Protection of Bone Mass by Estrogens and Raloxifene during Exercise-Induced Weight Loss
http://www.100md.com
《临床内分泌与代谢杂志》
Abstract
The aim of this study was to determine whether estrogen and/or raloxifene help to conserve bone mineral density (BMD) during moderate weight loss. Postmenopausal women (n = 68) participated in a 6-month weight loss program that consisted primarily of supervised exercise training. Another 26 women were studied over 6 months of weight stability. All participants were randomized to three treatment arms: placebo, raloxifene (60 mg/d), or hormone therapy (HT; conjugated estrogens, 0.625 mg/d; trimonthly medroxyprogesterone acetate, 5 mg/d for 13 d, for women with a uterus). Changes in body weight (mean ± SE) averaged 0.8 ± 0.5 kg in the weight-stable group and –4.1 ± 0.4 kg in the weight loss group. Across all measured skeletal sites, average changes in BMD in weight stable women were –0.6 ± 1.1% (n = 7), 0.9 ± 0.6% (n = 9), and 3.0 ± 0.7% (n = 10) in the placebo, raloxifene, and HT groups, respectively; comparable BMD changes in the weight loss groups were –1.5 ± 0.5% (n = 22), –0.5 ± 0.5% (n = 23), and 1.1 ± 0.4% (n = 23). There were no significant interactions between weight loss and drug treatment on changes in BMD, but there were significant main effects of weight loss on lumbar spine (P = 0.022), total hip (P = 0.010), and trochanter BMD (P < 0.001). These findings suggest that weight loss, even when modest in magnitude and induced by exercise training, causes a reduction in BMD, particularly in women not taking raloxifene or HT. It is not known whether reductions in BMD of this magnitude increase the risk for osteoporotic fracture.
Introduction
BONE MINERAL DENSITY (BMD), measured by dual energy x-ray absorptiometry (DXA), is a strong determinant of the risk for osteoporotic fracture. According to a recent meta-analysis, a reduction in hip BMD T-score of 1 (i.e. equivalent to 1 SD of BMD in young Caucasian women) is associated with a 140% increase in relative risk for hip fracture in Caucasian women (1).
There is a positive association between body weight and BMD, and increased body weight is thought to confer protection against osteoporotic fractures (2). However, prospective data indicate that bone loss occurs in women even when body weight is increasing (3, 4), and weight loss accelerates the decline in BMD of older women and men (5). In a Finnish population study of randomly selected peri- and postmenopausal women, the most significant determinants of changes in BMD of the lumbar spine and femoral neck over a 5-yr follow-up period were menopausal transition, use of hormone therapy (HT), and change in body weight (6, 7). It was also noted that the weight loss-related decline in BMD was significantly attenuated in women who reported any use of HT (7). Observational studies have also suggested that HT at least partially counteracts the loss of bone mineral associated with weight loss (3, 5).
To our knowledge, the protection of bone mass by HT during weight loss has not been investigated in a randomized, controlled fashion. In light of recent evidence that the risks of HT are greater than once thought (8, 9, 10, 11), current recommendations are that HT be used primarily for the management of menopausal symptoms (12, 13, 14). Therefore, when evaluating potential beneficial effects of HT on bone, there is now greater need to determine whether such effects also occur in response to selective estrogen receptor modulators. In this context, the purpose of this study was to determine whether the reduction in BMD in response to moderate weight loss is attenuated by HT and/or raloxifene in postmenopausal women. All participants were randomized to receive placebo, raloxifene, or HT, and those in the weight loss arm participated in a 6-month, supervised, exercise training program. Weight loss was induced primarily via exercise training, because this was part of a larger study examining certain metabolic responses to exercise training and drug intervention.
Subjects and Methods
Subjects
The study participants were sedentary, healthy, postmenopausal women, aged 50–70 yr, who were nonsmokers and overweight or moderately obese. Postmenopausal status was defined as the absence of menses for at least 1 yr or, in women who had undergone hysterectomy, a serum FSH level greater than 30 IU/liter. All participants had a normal mammogram and Pap test within 1 yr of enrolling in the study.
Screening tests included medical history, physical examination, blood chemistries, 12-lead electrocardiogram (ECG), and an exercise stress test. All subjects were confirmed to be euthyroid or receiving adequate replacement therapy based on a normal ultrasensitive TSH level. Volunteers were excluded from the study if they had contraindications to estrogen or raloxifene treatment, including history of breast cancer or other estrogen-dependent neoplasm, liver disease, undiagnosed vaginal bleeding, and history of venous thromboembolism. Other exclusion criteria included coronary artery disease, clinically significant abnormal resting ECG, angina and/or ECG evidence of myocardial ischemia during the maximal exercise stress test, resting blood pressure above 150 mm Hg systolic or 90 mm Hg diastolic, clinically significant arrhythmias, congestive heart failure, aortic stenosis, or unstable health status. Volunteers were also excluded if they had orthopedic or other problems that would interfere with exercise testing or training. Women who had been receiving HT or raloxifene within 6 months of screening were not enrolled.
Women were recruited separately for the weight loss and weight-stable arms of the study, but all met the inclusion and exclusion criteria. The Colorado Multiple Institutional Review Board approved the study. All volunteers who underwent screening for the study provided written informed consent to participate. In addition, active participants reconsented to continued participation in the study on two occasions as a result of new information regarding risks of continuous, combined hormone therapy (8, 9, 10, 11).
Weight loss arm
For the weight loss arm of the study, 138 volunteers underwent an orientation session to learn about the study; 20 elected not to participate, 36 were found to be ineligible, and the remaining 82 were randomized to drug treatment. During the intervention period, 14 participants (five placebo, seven raloxifene, and two HT) were lost to follow-up evaluation for personal (n = 10), medical (n = 1; worsening of multiple sclerosis), or unknown (n = 2) reasons or because of new information regarding the risks of HT (n = 1). Data are reported for the 68 women who completed follow-up evaluations. The racial composition of this cohort was predominantly Caucasian (Caucasian, n = 57; black/African-American, n = 9; Native American/Alaska Native, n = 2); five women reported being of Spanish/Latino/Hispanic ethnicity.
Weight-stable arm
For the weight-stable arm of the study, 61 volunteers underwent an orientation session to learn about the study; 16 elected not to participate, 14 were found to be ineligible, and the remaining 31 were randomized to drug treatment. Three participants (two raloxifene and one HT) were lost to follow-up evaluation during the intervention period (two did not tolerate drug treatment; one because of new information regarding risks of HT), and two finishers were excluded from the final data analysis (one started bisphosphonate therapy; one had a 20-kg increase in body weight). Data are reported for the 26 participants included in the final analyses. The racial composition of this cohort was predominantly Caucasian (Caucasian, n = 23; Asian/Native Hawaiian/other Pacific Islander, n = 2; Native American/Alaska Native, n = 1); one woman reported being of Spanish/Latino/Hispanic ethnicity.
Drug interventions
Eligible volunteers were randomized to three treatment arms: placebo, raloxifene, and HT. The randomizations were performed separately for the weight loss and weight-stable arms. The intervention was administered in a double-blinded fashion, to the extent possible given that side-effects often reveal treatment status to the participant. HT was daily conjugated estrogens (0.625 mg); women with an intact uterus also received trimonthly medroxyprogesterone acetate (5 mg/d for 13 consecutive days). Because the HT regimen minimized exposure to progestins while still protecting the endometrium (15), metabolic responses to HT are thought to reflect the actions of estrogens. The raloxifene treatment dose was 60 mg daily, and placebo treatment was a daily placebo tablet; women in these groups with an intact uterus also received trimonthly placebo medroxyprogesterone acetate treatment.
Weight loss intervention
Weight loss was induced primarily through a 6-month, supervised, endurance exercise training program. However, because weight loss with exercise occurs slowly, the program included 1-wk periods of a reduced calorie diet in the first, third, and fifth months. It has been reported that 1 wk of a reduced calorie diet serves as a jump start to losing weight and provides motivation for participants to continue to lose weight (16).
Participants were expected to attend three supervised exercise sessions per week, but were encouraged to attend more frequently and to exercise at home. During the first few weeks of the program, the goal was to exercise at a moderate intensity (i.e. 60–70% of maximal heart rate) and gradually increase duration to approximately 50 min/d. Thereafter, the goal for an exercise session was to generate an increase in energy expenditure of approximately 400 kcal. To enhance compliance with the exercise program, participants were allowed to select the mode(s) of exercise (i.e. treadmill walking/running, rowing, cycling, and/or elliptical exercise).
The General Clinical Research Center provided take-out meals for the 1-wk periods of reduced food intake in months 1, 3, and 5. Energy intake was reduced to 25 kcal/kg fat-free mass/d, but not less than 1200 kcal/d, with 60% of the energy as carbohydrate, 25% as protein, and 15% as fat.
