sensitivity after discontinuation of HRT. In a cross-sectional study, we reported no difference in insulin sensitivity between recent HRT users compared with never users after 2 months of HRT discontinuation (37). Thus, the decrease in insulin sensitivity

FIG. 4. Effect of HRT discontinuation on insulin sensitivity (A) and on insulin sensitivity per kilogram fat-free mass (B) in women randomized to placebo (f) or HRT (Œ). Values are means ⫾ SEM (n ⫽ 8 in placebo group and n ⫽ 9 in HRT group). The decrease in insulin sensitivity at the conclusion of the study at 24 months completely reversed at 1 yr after HRT discontinuation (36 months).

2706

J Clin Endocrinol Metab, May 2005, 90(5):2701–2707

with HRT occurs early (by 6 months), and appears to reverse readily in postmenopausal women upon discontinuation of medication. The health implications of a decrease in insulin sensitivity with ongoing HRT in postmenopausal women are unknown. Our findings of a decrease in insulin-stimulated glucose disposal with HRT but no change in endogenous glucose production mimic the changes noted by Weyer et al. (38) in the transition from normal to impaired glucose tolerance. Of note, none of the women in our study became diabetic during the 2-yr trial based on fasting glucose levels at each time point (for HRT, means ⫾ sd ⫽ 79.3 ⫾ 4.5, 78.9 ⫾ 4.2, 77.4 ⫾ 4.4, 78.5 ⫾ 4.7 mg/dl and for placebo, means ⫾ sd ⫽ 83.0 ⫾ 6.1, 81.7 ⫾ 5.4, 79.9 ⫾ 4.5, 81.2 ⫾ 7.1 mg/dl; conversion factor to SI units ⫽ 0.05551). In older postmenopausal women, HRT has been reported to reduce the incidence of diabetes mellitus by 17% in women without known heart disease (39) and by 35% in women with known heart disease (40). Whether this reduction in the incidence of diabetes occurs through changes in insulin resistance is unclear, because only 8% of the women without heart disease had insulin resistance reported by homeostasis model of assessment-insulin resistance (39), and too few women with heart disease had fasting glucose, fasting insulin, and homeostasis model of assessment-insulin resistance measurements performed (40). Our study would suggest that if HRT reduces the incidence of diabetes mellitus as reported, it may do so by mechanisms other than by improving insulin sensitivity. Furthermore, because insulin resistance is a factor in the progression to cardiovascular disease (41), it is possible that increased insulin resistance with HRT contributes to the finding of increased cardiovascular events seen early in the Women’s Health Initiative Trial (35). Increases in intraabdominal fat are associated with metabolic abnormalities and cardiovascular disease (42). During our 2-yr trial, both women on continuous combined HRT and placebo gained similar amounts of intraabdominal fat (6.5 cm2 vs. 11.9 cm2, respectively). In a recent open-label trial in Sweden, estradiol valerate, 2 mg/d, with 10 mg medroxyprogesterone acetate given for 10 d every 3 months reduced intraabdominal fat by 11% compared with no change with placebo after 1 yr (29). sc fat did not change significantly in their study. It is possible that HRT regimen differences explain the differences in intraabdominal fat findings between studies. Higher doses of estrogen with an infrequent progestin may reduce intraabdominal fat, but continuous exposure to progestin may dilute the benefit of estrogen. Perhaps of practical interest, neither weight nor waist circumference were reduced significantly by HRT in our study. This finding contrasts to those of Espeland et al. (43), and Kanaya et al. (40), who reported that HRT reduced waist circumference significantly by 1.0 –1.2 cm and weight significantly by 1.0 kg compared with placebo in large clinical trials. Our study was designed to examine the effect of HRT on body fat distribution by CT measured outcome variables rather than by anthropometric outcome variables. Thus, our inability to detect significant changes in weight or weight circumference with HRT was likely limited by our smaller sample size. Increased total and regional body fat are associated with increased insulin resistance in postmenopausal women (13),

