- Email: [email protected]
Difference in the effect of adiposity on bone density between pre- and postmenopausal women
Maturitas 34 (2000) 261 – 266 www.elsevier.com/locate/maturitas
Difference in the effect of adiposity on bone density between pre- and postmenopausal women Tsutomu Douchi *, Shinako Yamamoto, Toshimichi Oki, Kuninori Maruta, Riki Kuwahata, Hedeki Yamasaki, Yukihiro Nagata Department of Obstetrics and Gynecology, Faculty of Medicine, Kagoshima Uni6ersity, 8 -35 -1 Sakuragaoka, Kagoshima 890 -8520, Japan Received 24 August 1999; accepted 29 November 1999
Abstract Objecti6es: Elevated bone mineral density (BMD) in obese women is partially attributable to the higher circulating estrogen levels derived from extraglandular aromatization in adipose tissue. However, it remains unclear whether there is an effect of overall adiposity on BMD in both pre- and postmenopausal women. The difference in the effect of overall adiposity on BMD between pre- and postmenopausal women was investigated. Materials and methods: Subjects were 296 premenopausal women with regular menstruation and 233 postmenopausal women. Age, age at menarche, years since menopause (YSM, in postmenopausal women), weight, height, and body mass index were recorded. Total fat mass amount, lean mass amount, and percentage of body fat were measured by whole body scanning with dual-energy X-ray absorptiometry (DEXA). Lumbar spine BMD (L2 – L4) was measured by DEXA. In each group, significant determinants of BMD were investigated using univariate and stepwise multiple regression analysis. Results: In postmenopausal women, YSM, lean mass amount, total fat mass amount, and height were significant determinants of BMD (R 2 =0.273, P B0.001). In premenopausal women, only two variables including lean mass amount and age at menarche were significant determinants of lumbar spine BMD (R 2 = 0.110, PB 0.001), but total fat mass amount and percentage of body fat were not significant determinants of BMD. Conclusion: The effect of overall adiposity on BMD is more prominent in postmenopausal women than in premenopausal women. © 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Adiposity; Women; Bone mineral density
1. Introduction Postmenopausal osteoporosis is an important public health issue in Western countries. There * Corresponding author. Tel.: +81-99-2755422; fax: + 8199-2650507.
are many factors which affect bone mineral density (BMD). Estrogen is one of the principal determinants of BMD. It is well known that obese women have elevated BMD [1–4]. This may be partially attributable to the effect of increased serum estrogen levels derived from extraglandular aromatization of androgen in adipose tissue and
0378-5122/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 8 - 5 1 2 2 ( 9 9 ) 0 0 1 1 4 - 0
262
T. Douchi et al. / Maturitas 34 (2000) 261–266
the increased weight-bearing effect of fat mass on BMD. However, it remains unclear whether there is an effect of overall adiposity on BMD in both pre- and postmenopausal women. In the present study, we investigated the difference in the effect of overall adiposity on BMD between pre- and postmenopausal women.
2. Materials and methods Two hundred and ninety-six premenopausal and 233 postmenopausal Japanese women were recruited at the Department of Obstetrics and Gynecology, Kagoshima University Hospital between July 1996 and March 1999. These women were undergoing screening for cervical uterine cancer and ovarian cancer, and were in apparent good health by medical clinical evaluation. All of the postmenopausal women had experienced natural menopause. The subjects were judged to have entered menopause when they had not menstruated for at least 12 months prior to the investigation. Exclusion criteria were ovarian tumor, postmenopausal hormone replacement therapy either previously or currently, alcohol consumption, tobacco users, and endurance physical training. Women on diet or loosing or gaining weight were also excluded from the protocol. None of the subjects took any medications likely to affect BMD during the study. Baseline characteristics including age, age at menopause, years since menopause (YSM), age at menarche, weight, height, and body mass index (BMI) were recorded for each subject. BMI was calculated as weight (kg) divided by height squared (m2). Lumbar spine BMD (L2–L4), total fat mass amount, and the percentage of body fat were measured by dual-enery X-ray absorptiometry (DEXA, QDR 2000/W, Hologic, Waltham, MA). Baseline characteristics, anthropometric characteristics, and lumbar spine BMD were compared between the two groups. In each group, correlations between BMD and variables were investigated. DEXA measurements were performed between 09:00 and 12:00 h with a total body scanner and results were evaluated by the same examiner. This equipment uses switched pulsed stable dual-en-
ergy radiation with 70 and 140 kV. The machine performs serial transverse scans from head to toe at 1.2-cm intervals, providing a pixel size of 1.9 × 1.2 mm. The radiation dose is 0.05–0.15 mGy. The reproducibilities of total body fat mass amount and lean mass amount measurements were determined in 12 women, each of whom was measured twice at a 1-week interval. The coefficient of variation (CV) of total body fat mass measurements was B 3.0% [5]. CV for the lean mass measurements was B 4.5% [6]. All recordings were performed by the same experienced investigator. The examiner was blinded to the study status. Informed written consent was obtained in accordance with institutional guidelines. The study was also conducted in accordance with provisions of the Declaration of Helsinki. Intergroup comparisons were made using Student t test. On regression analysis, the dependent variable was lumbar spine BMD, while independent variables included other variables. On stepwise multiple regression analysis, however, weight and BMI were excluded from the independent variables. A 95% confidence interval (CI) was used to describe the strength of association. PB 0.05 was considered significant.