Subjects in the weight-stable group did not receive either the exercise or dietary interventions. These women were contacted only during the first, third, and fifth months of the study to assess how they were doing with their study medication and to provide medication refills.
DXA
All participants in the weight-stable arm and all but eight participants in the weight loss arm had baseline and follow-up DXA scans performed on a DPX-IQ instrument (Lunar Corp., Madison, WI). Because of a programatic plan at the institution to phase out the Lunar instrument, eight participants in the weight loss arm (four placebo, one raloxifene, and three HT) had both baseline and follow-up scans performed on a Delphi-W instrument (Hologic, Waltham, MA). Total body, lumbar spine (L2–L4), and proximal femur (total hip, femoral neck, trochanter, and femoral shaft) were the three regions scanned at baseline and after the intervention to determine bone mineral content (BMC) and BMD. Total mass, fat mass, and fat-free mass were measured during the total body scans (Lunar extended research analysis software version 4.7c; Hologic software version 11.2).
Diet evaluation
Participants completed 3-d food records (2 week days and 1 weekend day) at baseline (n = 81; 86%) and at the completion of the study (n = 59; 63%). Subjects received detailed instructions about the procedures for recording foods and portion sizes. Records were analyzed for macronutrient composition and calcium content (Nutritionist IV, version 2.2, First DataBank, Inc., San Bruno, CA). Subjects self-reported the presence or absence of calcium supplement use, but the precise mineral content of the supplements was not obtained.
Resting metabolic rate (RMR)
The RMR was measured in the morning after an overnight fast by indirect calorimetry with a ventilated hood (Vmax system, SensorMedics, Yorba Linda, CA). After 15 min of rest and a 5-min habituation period under the hood, oxygen uptake and carbon dioxide production were measured for 25–30 min and used to calculate the RMR (17).
Calculations and statistical analyses
The energy expenditure during walking, running, and cycling was estimated from the metabolic equations recommended by the American College of Sports Medicine that predict the oxygen cost of these activities and from measurements of the energy cost of fast walking (18, 19). It was assumed that energy expenditure averaged 5 kcal/liter oxygen consumed. Because standardized equations are not available for rowing and elliptical exercise, estimates of energy expenditure for these activities were taken directly from the ergometers. For all activities, the values reflected total energy expenditure. The exercise-induced increase in energy expenditure was estimated as the increase above the RMR.
Differences among the groups in baseline characteristics were evaluated by analyses of variance and Tukey post hoc tests when indicated. Changes in outcomes of interest in response to the interventions were analyzed by two-way (weight group and drug group) ANOVA. The sample size provided 80% power to detect group differences of a 0.030 g/cm2 change in BMD. A composite BMD score was generated to evaluate the overall effects of weight loss and drug treatment. The composite score, calculated for each individual, was the average of the relative changes in BMD at the skeletal sites measured (i.e. lumbar spine, total hip, femoral neck, trochanter, and femoral shaft). Analysis of covariance was used to determine whether the magnitude of change in body weight over the 6-month treatment period was a determinant of the changes in BMD after adjustment for drug treatment status. For all analyses, statistical significance was defined as 0.05. All data are reported as the mean ± SD unless otherwise stated.
Results
Baseline characteristics
Weight loss vs. weight-stable groups. The weight loss and weight-stable groups were similar with respect to age, age of menopause, previous HT use, dietary calcium intake, and calcium supplement use. Applying the World Health Organization criteria (20), 22 women in the weight loss group were osteopenic, and six were osteoporotic; 12 women in the weight-stable group were osteopenic, and five were osteoporotic. On entry into the study, women who were recruited for the weight loss arm weighed more (79.2 ± 12.0 vs. 71.1 ± 13.1 kg; P = 0.005) and had more fat mass (35.3 ± 8.7 vs. 29.4 ± 10.7 kg; P = 0.008) and fat-free mass (43.9 ± 4.7 vs. 41.7 ± 5.3 kg; P = 0.047) than those in the weight-stable arm. At study entry, BMD also tended to be higher in the weight loss group compared with the weight-stable group, with a significantly greater femoral shaft BMD (1.140 ± 0.157 vs. 1.064 ± 0.183 g/cm2; P = 0.047) and strong trends for greater lumbar spine BMD (1.143 ± 0.178 vs. 1.062 ± 0.193 g/cm2; P = 0.055), total hip BMD (0.975 ± 0.127 vs. 0.907 ± 0.151 g/cm2; P = 0.060), and whole body BMC (2436 ± 42 vs. 2300 ± 73 g; P = 0.095).
Drug treatment groups. There were no significant differences among the drug treatment groups at baseline in age, age at menopause, body composition, previous HT use, or BMD in either the weight loss (Table 1) or the weight-stable (Table 2) group. Dietary calcium intake was lower (P = 0.039) in the HT group than in the placebo group in both the weight loss and weight-stable groups.
Exercise training and changes in body composition
The volume of exercise performed during the intervention period by women in the weight loss group did not differ significantly among the drug treatment groups. Women in the placebo, raloxifene, and HT groups exercised an average of 3.3 ± 1.1, 3.6 ± 1.1, and 3.4 ± 0.9 d/wk for 49 ± 22, 50 ± 19, and 49 ± 16 min/d at an average heart rate of 133 ± 14, 132 ± 15, and 133 ± 11 beats/min, respectively. The increase in energy expenditure attributable to exercise was estimated to be 1189 ± 637, 1272 ± 565, and 1327 ± 507 kcal/wk in the placebo, raloxifene, and HT groups, respectively. This resulted in an average weight loss of –4.1 ± 3.4 kg and a fat loss of –4.1 ± 3.4 kg with a preservation of fat-free mass; changes were not different among the drug treatment groups, but were significantly different (P < 0.001) from the changes in the weight-stable group (Table 3). The increase in energy expenditure attributable to exercise accounted for a reduction in fat mass of –3.7 kg when based on the conventional estimate of 7718 kcal/kg fat (i.e. 3500 kcal per pound), or a reduction of –3.2 kg when based on an estimate of 9000 kcal/kg fat (i.e. average caloric density for fat of 9 kcal/g). Although paired diet records were available for only 56 subjects (60%), these limited data revealed no significant change in dietary calcium intake from baseline to 6 months. After the weight loss intervention, there were no significant differences (P > 0.05) between the weight loss and weight-stable groups in dietary calcium intake, use of calcium supplements, body weight, body composition, or BMD.
Changes in BMD
Effects of drug treatment. There were no significant interactions between drug treatment and weight group for any of the skeletal measurements (P = 0.156 to 0.899). As would be expected, there were significant main effects (P < 0.001) of drug treatment on BMD, such that the largest increases in BMD occurred in the HT group, and the effects of raloxifene were intermediate to those of HT and placebo treatment (Figs. 1 and 2). Specifically, HT was significantly (P < 0.05) more effective than placebo treatment in increasing BMD at all sites measured with the exception of the femoral neck; HT was also significantly (P < 0.05) more effective than raloxifene in increasing BMD of the total hip and the trochanter and shaft regions of the proximal femur. Raloxifene was significantly (P < 0.05) more effective than placebo in increasing BMD of the lumbar spine.
Discussion
To our knowledge, this was the first study to determine the effects of raloxifene and HT on weight loss-related changes in BMD in a randomized, controlled trial. The relative changes in composite BMD scores in response to moderate weight loss in the placebo, raloxifene, and HT groups were –1.5 ± 0.5%, –0.5 ± 0.5%, and 1.1 ± 0.4%, respectively; comparable changes in the weight-stable group were –0.6 ± 1.1%, 0.9 ± 0.6%, and 3.0 ± 0.7%. Despite the fact that the magnitude of weight loss was modest and was induced primarily through exercise training, only the HT group maintained a positive BMD balance at the most clinically relevant skeletal sites (i.e. lumbar spine, total hip, femoral neck, and trochanter) during weight loss.
It seems plausible that bone mass might be better preserved when weight loss is induced primarily by increasing energy expenditure through exercise, as in the current study, than by restricting energy intake, because of both the mechanical loading forces on the skeleton during exercise and the potential for better preservation of lean body mass during weight loss. We are aware of only one study, of overweight men, that evaluated this in a randomized controlled design (21). The results indicated that when adjusted for fat loss, which was more than 2-fold greater in dieters, there were similar reductions in total body BMC of 11.7 and 11.4 g/kg fat loss in dieters and exercisers, respectively. In the current study only the placebo-treated group had a decline in total body BMC, which averaged 6.2 g/kg fat loss. There were differences between these studies in exercise intensity that could contribute to the apparent discordance in the extent of bone loss. Exercise intensity averaged 80–85% of maximal heart rate in the current study compared with a prescribed intensity of 65–75% of maximal heart rate in the study by Pritchard and colleagues (21); whether that level of intensity was achieved and maintained in that study was not reported. Although the relative heart rate achieved during exercise is not a direct determinant of the mechanical stimulus to bone, it is probably a good surrogate measure of mechanical stress during weight-bearing exercise. The ground reaction forces that are generated during walking and running increase as speed increases (22), as does heart rate, and the magnitude of loading forces is an important determinant of the osteogenic response (23). It should be noted that because the current study was part of a parent study of the effects of exercise-induced weight loss on metabolic factors other than bone, the exercise program was not specifically designed to mechanically stress bone.