Sites et al. • HRT, Fat, and Insulin Sensitivity

yet our study indicates that these outcome variables may not necessarily be linked. HRT reduced insulin sensitivity without affecting body composition or body fat distribution, contrary to our hypothesis. It has been hypothesized that free fatty acids released from visceral fat to the liver may increase lipid and lipoprotein production by the liver, thus contributing to the metabolic syndrome with its associated insulin resistance (44, 45). Visceral fat increases were no different between HRT and placebo, suggesting an alternative mechanism for the insulin resistance we observed, and furthermore suggesting that a peripheral mechanism involving skeletal muscle is involved. Indeed, other factors such as exercise capacity are more important than are total and regional adiposity to the determination of insulin sensitivity in nonobese women (16). Furthermore, in other studies of late premenopausal and early postmenopausal women similar in age, we reported that postmenopausal women have greater intraabdominal and total fat than premenopausal women, although insulin sensitivity is not different between groups (3, 14). These data suggest that insulin sensitivity and body composition may not necessarily be linked, and that other factors determine variability in insulin sensitivity. Self-reported leisure time activity did not change with HRT, suggesting that the decrease in insulin sensitivity with HRT is not related to a decrease in physical activity with HRT. We found a trend toward a self-reported decrease in carbohydrate and protein intake with HRT, although these changes were not significant. These dietary findings suggest that changes in insulin sensitivity with HRT are not due to changes in caloric intake. In summary, we report that oral combined conjugated estrogens plus medroxyprogesterone acetate reduce insulin sensitivity without significantly affecting body composition or body fat distribution in nonobese, early postmenopausal women. The mechanism of insulin resistance with HRT involves a decrease in peripheral insulin sensitivity rather than a change in endogenous glucose production. The effect is reversible after discontinuing HRT. The majority of volunteers in our study were Caucasian; it is unknown if HRT affects other racial and ethnic groups in a similar manner. Our study is limited in that it involved a relatively small sample size which may have precluded finding differences between groups. Further studies are necessary to understand the cellular and molecular mechanisms involved in this insulin resistance, and to understand the effect of HRT on insulin sensitivity in obese postmenopausal women. Acknowledgments The authors thank Maureen O’Connell, M.S., and Janice Bunn, Ph.D., for statistical assistance with this manuscript. Received July 27, 2004. Accepted January 26, 2005. Address all correspondence and requests for reprints to: Cynthia K. Sites, M.D., Department of Obstetrics and Gynecology, University of Vermont College of Medicine, Given Building C250, 89 Beaumont Avenue, Burlington, Vermont 05405. E-mail: [email protected]. This work was supported by the National Institutes of Health Grants R29 AG15121 and M01RR10932S2 (both to C.K.S.), and by the General Clinical Research Center at the University of Vermont (M01 RR109).

Sites et al. • HRT, Fat, and Insulin Sensitivity

J Clin Endocrinol Metab, May 2005, 90(5):2701–2707

References 1. Ley CJ, Lees B, Stevenson JC 1992 Sex- and menopause-associated changes in body fat distribution. Am J Clin Nutr 5:950 –954 2. Svendsen OL, Hassager C, Christiansen C 1995 Age- and menopause-associated variations in body composition and fat distribution in healthy women as measured by dual-energy x-ray absorptiometry. Metabolism 44:369 –373 3. Toth MJ, Tchernof A, Sites CK, Poehlman ET 2000 Effect of menopausal status on body composition and abdominal fat distribution. Int J Obes Relat Metab Disord 24:226 –231 4. Wu SI, Chou P, Tsai ST 2001 The impact of years since menopause on the development of impaired glucose tolerance. J Clin Epidemiol 54:117–120 5. Lorenzo C, Okoloise M, Williams K, Stern MP, Haffner SM; San Antonio Heart Study 2003 The metabolic syndrome as predictor of type 2 diabetes: the San Antonio Heart Study. Diabetes Care 26:3153–3159 6. Rewers M, Zaccaro D, D’Agnostino R, Haffner S, Saad MF, Selby JV, Bergman R, Savage P; Insulin Resistance Atherosclerosis Study Investigators. 2004 Insulin sensitivity, insulinemia, and coronary artery disease: the Insulin Resistance Atherosclerosis Study. Diabetes Care 27:781–787 7. Haarbo J, Gotfredsen A, Christiansen C 1991 Postmenopausal hormone replacement therapy prevents central distribution of body fat after menopause. Metabolism 40:1323–1326 8. Reubinoff BE, Wurtman J, Rojansky N, Adler D, Stein P, Schenker JG, Brzezinski A 1995 Effects of hormone replacement therapy on weight, body composition, fat distribution, and food intake in early postmenopausal women: a prospective study. Fertil Steril 64:963–968 9. Gambacciani M, Ciaponi M, Cappagli B, Piaggesi L, De Simone L, Orlandi R, Genazzani AR 1997 Body weight, body fat distribution, and hormonal replacement therapy in early postmenopausal women. J Clin Endocrinol Metab 82:414 – 417 10. Lindheim SR, Presser SC, Ditkoff EC 1993 A possible bimodal effect of estrogen on insulin sensitivity in postmenopausal women and the attenuating effect of added progestin. Fertil Steril 60:664 – 667 11. Ferrara A, Barrett-Connor E, Wingard DL, Edelstein SL 1995 Sex differences in insulin levels in older adults and the effect of body size, estrogen replacement therapy, and glucose tolerance status. The Rancho Bernardo Study, 1984 –1987. Diabetes Care 18:220 –225 12. DeFronzo RA, Tobin JD, Andres R 1979 Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 237:E214 –E223 13. Sites CK, Calles-Escandon J, Brochu M, Butterfield M, Ashikaga T, Poehlman ET 2000 Relation of regional fat distribution to insulin sensitivity in postmenopausal women. Fertil Steril 73:61– 65 14. Toth MJ, Sites CK, Eltabbakh GH, Poehlman ET 2000 Effect of menopausal status on insulin-stimulated glucose disposal. Diabetes Care 23:801– 806 15. Sites CK, Toth MJ, Cushman M, L’Hommedieu GD, Tchernof A, Tracy RP, Poehlman ET 2002 Menopause-related differences in inflammation markers and their relationship to body fat distribution and insulin-stimulated glucose disposal. Fertil Steril 77:128 –135 16. Toth MJ, Sites CK, Cefalu WT, Matthews DE, Poehlman ET 2001 Determinants of insulin-stimulated glucose disposal in middle-aged, premenopausal women. Am J Physiol Endocrinol Metab 281:E113–E121 17. Basiotis PP, Welsh SO, Cronin FJ, Kelsay JL, Mertz W 1987 Number of days of food intake records required to estimate individual and group nutrient intakes with defined confidence. J Nutr 117:1638 –1641 18. Thompson FE, Byers T 1994 Dietary assessment resource manual. J Nutr 124:2245S–2317S 19. Taylor BL, Jacobs DR, Schucker B 1978 A questionnaire for the assessment of leisure time physical activity. J Chronic Dis 31:741–755 20. Albanes D, Conway JM, Taylor PR 1990 Validation and comparison of eight physical activity questionnaires. Epidemiology 1:65–71 21. Folsom AR, Jacobs DR, Caspersen CJ 1986 Test-retest reliability of the Minnesota Leisure Time Physical Activity Questionnaire. J Chronic Dis 39:505–511 22. 1985 SAS procedure guide for personal computers, version 6 edition. Cary, NC: SAS Institute Inc. 23. Dixon WJ 1988 BMDP statistical software manual, volume 2. Berkeley, CA: University of California Press 24. O’ Sullivan AJ, Ho KKY 1995 A comparison of the effects of oral and transdermal estrogen replacement on insulin sensitivity in postmenopausal women. J Clin Endocrinol Metab 80:1783–1788 25. Cagnacci A, Soldani R, Carriero PL, Paoletti AM, Fioretti P, Melis GB 1992

26.

27.

28.

29.

30.

31. 32. 33. 34. 35.

36. 37. 38. 39.

40.

41. 42. 43. 44. 45.