Decreases in BMD in response to weight loss, typically induced by energy restriction alone or in combination with exercise, are commonly observed (21, 24, 25, 26, 27, 28, 29, 30). It has been suggested that the loss of bone is due in part to an artifact of the measurement, because changes in the thickness and composition of tissue that surrounds bone as a result of weight loss can influence the measurement of bone mass by DXA (31, 32). The magnitude of this effect has been evaluated by acutely manipulating thickness and composition, using materials that have similar x-ray attenuation characteristics as fat and lean tissue (e.g. lard and water). The results of such experiments have been equivocal (24, 32, 33, 34), but suggest that the effects on the measurement of BMD are very small when weight change is moderate, as in the current study.
There is additional evidence that weight loss-induced bone loss is not simply an artifact of measurement. Reductions in body weight of 5–10% have been found to result in increases in serum and urinary markers of bone turnover and in urinary calcium excretion, with accompanying declines in BMC and/or BMD (28, 35). It has been suggested that these responses occur as a result of reduced calcium intake during diet-induced weight loss, and in fact, calcium supplementation has been found to have an ameliorating effect (27, 28, 36). However, in men who lost weight via either diet or exercise, the decrease in BMC per kilogram of fat loss was similar in the two groups despite an increase in self-reported dietary calcium intake in the exercise group and no change in the diet group; information regarding supplemental calcium intake was not provided (21). In the current study dietary calcium intake and the prevalence of calcium supplement use were similar in the weight-stable and weight loss groups at baseline and 6 months. However, only 60% of our subjects completed 6-month dietary assessments, and detailed information on the mineral content of calcium supplements was not obtained, thus limiting the reliability of our data for total daily calcium intake.
Estrogen status has been suggested to be an important determinant of bone loss in response to weight loss. In adult female rats, reducing energy intake while maintaining calcium intake had deleterious effects on bone mass and strength, and it was suggested that this was related to the concomitant decrease in serum estradiol levels that occurred (37). In an observational study of older women and men, weight loss was a major determinant of bone loss, but the magnitude of the effect was dampened in women receiving estrogen therapy (5). Another observational study of postmenopausal women suggested that HT protected against weight loss-related bone loss (7). Ricci and colleagues found that the negative effects of weight loss on bone mass in postmenopausal women (27, 35) did not occur in premenopausal women (36). They suggested that the increased susceptibility of postmenopausal women not receiving HT to weight loss-induced bone loss could be estrogen mediated, because adipose tissue is an important site of estrogen production via the aromatization of androgens (38). This was based on the observation that changes in fat mass were significantly related to changes in serum estrone levels in postmenopausal women (35).
In one respect the results of the current randomized, controlled trial may appear to support the concept that estrogens provide osteoprotection during weight loss in postmenopausal women. Among women in the weight loss arm, those receiving placebo treatment had significantly larger reductions in BMD of the total body, lumbar spine, total hip, and trochanter and less increase in femoral shaft BMD than those receiving HT. Indeed, only the HT group maintained a positive BMD balance at all sites of measurement in response to moderate weight loss. However, compared with women in the weight-stable arm, the effects of weight loss were apparent in all drug treatment groups. An osteoprotective effect of HT during weight loss would have been supported by interactions between drug treatment and weight group, which were not significant. However, it should be noted that the parent study from which these data were generated was not designed to detect such interactions. The power to detect a significant interaction effect at any of the sites of BMD measurement was less than 40%; the results must therefore be interpreted cautiously. It is possible that the baseline differences between the weight-stable and weight loss groups in body composition and BMD, which were probably related to recruiting women separately for these two intervention arms, may have further limited our ability to detect an interaction between drug treatment and weight group. The mechanisms by which weight loss alters skeletal metabolism and results in a loss of bone mineral remain to be determined. Because body weight influences the magnitude of skeletal loading forces during all ambulatory activities, it is possible that a decrease in BMD represents an appropriately coupled response. This did not appear to be the case in the current study, because a change in body weight was not a significant determinant of a change in BMD. However, it cannot be ruled out that such an association was masked by the superimposed effects of the exercise per se and drug treatment on BMD. Alternatively, it is possible that weight loss is accompanied by changes in metabolic or hormonal factors that independently influence bone metabolism.
Regardless of the mechanisms, the fact that weight reduction causes bone loss in postmenopausal women suggests that weight loss could increase bone fragility in a population already at risk for osteoporosis. Women in the weight loss arm were overweight to mildly obese at the time of study enrollment, yet 46% of them were osteopenic or osteoporotic. Furthermore, the observations in premenopausal women that BMD levels are lower in those with a history of weight cycling than in nonweight cyclers (39) and that preweight loss BMD levels may not be restored with weight regain (26) suggest that weight cycling may be harmful for postmenopausal women at risk for osteoporosis. In the current study the effects of weight loss were most detrimental at skeletal sites that have high trabecular bone content (i.e. lumbar spine and trochanter) and are susceptible to osteoporotic fractures. In this context, repeated attempts at weight loss could be particularly devastating for postmenopausal women if irreversible changes in the microarchitecture and strength of certain skeletal regions occur with each attempt.
In summary, our study documents that moderate weight loss in women nearly a decade beyond the menopause transition results in a significant BMD decline in skeletal regions susceptible to osteoporotic fracture. Because weight loss was induced through exercise training, which presumably has osteoprotective effects via mechanical loading and the maintenance of fat-free mass, randomized trials are needed to determine whether weight loss mediated by energy restriction is even more detrimental to BMD than weight loss mediated by exercise. Although intervention with raloxifene did not prevent a decrease in BMD with exercise-induced weight loss, the expected skeletal benefits of raloxifene and HT were superimposed on the negative effects of weight loss. Thus, raloxifene was effective in attenuating the decrease in BMD during moderate weight loss, whereas the more potent HT was effective in preventing a decrease in BMD. Examining whether other agents (e.g. bisphosphonates) known to prevent osteoporotic fracture (40) may also shield women from bone loss during weight reduction is an important area for future inquiry. Clinicians should remain cognizant of the fact that being overweight does not necessarily confer protection against low BMD in postmenopausal women. It will be important to determine how overweight and obese postmenopausal women can gain the cardiovascular benefits of moderate weight loss without simultaneously increasing risk for osteoporotic fracture.
Acknowledgments
We express our gratitude to the nursing, bionutrition, core laboratory, information systems, and administrative staffs of the General Clinical Research Center and the energy balance core and administrative staffs of the Clinical Nutrition Research Unit for their assistance in conducting this study. We also acknowledge the members of our research group who carried out the day to day activities for the project, including D. Dahl, D. Day, S. Howar, G. Keisling, J. Lynn, A. Moquin, J. Murray, J. Quick, N. Scherer, E. Sdrulla, J. Smoldt, and K. Sutika. Finally, we thank the women who volunteered to participate in the study for their time and efforts.
Footnotes
This work was supported by awards from the NIH, including R01-AG-18198 and F32-AG-05899 (to W.S.G.), K01-AG-19630 (to R.E.V.P.), M01-RR-00051 (to the General Clinical Research Center), and P30-DK-48520 (to the Clinical Nutrition Research Unit). W.S.G. was supported in part by a North American Menopause Society/Wyeth-Ayerst Clinical Research Fellowship and a Hartford/Jahnigen Center of Excellence career award.
First Published Online October 19, 2004
Abbreviations: BMC, Bone mineral content; BMD, bone mineral density; DXA, dual energy x-ray absorptiometry; ECG, electrocardiogram; HT, hormone therapy; RMR, resting metabolic rate.
Received April 16, 2004.
Accepted September 28, 2004.
References
Cummings SR, Bates D, Black DM 2002 Clinical use of bone densitometry: scientific review. JAMA 288:1889–1897
Lane JM, Russell L, Khan SN 2000 Osteoporosis. Clin Orthop 372:139–150
Wu F, Ames R, Clearwater J, Evans MC, Gamble G, Reid IR 2002 Prospective 10-year study of the determinants of bone density and bone loss in normal postmenopausal women, including the effect of hormone replacement therapy. Clin Endocrinol (Oxf) 56:703–711
Hui SL, Perkins AJ, Zhou L, Longcope C, Econs MJ, Peacock M, McClintock C, Johnston Jr CC 2002 Bone loss at the femoral neck in premenopausal white women: effects of weight change and sex-hormone levels. J Clin Endocrinol Metab 87:1539–1543
Hannan MT, Felson DT, Dawson-Hughes B, Tucker KL, Cupples LA, Wilson PW, Kiel DP 2000 Risk factors for longitudinal bone loss in elderly men and women: the Framingham Osteoporosis Study. J Bone Miner Res 15:710–720
Sirola J, Kroger H, Honkanen R, Jurvelin JS, Sandini L, Tuppurainen MT, Saarikoski S 2003 Factors affecting bone loss around menopause in women without HRT: a prospective study. Maturitas 45:159–167
Sirola J, Kroger H, Honkanen R, Sandini L, Tuppurainen M, Jurvelin JS, Saarikoski S 2003 Risk factors associated with peri- and postmenopausal bone loss: does HRT prevent weight loss-related bone loss? Osteoporos Int 14:27–33
Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Kooperberg C, Stefanick ML, Jackson RD, Beresford SA, Howard BV, Johnson KC, Kotchen JM, Ockene J 2002 Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA 288:321–333
Chlebowski RT, Hendrix SL, Langer RD, Stefanick ML, Gass M, Lane D, Rodabough RJ, Gilligan MA, Cyr MG, Thomson CA, Khandekar J, Petrovitch H, McTiernan A 2003 Influence of estrogen plus progestin on breast cancer and mammography in healthy postmenopausal women: the Women’s Health Initiative Randomized Trial. JAMA 289:3243–3253
Rapp SR, Espeland MA, Shumaker SA, Henderson VW, Brunner RL, Manson JE, Gass ML, Stefanick ML, Lane DS, Hays J, Johnson KC, Coker LH, Dailey M, Bowen D 2003 Effect of estrogen plus progestin on global cognitive function in postmenopausal women: the Women’s Health Initiative Memory Study: a randomized controlled trial. JAMA 289:2663–2672
Shumaker SA, Legault C, Thal L, Wallace RB, Ockene JK, Hendrix SL, Jones III BN, Assaf AR, Jackson RD, Kotchen JM, Wassertheil-Smoller S, Wactawski-Wende J 2003 Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women: the Women’s Health Initiative Memory Study: a randomized controlled trial. JAMA 289:2651–2662
American College of Obstetricians and Gynecologists 2002 ACOG committee opinion. Risk of breast cancer with estrogen-progestin replacement therapy. Int J Gynaecol Obstet 76:333–335
North American Menopause Society 2003 Amended report from the NAMS Advisory Panel on Postmenopausal Hormone Therapy. Menopause 10:6–12
Mosca L, Collins P, Herrington DM, Mendelsohn ME, Pasternak RC, Robertson RM, Schenck-Gustafsson K, Smith Jr SC, Taubert KA, Wenger NK 2001 Hormone replacement therapy and cardiovascular disease: a statement for healthcare professionals from the American Heart Association. Circulation 104:499–503
Williams DB, Voigt BJ, Fu YS, Schoenfeld MJ, Judd HL 1994 Assessment of less than monthly progestin therapy in postmenopausal women given estrogen replacement. Obstet Gynecol 84:787–793
Racette SB, Weiss EP, Obert KA, Kohrt WM, Holloszy JO 2001 Modest lifestyle intervention and glucose tolerance in obese African Americans. Obes Res 9:348–355
Weir JB 1949 New method for calculating metabolic rate with special reference to protein metabolism. J Physiol 109:1–9
Kohrt WM, Spina RJ, Holloszy JO, Ehsani AA 1998 Prescribing exercise intensity for older women. J Am Geriatr Soc 46:1–5
American College of Sports Medicine 2000 Guidelines for exercise testing and prescription, 6th Ed. Philadelphia: Lippincott Williams & Wilkins.
Genant HK, Cooper C, Poor G, Reid I, Ehrlich G, Kanis J, Nordin BE, Barrett-Connor E, Black D, Bonjour JP, Dawson-Hughes B, Delmas PD, Dequeker J, Ragi ES, Gennari C, Johnell O, Johnston Jr CC, Lau EM, Liberman UA, Lindsay R, Martin TJ, Masri B, Mautalen CA, Meunier PJ, Khaltaev N 1999 Interim report and recommendations of the World Health Organization Task-Force for Osteoporosis. Osteoporos Int 10:259–264
Pritchard JE, Nowson CA, Wark JD 1996 Bone loss accompanying diet- induced or exercise-induced weight loss: a randomised controlled study. Int J Obes Relat Metab Disord 20:513–520
Keller TS, Weisberger AM, Ray JL, Hasan SS, Shiavi RG, Spengler DM 1996 Relationship between vertical ground reaction force and speed during walking, slow jogging, and running. Clin Biomech 11:253–259
Lanyon LE 1996 Using functional loading to influence bone mass and architecture: objectives, mechanisms, and relationship with estrogen of the mechanically adaptive process in bone. Bone 18:37S–43S
Jensen LB, Quaade F, Sorensen OH 1994 Bone loss accompanying voluntary weight loss in obese humans. J Bone Miner Res 9:459–463
Chao D, Espeland MA, Farmer D, Register TC, Lenchik L, Applegate WB, Ettinger Jr WH 2000 Effect of voluntary weight loss on bone mineral density in older overweight women. J Am Geriatr Soc 48:753–759
Fogelholm GM, Sievanen HT, Kukkonen-Harjula TK, Pasanen ME 2001 Bone mineral density during reduction, maintenance and regain of body weight in premenopausal, obese women. Osteoporos Int 12:199–206
Ricci TA, Chowdhury HA, Heymsfield SB, Stahl T, Pierson RNJ, Shapses SA 1998 Calcium supplementation suppresses bone turnover during weight reduction in postmenopausal women. J Bone Miner Res 13:1045–1050
Jensen LB, Kollerup G, Quaade F, Sorensen OH 2001 Bone minerals changes in obese women during a moderate weight loss with and without calcium supplementation. J Bone Miner Res 16:141–147
Svendsen OL, Hassager C, Christiansen C 1993 Effect of an energy-restrictive diet, with or without exercise, on lean tissue mass, resting metabolic rate, cardiovascular risk factors, and bone in overweight postmenopausal women. Am J Med 95:131–140
Ryan AS, Nicklas BJ, Dennis KE 1998 Aerobic exercise maintains regional bone mineral density during weight loss in postmenopausal women. J Appl Physiol 84:1305–1310
Svendsen OL, Hendel HW, Gotfredsen A, Pedersen BH, Andersen T 2002 Are soft tissue composition of bone and non-bone pixels in spinal bone mineral measurements by DXA similar? Impact of weight loss. Clin Physiol Funct Imaging 22:72–77
Kohrt WM 1998 Preliminary evidence that DEXA provides an accurate assessment of body composition. J Appl Physiol 84:372–377
Tothill P, Hannan WJ, Wilkinson S 2001 Comparisons between a pencil beam and two fan beam dual energy X-ray absorptiometers used for measuring total body bone and soft tissue. Br J Radiol 74:166–176
Tothill P, Hannan WJ 2000 Comparisons between Hologic QDR 1000W, QDR 4500A, and Lunar Expert dual-energy x-ray absorptiometry scanners used for measuring total body bone and soft tissue. Ann NY Acad Sci 904:63–71
Ricci TA, Heymsfield SB, Pierson RNJ, Stahl T, Chowdhury HA, Shapses SA 2001 Moderate energy restriction increases bone resorption in obese postmenopausal women. Am J Clin Nutr 73:347–352
Shapses SA, Von Thun NL, Heymsfield SB, Ricci TA, Ospina M, Pierson Jr RN, Stahl T 2001 Bone turnover and density in obese premenopausal women during moderate weight loss and calcium supplementation. J Bone Miner Res 16:1329–1336
Talbott SM, Cifuentes M, Dunn MG, Shapses SA 2001 Energy restriction reduces bone density and biomechanical properties in aged female rats. J Nutr 131:2382–2387
Burger HG 2001 Physiological principles of endocrine replacement: estrogen. Horm Res 56(Suppl 1):82–85
Fogelholm M, Sievanen H, Heinonen A, Virtanen M, Uusi-Rasi K, Pasanen M, Vuori I 1997 Association between weight cycling history and bone mineral density in premenopausal women. Osteoporos Int 7:354–358
Cranney A, Guyatt G, Griffith L, Wells G, Tugwell P, Rosen C 2002 Meta-analyses of therapies for postmenopausal osteoporosis. IX. Summary of meta-analyses of therapies for postmenopausal osteoporosis. Endocr Rev 23:570–578(W. S. Gozansky, R. E. Van)
The aim of this study was to determine whether estrogen and/or raloxifene help to conserve bone mineral density (BMD) during moderate weight loss. Postmenopausal women (n = 68) participated in a 6-month weight loss program that consisted primarily of supervised exercise training. Another 26 women were studied over 6 months of weight stability. All participants were randomized to three treatment arms: placebo, raloxifene (60 mg/d), or hormone therapy (HT; conjugated estrogens, 0.625 mg/d; trimonthly medroxyprogesterone acetate, 5 mg/d for 13 d, for women with a uterus). Changes in body weight (mean ± SE) averaged 0.8 ± 0.5 kg in the weight-stable group and –4.1 ± 0.4 kg in the weight loss group. Across all measured skeletal sites, average changes in BMD in weight stable women were –0.6 ± 1.1% (n = 7), 0.9 ± 0.6% (n = 9), and 3.0 ± 0.7% (n = 10) in the placebo, raloxifene, and HT groups, respectively; comparable BMD changes in the weight loss groups were –1.5 ± 0.5% (n = 22), –0.5 ± 0.5% (n = 23), and 1.1 ± 0.4% (n = 23). There were no significant interactions between weight loss and drug treatment on changes in BMD, but there were significant main effects of weight loss on lumbar spine (P = 0.022), total hip (P = 0.010), and trochanter BMD (P < 0.001). These findings suggest that weight loss, even when modest in magnitude and induced by exercise training, causes a reduction in BMD, particularly in women not taking raloxifene or HT. It is not known whether reductions in BMD of this magnitude increase the risk for osteoporotic fracture.