2707

Effects of low doses of transdermal 17␤-estradiol on carbohydrate metabolism in postmenopausal women. J Clin Endocrinol Metab 74:1396 –1400 Kimmerle R, Heinemann L, Heise T, Bender R, Weyer C, Hirschberger S, Berger M 1999 Influence of continuous combined estradiol-norethisterone acetate preparations on insulin sensitivity in postmenopausal nondiabetic women. Menopause 6:36 – 42 Duncan AC, Lyall H, Roberts RN, Petrie JR, Perera MJ, Monaghan S, Hart DM, Connell JM, Lumsden MA 1999 The effect of estradiol and a combined estradiol/progestagen preparation on insulin sensitivity in healthy postmenopausal women. J Clin Endocrinol Metab 84:2402–2407 Cucinelli F, Paparella P, Soranna L, Barini A, Cinque B, Mancuso S, Lanzone A 1999 Differential effect of transdermal estrogen plus progestagen replacement therapy on insulin metabolism in postmenopausal women: relation to their insulinemic secretion. Eur J Endocrinol 140:215–223 Mattiasson I, Rendell M, Tornquist C, Jeppsson S, Hulthen UL 2002 Effects of estrogen replacement therapy on abdominal fat compartments as related to glucose and lipid metabolism in early postmenopausal women. Horm Metab Res 34:583–588 Walker RJ, Lewis-Barned NJ, Sutherland WH, Goulding A, Edwards EA, deJong SA, Gold E, Walker HL 2001 The effects of sequential combined oral 17␤-estradiol and norethisterone acetate on insulin sensitivity and body composition in healthy postmenopausal women: a randomized single blind placebo-controlled study. Menopause 8:27–32 Hevener AL, He W, Barak Y, Le J, Bandyopadhyay G, Olson P, Wilkes J, Evans RM, Olefsky J 2003 Muscle-specific Pparg deletion causes insulin resistance. Nat Med 9:1491–1497 Kumagai S, Hommang A, Bjorntorp P 1993 The effects of oestrogen and progesterone on insulin sensitivity in female rats. Acta Physiol Scand 149:91–97 ACOG Committee on Gynecological Practice 2001 Committee opinion no. 262, December 2001. Risk of breast cancer with estrogen-progestin replacement therapy. Obstet Gynecol 98:1181–1184 Stephenson J 2003 FDA orders estrogen safety warnings: agency offers guidance for HRT use. JAMA 289:537–538 Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Koogerberg C, Stefanick ML, Jackson RD, Beresford SA, Howard VT, Johnson KC, Kotchen JM, Ockere J, Writing Group for the Women’s Health Initiative Investigators 2002 Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA 288:2321–2333 Hersh AL, Stefanick ML, Stafford RS 2004 National use of postmenopausal hormone therapy: annual trend and response to recent evidence. JAMA 1:47–53 Sites CK, L’ Hommedieu GD, Brochu M, Poehlman ET 2001 Previous exposure to hormone replacement therapy and confounders in metabolic studies. Menopause 8:281–285 Weyer C, Bogardus C, Mott DM, Pratley RE 1999 The natural history of insulin secretory dysfunction and insulin resistance in the pathogenesis of type 2 diabetes mellitus. J Clin Invest 104:787–794 Margolis KL, Bonds DE, Rodabough RJ, Tinker L, Phillips LS, Allen C, Bassford T, Burke G, Torrens J, Howard BV 2004 Effect of oestrogen plus progestin on the incidence of diabetes in postmenopausal women: results from the Women’s Health Initiative Hormone Trial. Diabetologia 47:1175–1187 Kanaya AM, Herrington D, Vittinghoff E, Lin F, Grady D, Bittner V, Cauley JA, Barrett-Connor E 2003 Glycemic effects of postmenopausal hormone therapy: the Heart and Estrogen/Progestin Replacement Study. Ann Intern Med 138:1–9 Hegazi RA, Sutton-Tyrrell K, Evans RW, Kuller LH, Belle S, Yamamoto M, Edmundowicz D, Kelley DE 2003 Relationship of adiposity to subclinical atherosclerosis in obese patients with type 2 diabetes. Obes Res 11:1597–1605 Lamarche B 1998 Abdominal obesity and its metabolic complications: implications for the risk of ischaemic heart disease. Coronary Artery Dis 9:473– 481 Espeland MA, Stefanick ML, Kritz-Silverstein D, Finberg SE, Waclawiw MA, James MK 1997 Effect of postmenopausal hormone therapy on body weight and waist and hip girths. J Clin Endocrinol Metab 82:1549 –1556 Mauriege P, Prud’homme D, Lemieux S, Tremblay A, Despres JP 1995 Regional differences in adipose tissue lipolysis from lean and obese women: existence of postreceptor alterations. Am J Physiol 269:E341–E350 Bjorntorp P 1990 Portal adipose tissue as a generator of risk factors for cardiovascular disease and diabetes. Arterioscler Thromb 10:493– 496

JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.