Introduction
BONE MINERAL DENSITY (BMD), measured by dual energy x-ray absorptiometry (DXA), is a strong determinant of the risk for osteoporotic fracture. According to a recent meta-analysis, a reduction in hip BMD T-score of 1 (i.e. equivalent to 1 SD of BMD in young Caucasian women) is associated with a 140% increase in relative risk for hip fracture in Caucasian women (1).
There is a positive association between body weight and BMD, and increased body weight is thought to confer protection against osteoporotic fractures (2). However, prospective data indicate that bone loss occurs in women even when body weight is increasing (3, 4), and weight loss accelerates the decline in BMD of older women and men (5). In a Finnish population study of randomly selected peri- and postmenopausal women, the most significant determinants of changes in BMD of the lumbar spine and femoral neck over a 5-yr follow-up period were menopausal transition, use of hormone therapy (HT), and change in body weight (6, 7). It was also noted that the weight loss-related decline in BMD was significantly attenuated in women who reported any use of HT (7). Observational studies have also suggested that HT at least partially counteracts the loss of bone mineral associated with weight loss (3, 5).
To our knowledge, the protection of bone mass by HT during weight loss has not been investigated in a randomized, controlled fashion. In light of recent evidence that the risks of HT are greater than once thought (8, 9, 10, 11), current recommendations are that HT be used primarily for the management of menopausal symptoms (12, 13, 14). Therefore, when evaluating potential beneficial effects of HT on bone, there is now greater need to determine whether such effects also occur in response to selective estrogen receptor modulators. In this context, the purpose of this study was to determine whether the reduction in BMD in response to moderate weight loss is attenuated by HT and/or raloxifene in postmenopausal women. All participants were randomized to receive placebo, raloxifene, or HT, and those in the weight loss arm participated in a 6-month, supervised, exercise training program. Weight loss was induced primarily via exercise training, because this was part of a larger study examining certain metabolic responses to exercise training and drug intervention.
Subjects and Methods
Subjects
The study participants were sedentary, healthy, postmenopausal women, aged 50–70 yr, who were nonsmokers and overweight or moderately obese. Postmenopausal status was defined as the absence of menses for at least 1 yr or, in women who had undergone hysterectomy, a serum FSH level greater than 30 IU/liter. All participants had a normal mammogram and Pap test within 1 yr of enrolling in the study.
Screening tests included medical history, physical examination, blood chemistries, 12-lead electrocardiogram (ECG), and an exercise stress test. All subjects were confirmed to be euthyroid or receiving adequate replacement therapy based on a normal ultrasensitive TSH level. Volunteers were excluded from the study if they had contraindications to estrogen or raloxifene treatment, including history of breast cancer or other estrogen-dependent neoplasm, liver disease, undiagnosed vaginal bleeding, and history of venous thromboembolism. Other exclusion criteria included coronary artery disease, clinically significant abnormal resting ECG, angina and/or ECG evidence of myocardial ischemia during the maximal exercise stress test, resting blood pressure above 150 mm Hg systolic or 90 mm Hg diastolic, clinically significant arrhythmias, congestive heart failure, aortic stenosis, or unstable health status. Volunteers were also excluded if they had orthopedic or other problems that would interfere with exercise testing or training. Women who had been receiving HT or raloxifene within 6 months of screening were not enrolled.
Women were recruited separately for the weight loss and weight-stable arms of the study, but all met the inclusion and exclusion criteria. The Colorado Multiple Institutional Review Board approved the study. All volunteers who underwent screening for the study provided written informed consent to participate. In addition, active participants reconsented to continued participation in the study on two occasions as a result of new information regarding risks of continuous, combined hormone therapy (8, 9, 10, 11).
Weight loss arm
For the weight loss arm of the study, 138 volunteers underwent an orientation session to learn about the study; 20 elected not to participate, 36 were found to be ineligible, and the remaining 82 were randomized to drug treatment. During the intervention period, 14 participants (five placebo, seven raloxifene, and two HT) were lost to follow-up evaluation for personal (n = 10), medical (n = 1; worsening of multiple sclerosis), or unknown (n = 2) reasons or because of new information regarding the risks of HT (n = 1). Data are reported for the 68 women who completed follow-up evaluations. The racial composition of this cohort was predominantly Caucasian (Caucasian, n = 57; black/African-American, n = 9; Native American/Alaska Native, n = 2); five women reported being of Spanish/Latino/Hispanic ethnicity.
Weight-stable arm
For the weight-stable arm of the study, 61 volunteers underwent an orientation session to learn about the study; 16 elected not to participate, 14 were found to be ineligible, and the remaining 31 were randomized to drug treatment. Three participants (two raloxifene and one HT) were lost to follow-up evaluation during the intervention period (two did not tolerate drug treatment; one because of new information regarding risks of HT), and two finishers were excluded from the final data analysis (one started bisphosphonate therapy; one had a 20-kg increase in body weight). Data are reported for the 26 participants included in the final analyses. The racial composition of this cohort was predominantly Caucasian (Caucasian, n = 23; Asian/Native Hawaiian/other Pacific Islander, n = 2; Native American/Alaska Native, n = 1); one woman reported being of Spanish/Latino/Hispanic ethnicity.
Drug interventions
Eligible volunteers were randomized to three treatment arms: placebo, raloxifene, and HT. The randomizations were performed separately for the weight loss and weight-stable arms. The intervention was administered in a double-blinded fashion, to the extent possible given that side-effects often reveal treatment status to the participant. HT was daily conjugated estrogens (0.625 mg); women with an intact uterus also received trimonthly medroxyprogesterone acetate (5 mg/d for 13 consecutive days). Because the HT regimen minimized exposure to progestins while still protecting the endometrium (15), metabolic responses to HT are thought to reflect the actions of estrogens. The raloxifene treatment dose was 60 mg daily, and placebo treatment was a daily placebo tablet; women in these groups with an intact uterus also received trimonthly placebo medroxyprogesterone acetate treatment.
Weight loss intervention
Weight loss was induced primarily through a 6-month, supervised, endurance exercise training program. However, because weight loss with exercise occurs slowly, the program included 1-wk periods of a reduced calorie diet in the first, third, and fifth months. It has been reported that 1 wk of a reduced calorie diet serves as a jump start to losing weight and provides motivation for participants to continue to lose weight (16).
Participants were expected to attend three supervised exercise sessions per week, but were encouraged to attend more frequently and to exercise at home. During the first few weeks of the program, the goal was to exercise at a moderate intensity (i.e. 60–70% of maximal heart rate) and gradually increase duration to approximately 50 min/d. Thereafter, the goal for an exercise session was to generate an increase in energy expenditure of approximately 400 kcal. To enhance compliance with the exercise program, participants were allowed to select the mode(s) of exercise (i.e. treadmill walking/running, rowing, cycling, and/or elliptical exercise).
The General Clinical Research Center provided take-out meals for the 1-wk periods of reduced food intake in months 1, 3, and 5. Energy intake was reduced to 25 kcal/kg fat-free mass/d, but not less than 1200 kcal/d, with 60% of the energy as carbohydrate, 25% as protein, and 15% as fat.
Subjects in the weight-stable group did not receive either the exercise or dietary interventions. These women were contacted only during the first, third, and fifth months of the study to assess how they were doing with their study medication and to provide medication refills.
DXA
All participants in the weight-stable arm and all but eight participants in the weight loss arm had baseline and follow-up DXA scans performed on a DPX-IQ instrument (Lunar Corp., Madison, WI). Because of a programatic plan at the institution to phase out the Lunar instrument, eight participants in the weight loss arm (four placebo, one raloxifene, and three HT) had both baseline and follow-up scans performed on a Delphi-W instrument (Hologic, Waltham, MA). Total body, lumbar spine (L2–L4), and proximal femur (total hip, femoral neck, trochanter, and femoral shaft) were the three regions scanned at baseline and after the intervention to determine bone mineral content (BMC) and BMD. Total mass, fat mass, and fat-free mass were measured during the total body scans (Lunar extended research analysis software version 4.7c; Hologic software version 11.2).
Diet evaluation
Participants completed 3-d food records (2 week days and 1 weekend day) at baseline (n = 81; 86%) and at the completion of the study (n = 59; 63%). Subjects received detailed instructions about the procedures for recording foods and portion sizes. Records were analyzed for macronutrient composition and calcium content (Nutritionist IV, version 2.2, First DataBank, Inc., San Bruno, CA). Subjects self-reported the presence or absence of calcium supplement use, but the precise mineral content of the supplements was not obtained.
Resting metabolic rate (RMR)
The RMR was measured in the morning after an overnight fast by indirect calorimetry with a ventilated hood (Vmax system, SensorMedics, Yorba Linda, CA). After 15 min of rest and a 5-min habituation period under the hood, oxygen uptake and carbon dioxide production were measured for 25–30 min and used to calculate the RMR (17).
Calculations and statistical analyses
The energy expenditure during walking, running, and cycling was estimated from the metabolic equations recommended by the American College of Sports Medicine that predict the oxygen cost of these activities and from measurements of the energy cost of fast walking (18, 19). It was assumed that energy expenditure averaged 5 kcal/liter oxygen consumed. Because standardized equations are not available for rowing and elliptical exercise, estimates of energy expenditure for these activities were taken directly from the ergometers. For all activities, the values reflected total energy expenditure. The exercise-induced increase in energy expenditure was estimated as the increase above the RMR.
Differences among the groups in baseline characteristics were evaluated by analyses of variance and Tukey post hoc tests when indicated. Changes in outcomes of interest in response to the interventions were analyzed by two-way (weight group and drug group) ANOVA. The sample size provided 80% power to detect group differences of a 0.030 g/cm2 change in BMD. A composite BMD score was generated to evaluate the overall effects of weight loss and drug treatment. The composite score, calculated for each individual, was the average of the relative changes in BMD at the skeletal sites measured (i.e. lumbar spine, total hip, femoral neck, trochanter, and femoral shaft). Analysis of covariance was used to determine whether the magnitude of change in body weight over the 6-month treatment period was a determinant of the changes in BMD after adjustment for drug treatment status. For all analyses, statistical significance was defined as 0.05. All data are reported as the mean ± SD unless otherwise stated.
Results
Baseline characteristics
Weight loss vs. weight-stable groups. The weight loss and weight-stable groups were similar with respect to age, age of menopause, previous HT use, dietary calcium intake, and calcium supplement use. Applying the World Health Organization criteria (20), 22 women in the weight loss group were osteopenic, and six were osteoporotic; 12 women in the weight-stable group were osteopenic, and five were osteoporotic. On entry into the study, women who were recruited for the weight loss arm weighed more (79.2 ± 12.0 vs. 71.1 ± 13.1 kg; P = 0.005) and had more fat mass (35.3 ± 8.7 vs. 29.4 ± 10.7 kg; P = 0.008) and fat-free mass (43.9 ± 4.7 vs. 41.7 ± 5.3 kg; P = 0.047) than those in the weight-stable arm. At study entry, BMD also tended to be higher in the weight loss group compared with the weight-stable group, with a significantly greater femoral shaft BMD (1.140 ± 0.157 vs. 1.064 ± 0.183 g/cm2; P = 0.047) and strong trends for greater lumbar spine BMD (1.143 ± 0.178 vs. 1.062 ± 0.193 g/cm2; P = 0.055), total hip BMD (0.975 ± 0.127 vs. 0.907 ± 0.151 g/cm2; P = 0.060), and whole body BMC (2436 ± 42 vs. 2300 ± 73 g; P = 0.095).
Drug treatment groups. There were no significant differences among the drug treatment groups at baseline in age, age at menopause, body composition, previous HT use, or BMD in either the weight loss (Table 1) or the weight-stable (Table 2) group. Dietary calcium intake was lower (P = 0.039) in the HT group than in the placebo group in both the weight loss and weight-stable groups.
Exercise training and changes in body composition
The volume of exercise performed during the intervention period by women in the weight loss group did not differ significantly among the drug treatment groups. Women in the placebo, raloxifene, and HT groups exercised an average of 3.3 ± 1.1, 3.6 ± 1.1, and 3.4 ± 0.9 d/wk for 49 ± 22, 50 ± 19, and 49 ± 16 min/d at an average heart rate of 133 ± 14, 132 ± 15, and 133 ± 11 beats/min, respectively. The increase in energy expenditure attributable to exercise was estimated to be 1189 ± 637, 1272 ± 565, and 1327 ± 507 kcal/wk in the placebo, raloxifene, and HT groups, respectively. This resulted in an average weight loss of –4.1 ± 3.4 kg and a fat loss of –4.1 ± 3.4 kg with a preservation of fat-free mass; changes were not different among the drug treatment groups, but were significantly different (P < 0.001) from the changes in the weight-stable group (Table 3). The increase in energy expenditure attributable to exercise accounted for a reduction in fat mass of –3.7 kg when based on the conventional estimate of 7718 kcal/kg fat (i.e. 3500 kcal per pound), or a reduction of –3.2 kg when based on an estimate of 9000 kcal/kg fat (i.e. average caloric density for fat of 9 kcal/g). Although paired diet records were available for only 56 subjects (60%), these limited data revealed no significant change in dietary calcium intake from baseline to 6 months. After the weight loss intervention, there were no significant differences (P > 0.05) between the weight loss and weight-stable groups in dietary calcium intake, use of calcium supplements, body weight, body composition, or BMD.
Changes in BMD
Effects of drug treatment. There were no significant interactions between drug treatment and weight group for any of the skeletal measurements (P = 0.156 to 0.899). As would be expected, there were significant main effects (P < 0.001) of drug treatment on BMD, such that the largest increases in BMD occurred in the HT group, and the effects of raloxifene were intermediate to those of HT and placebo treatment (Figs. 1 and 2). Specifically, HT was significantly (P < 0.05) more effective than placebo treatment in increasing BMD at all sites measured with the exception of the femoral neck; HT was also significantly (P < 0.05) more effective than raloxifene in increasing BMD of the total hip and the trochanter and shaft regions of the proximal femur. Raloxifene was significantly (P < 0.05) more effective than placebo in increasing BMD of the lumbar spine.
Discussion
To our knowledge, this was the first study to determine the effects of raloxifene and HT on weight loss-related changes in BMD in a randomized, controlled trial. The relative changes in composite BMD scores in response to moderate weight loss in the placebo, raloxifene, and HT groups were –1.5 ± 0.5%, –0.5 ± 0.5%, and 1.1 ± 0.4%, respectively; comparable changes in the weight-stable group were –0.6 ± 1.1%, 0.9 ± 0.6%, and 3.0 ± 0.7%. Despite the fact that the magnitude of weight loss was modest and was induced primarily through exercise training, only the HT group maintained a positive BMD balance at the most clinically relevant skeletal sites (i.e. lumbar spine, total hip, femoral neck, and trochanter) during weight loss.
It seems plausible that bone mass might be better preserved when weight loss is induced primarily by increasing energy expenditure through exercise, as in the current study, than by restricting energy intake, because of both the mechanical loading forces on the skeleton during exercise and the potential for better preservation of lean body mass during weight loss. We are aware of only one study, of overweight men, that evaluated this in a randomized controlled design (21). The results indicated that when adjusted for fat loss, which was more than 2-fold greater in dieters, there were similar reductions in total body BMC of 11.7 and 11.4 g/kg fat loss in dieters and exercisers, respectively. In the current study only the placebo-treated group had a decline in total body BMC, which averaged 6.2 g/kg fat loss. There were differences between these studies in exercise intensity that could contribute to the apparent discordance in the extent of bone loss. Exercise intensity averaged 80–85% of maximal heart rate in the current study compared with a prescribed intensity of 65–75% of maximal heart rate in the study by Pritchard and colleagues (21); whether that level of intensity was achieved and maintained in that study was not reported. Although the relative heart rate achieved during exercise is not a direct determinant of the mechanical stimulus to bone, it is probably a good surrogate measure of mechanical stress during weight-bearing exercise. The ground reaction forces that are generated during walking and running increase as speed increases (22), as does heart rate, and the magnitude of loading forces is an important determinant of the osteogenic response (23). It should be noted that because the current study was part of a parent study of the effects of exercise-induced weight loss on metabolic factors other than bone, the exercise program was not specifically designed to mechanically stress bone.
Decreases in BMD in response to weight loss, typically induced by energy restriction alone or in combination with exercise, are commonly observed (21, 24, 25, 26, 27, 28, 29, 30). It has been suggested that the loss of bone is due in part to an artifact of the measurement, because changes in the thickness and composition of tissue that surrounds bone as a result of weight loss can influence the measurement of bone mass by DXA (31, 32). The magnitude of this effect has been evaluated by acutely manipulating thickness and composition, using materials that have similar x-ray attenuation characteristics as fat and lean tissue (e.g. lard and water). The results of such experiments have been equivocal (24, 32, 33, 34), but suggest that the effects on the measurement of BMD are very small when weight change is moderate, as in the current study.
There is additional evidence that weight loss-induced bone loss is not simply an artifact of measurement. Reductions in body weight of 5–10% have been found to result in increases in serum and urinary markers of bone turnover and in urinary calcium excretion, with accompanying declines in BMC and/or BMD (28, 35). It has been suggested that these responses occur as a result of reduced calcium intake during diet-induced weight loss, and in fact, calcium supplementation has been found to have an ameliorating effect (27, 28, 36). However, in men who lost weight via either diet or exercise, the decrease in BMC per kilogram of fat loss was similar in the two groups despite an increase in self-reported dietary calcium intake in the exercise group and no change in the diet group; information regarding supplemental calcium intake was not provided (21). In the current study dietary calcium intake and the prevalence of calcium supplement use were similar in the weight-stable and weight loss groups at baseline and 6 months. However, only 60% of our subjects completed 6-month dietary assessments, and detailed information on the mineral content of calcium supplements was not obtained, thus limiting the reliability of our data for total daily calcium intake.
Estrogen status has been suggested to be an important determinant of bone loss in response to weight loss. In adult female rats, reducing energy intake while maintaining calcium intake had deleterious effects on bone mass and strength, and it was suggested that this was related to the concomitant decrease in serum estradiol levels that occurred (37). In an observational study of older women and men, weight loss was a major determinant of bone loss, but the magnitude of the effect was dampened in women receiving estrogen therapy (5). Another observational study of postmenopausal women suggested that HT protected against weight loss-related bone loss (7). Ricci and colleagues found that the negative effects of weight loss on bone mass in postmenopausal women (27, 35) did not occur in premenopausal women (36). They suggested that the increased susceptibility of postmenopausal women not receiving HT to weight loss-induced bone loss could be estrogen mediated, because adipose tissue is an important site of estrogen production via the aromatization of androgens (38). This was based on the observation that changes in fat mass were significantly related to changes in serum estrone levels in postmenopausal women (35).
In one respect the results of the current randomized, controlled trial may appear to support the concept that estrogens provide osteoprotection during weight loss in postmenopausal women. Among women in the weight loss arm, those receiving placebo treatment had significantly larger reductions in BMD of the total body, lumbar spine, total hip, and trochanter and less increase in femoral shaft BMD than those receiving HT. Indeed, only the HT group maintained a positive BMD balance at all sites of measurement in response to moderate weight loss. However, compared with women in the weight-stable arm, the effects of weight loss were apparent in all drug treatment groups. An osteoprotective effect of HT during weight loss would have been supported by interactions between drug treatment and weight group, which were not significant. However, it should be noted that the parent study from which these data were generated was not designed to detect such interactions. The power to detect a significant interaction effect at any of the sites of BMD measurement was less than 40%; the results must therefore be interpreted cautiously. It is possible that the baseline differences between the weight-stable and weight loss groups in body composition and BMD, which were probably related to recruiting women separately for these two intervention arms, may have further limited our ability to detect an interaction between drug treatment and weight group. The mechanisms by which weight loss alters skeletal metabolism and results in a loss of bone mineral remain to be determined. Because body weight influences the magnitude of skeletal loading forces during all ambulatory activities, it is possible that a decrease in BMD represents an appropriately coupled response. This did not appear to be the case in the current study, because a change in body weight was not a significant determinant of a change in BMD. However, it cannot be ruled out that such an association was masked by the superimposed effects of the exercise per se and drug treatment on BMD. Alternatively, it is possible that weight loss is accompanied by changes in metabolic or hormonal factors that independently influence bone metabolism.
Regardless of the mechanisms, the fact that weight reduction causes bone loss in postmenopausal women suggests that weight loss could increase bone fragility in a population already at risk for osteoporosis. Women in the weight loss arm were overweight to mildly obese at the time of study enrollment, yet 46% of them were osteopenic or osteoporotic. Furthermore, the observations in premenopausal women that BMD levels are lower in those with a history of weight cycling than in nonweight cyclers (39) and that preweight loss BMD levels may not be restored with weight regain (26) suggest that weight cycling may be harmful for postmenopausal women at risk for osteoporosis. In the current study the effects of weight loss were most detrimental at skeletal sites that have high trabecular bone content (i.e. lumbar spine and trochanter) and are susceptible to osteoporotic fractures. In this context, repeated attempts at weight loss could be particularly devastating for postmenopausal women if irreversible changes in the microarchitecture and strength of certain skeletal regions occur with each attempt.
In summary, our study documents that moderate weight loss in women nearly a decade beyond the menopause transition results in a significant BMD decline in skeletal regions susceptible to osteoporotic fracture. Because weight loss was induced through exercise training, which presumably has osteoprotective effects via mechanical loading and the maintenance of fat-free mass, randomized trials are needed to determine whether weight loss mediated by energy restriction is even more detrimental to BMD than weight loss mediated by exercise. Although intervention with raloxifene did not prevent a decrease in BMD with exercise-induced weight loss, the expected skeletal benefits of raloxifene and HT were superimposed on the negative effects of weight loss. Thus, raloxifene was effective in attenuating the decrease in BMD during moderate weight loss, whereas the more potent HT was effective in preventing a decrease in BMD. Examining whether other agents (e.g. bisphosphonates) known to prevent osteoporotic fracture (40) may also shield women from bone loss during weight reduction is an important area for future inquiry. Clinicians should remain cognizant of the fact that being overweight does not necessarily confer protection against low BMD in postmenopausal women. It will be important to determine how overweight and obese postmenopausal women can gain the cardiovascular benefits of moderate weight loss without simultaneously increasing risk for osteoporotic fracture.
Acknowledgments
We express our gratitude to the nursing, bionutrition, core laboratory, information systems, and administrative staffs of the General Clinical Research Center and the energy balance core and administrative staffs of the Clinical Nutrition Research Unit for their assistance in conducting this study. We also acknowledge the members of our research group who carried out the day to day activities for the project, including D. Dahl, D. Day, S. Howar, G. Keisling, J. Lynn, A. Moquin, J. Murray, J. Quick, N. Scherer, E. Sdrulla, J. Smoldt, and K. Sutika. Finally, we thank the women who volunteered to participate in the study for their time and efforts.
Footnotes
This work was supported by awards from the NIH, including R01-AG-18198 and F32-AG-05899 (to W.S.G.), K01-AG-19630 (to R.E.V.P.), M01-RR-00051 (to the General Clinical Research Center), and P30-DK-48520 (to the Clinical Nutrition Research Unit). W.S.G. was supported in part by a North American Menopause Society/Wyeth-Ayerst Clinical Research Fellowship and a Hartford/Jahnigen Center of Excellence career award.
First Published Online October 19, 2004
Abbreviations: BMC, Bone mineral content; BMD, bone mineral density; DXA, dual energy x-ray absorptiometry; ECG, electrocardiogram; HT, hormone therapy; RMR, resting metabolic rate.
Received April 16, 2004.
Accepted September 28, 2004.
References
Cummings SR, Bates D, Black DM 2002 Clinical use of bone densitometry: scientific review. JAMA 288:1889–1897
Lane JM, Russell L, Khan SN 2000 Osteoporosis. Clin Orthop 372:139–150
Wu F, Ames R, Clearwater J, Evans MC, Gamble G, Reid IR 2002 Prospective 10-year study of the determinants of bone density and bone loss in normal postmenopausal women, including the effect of hormone replacement therapy. Clin Endocrinol (Oxf) 56:703–711
Hui SL, Perkins AJ, Zhou L, Longcope C, Econs MJ, Peacock M, McClintock C, Johnston Jr CC 2002 Bone loss at the femoral neck in premenopausal white women: effects of weight change and sex-hormone levels. J Clin Endocrinol Metab 87:1539–1543
Hannan MT, Felson DT, Dawson-Hughes B, Tucker KL, Cupples LA, Wilson PW, Kiel DP 2000 Risk factors for longitudinal bone loss in elderly men and women: the Framingham Osteoporosis Study. J Bone Miner Res 15:710–720
Sirola J, Kroger H, Honkanen R, Jurvelin JS, Sandini L, Tuppurainen MT, Saarikoski S 2003 Factors affecting bone loss around menopause in women without HRT: a prospective study. Maturitas 45:159–167
Sirola J, Kroger H, Honkanen R, Sandini L, Tuppurainen M, Jurvelin JS, Saarikoski S 2003 Risk factors associated with peri- and postmenopausal bone loss: does HRT prevent weight loss-related bone loss? Osteoporos Int 14:27–33
Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Kooperberg C, Stefanick ML, Jackson RD, Beresford SA, Howard BV, Johnson KC, Kotchen JM, Ockene J 2002 Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA 288:321–333
Chlebowski RT, Hendrix SL, Langer RD, Stefanick ML, Gass M, Lane D, Rodabough RJ, Gilligan MA, Cyr MG, Thomson CA, Khandekar J, Petrovitch H, McTiernan A 2003 Influence of estrogen plus progestin on breast cancer and mammography in healthy postmenopausal women: the Women’s Health Initiative Randomized Trial. JAMA 289:3243–3253
Rapp SR, Espeland MA, Shumaker SA, Henderson VW, Brunner RL, Manson JE, Gass ML, Stefanick ML, Lane DS, Hays J, Johnson KC, Coker LH, Dailey M, Bowen D 2003 Effect of estrogen plus progestin on global cognitive function in postmenopausal women: the Women’s Health Initiative Memory Study: a randomized controlled trial. JAMA 289:2663–2672
Shumaker SA, Legault C, Thal L, Wallace RB, Ockene JK, Hendrix SL, Jones III BN, Assaf AR, Jackson RD, Kotchen JM, Wassertheil-Smoller S, Wactawski-Wende J 2003 Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women: the Women’s Health Initiative Memory Study: a randomized controlled trial. JAMA 289:2651–2662
American College of Obstetricians and Gynecologists 2002 ACOG committee opinion. Risk of breast cancer with estrogen-progestin replacement therapy. Int J Gynaecol Obstet 76:333–335
North American Menopause Society 2003 Amended report from the NAMS Advisory Panel on Postmenopausal Hormone Therapy. Menopause 10:6–12
Mosca L, Collins P, Herrington DM, Mendelsohn ME, Pasternak RC, Robertson RM, Schenck-Gustafsson K, Smith Jr SC, Taubert KA, Wenger NK 2001 Hormone replacement therapy and cardiovascular disease: a statement for healthcare professionals from the American Heart Association. Circulation 104:499–503
Williams DB, Voigt BJ, Fu YS, Schoenfeld MJ, Judd HL 1994 Assessment of less than monthly progestin therapy in postmenopausal women given estrogen replacement. Obstet Gynecol 84:787–793
Racette SB, Weiss EP, Obert KA, Kohrt WM, Holloszy JO 2001 Modest lifestyle intervention and glucose tolerance in obese African Americans. Obes Res 9:348–355
Weir JB 1949 New method for calculating metabolic rate with special reference to protein metabolism. J Physiol 109:1–9
Kohrt WM, Spina RJ, Holloszy JO, Ehsani AA 1998 Prescribing exercise intensity for older women. J Am Geriatr Soc 46:1–5
American College of Sports Medicine 2000 Guidelines for exercise testing and prescription, 6th Ed. Philadelphia: Lippincott Williams & Wilkins.
Genant HK, Cooper C, Poor G, Reid I, Ehrlich G, Kanis J, Nordin BE, Barrett-Connor E, Black D, Bonjour JP, Dawson-Hughes B, Delmas PD, Dequeker J, Ragi ES, Gennari C, Johnell O, Johnston Jr CC, Lau EM, Liberman UA, Lindsay R, Martin TJ, Masri B, Mautalen CA, Meunier PJ, Khaltaev N 1999 Interim report and recommendations of the World Health Organization Task-Force for Osteoporosis. Osteoporos Int 10:259–264
Pritchard JE, Nowson CA, Wark JD 1996 Bone loss accompanying diet- induced or exercise-induced weight loss: a randomised controlled study. Int J Obes Relat Metab Disord 20:513–520
Keller TS, Weisberger AM, Ray JL, Hasan SS, Shiavi RG, Spengler DM 1996 Relationship between vertical ground reaction force and speed during walking, slow jogging, and running. Clin Biomech 11:253–259
Lanyon LE 1996 Using functional loading to influence bone mass and architecture: objectives, mechanisms, and relationship with estrogen of the mechanically adaptive process in bone. Bone 18:37S–43S
Jensen LB, Quaade F, Sorensen OH 1994 Bone loss accompanying voluntary weight loss in obese humans. J Bone Miner Res 9:459–463
Chao D, Espeland MA, Farmer D, Register TC, Lenchik L, Applegate WB, Ettinger Jr WH 2000 Effect of voluntary weight loss on bone mineral density in older overweight women. J Am Geriatr Soc 48:753–759
Fogelholm GM, Sievanen HT, Kukkonen-Harjula TK, Pasanen ME 2001 Bone mineral density during reduction, maintenance and regain of body weight in premenopausal, obese women. Osteoporos Int 12:199–206
Ricci TA, Chowdhury HA, Heymsfield SB, Stahl T, Pierson RNJ, Shapses SA 1998 Calcium supplementation suppresses bone turnover during weight reduction in postmenopausal women. J Bone Miner Res 13:1045–1050
Jensen LB, Kollerup G, Quaade F, Sorensen OH 2001 Bone minerals changes in obese women during a moderate weight loss with and without calcium supplementation. J Bone Miner Res 16:141–147
Svendsen OL, Hassager C, Christiansen C 1993 Effect of an energy-restrictive diet, with or without exercise, on lean tissue mass, resting metabolic rate, cardiovascular risk factors, and bone in overweight postmenopausal women. Am J Med 95:131–140
Ryan AS, Nicklas BJ, Dennis KE 1998 Aerobic exercise maintains regional bone mineral density during weight loss in postmenopausal women. J Appl Physiol 84:1305–1310
Svendsen OL, Hendel HW, Gotfredsen A, Pedersen BH, Andersen T 2002 Are soft tissue composition of bone and non-bone pixels in spinal bone mineral measurements by DXA similar? Impact of weight loss. Clin Physiol Funct Imaging 22:72–77
Kohrt WM 1998 Preliminary evidence that DEXA provides an accurate assessment of body composition. J Appl Physiol 84:372–377
Tothill P, Hannan WJ, Wilkinson S 2001 Comparisons between a pencil beam and two fan beam dual energy X-ray absorptiometers used for measuring total body bone and soft tissue. Br J Radiol 74:166–176
Tothill P, Hannan WJ 2000 Comparisons between Hologic QDR 1000W, QDR 4500A, and Lunar Expert dual-energy x-ray absorptiometry scanners used for measuring total body bone and soft tissue. Ann NY Acad Sci 904:63–71
Ricci TA, Heymsfield SB, Pierson RNJ, Stahl T, Chowdhury HA, Shapses SA 2001 Moderate energy restriction increases bone resorption in obese postmenopausal women. Am J Clin Nutr 73:347–352
Shapses SA, Von Thun NL, Heymsfield SB, Ricci TA, Ospina M, Pierson Jr RN, Stahl T 2001 Bone turnover and density in obese premenopausal women during moderate weight loss and calcium supplementation. J Bone Miner Res 16:1329–1336
Talbott SM, Cifuentes M, Dunn MG, Shapses SA 2001 Energy restriction reduces bone density and biomechanical properties in aged female rats. J Nutr 131:2382–2387
Burger HG 2001 Physiological principles of endocrine replacement: estrogen. Horm Res 56(Suppl 1):82–85
Fogelholm M, Sievanen H, Heinonen A, Virtanen M, Uusi-Rasi K, Pasanen M, Vuori I 1997 Association between weight cycling history and bone mineral density in premenopausal women. Osteoporos Int 7:354–358
Cranney A, Guyatt G, Griffith L, Wells G, Tugwell P, Rosen C 2002 Meta-analyses of therapies for postmenopausal osteoporosis. IX. Summary of meta-analyses of therapies for postmenopausal osteoporosis. Endocr Rev 23:570–578(W. S. Gozansky, R. E. Van)