3. Results Table 1 presents baseline characteristics, anthropometric characteristics, and lumbar spine BMD of the two groups. Mean height, weight, lean mass amount, and lumbar spine BMD were significantly higher in premenopausal women than in postmenopausal women (PB 0.05), while percentage of body fat was significantly higher in postmenopausal women (PB 0.001). Mean BMI and total fat mass amount did not differ between the two groups. Table 2 presents the relationship of lumbar spine BMD to baseline characteristics and anthropometric characteristics on univariate regression analysis in premenopausal women. Height, weight, lean mass amount, and fat mass amount were positively correlated with lumbar spine BMD (PB 0.01), while age at menarche was in-
T. Douchi et al. / Maturitas 34 (2000) 261–266
versely correlated with BMD (P B 0.001). Table 3 presents the outcome of stepwise multiple regression analysis in premenopausal women. Only two variables including lean mass amount and age at menarche were significant determinants of lumbar spine BMD (R 2 =0.110, P B0.001). These relations were independent one from the other. Table 4 presents the relationship of lumbar spine BMD to baseline characteristics and anthropometric characteristics on univariate regression analysis in postmenopausal women. Height, weight, BMI, percentage of body fat, lean mass amount, and fat mass amount were positively correlated with lumbar spine BMD (P B 0.01),
263
while age and YSM were inversely correlated with BMD. Table 5 presents the outcome of stepwise multiple regression analysis in postmenopausal women. YSM, lean mass amount, total fat mass amount, and height were the significant determinants of lumbar spine BMD (R 2 = 0.273, PB 0.001). These relations were independent one from the other.
4. Discussion Factors that affect BMD do not always generally influence BMD throughout all stages of the
Table 1 Baseline characteristics, anthropometric characteristics, and lumbar spine BMD of the two groupsa
n Age (years) Age at menarche (years) Years since menopause Height (cm) Weight (kg) Body mass index (kg/m2) Total fat mass amount (kg) Percentage of body fat (%) Lean mass amount (kg) Lumbar spine BMD (g/cm2)
Premenopausal women
Postmenopausal women
P
296 38.99 8.7 (20–55) 13.09 1.4 (11–19) – 155.6 9 5.7 (140–170) 54.8911.0 (35–85) 22.8 94.3 (14–41) 17.9 97.7 (5.2–52) 31.397.5 (9.8–41.7) 34.59 4.8 (23.8–51.5) 1.09 9 0.16 (0.76–1.55)
233 62.09 7.4 (45–75) 14.891.7 (11–20) 11.9 97.8 (1–26) 150.895.5 (135–165) 53.0 9 8.1 (34–87) 23.2 9 3.3 (16–38) 18.4 9 6.0 (5.2–48) 34.197.3 (8.2–55.2) 32.3 9 4.4 (19.9–44.9) 0.84 9 0.18 (0.45–1.42)
– – B0.001 – B0.001 B0.05 ns ns B0.001 B0.001 B0.001
a Data are expressed as mean 9S.D. or n. Numbers in parentheses indicate ranges; BMD, bone mineral density; ns, not significant.
Table 2 Relationships of lumbar spine BMD to baseline and anthropometric characteristics on univariate regression analysis in premenopausal womena Dependent and independent variables
Lumbar spine BMD Age (years) Age at menarche (years) Height (cm) Weight (kg) Body mass index (k/m2) Fat mass amount (kg) Percentage of body fat (%) Lean mass amount (kg) a
Univariate regression
Standardized regression
Coefficient
95% CI
−0.001 −0.024 0.005 0.004 0.006 0.004 0.002 0.009
−0.003, −0.037, 0.002, 0.002, 0.002, 0.001, −0.000, 0.005,
0.002 −0.012 0.008 0.003 0.010 0.006 0.005 0.010
ns, not significant; BMD, bone mineral density; CI, confidence interval.
Coefficient
P
−0.029 −0.222 0.181 0.269 0.177 0.182 0.104 0.274
ns B0.001 B0.001 B0.001 B0.01 B0.01 ns B0.01
T. Douchi et al. / Maturitas 34 (2000) 261–266
264
Table 3 Outcome of stepwise multiple regression analysis in premenopausal women (n = 296)a Dependent and independent variable Lumbar spine BMD Lean mass amount +Age at menarche
R2
0.075 0.110
DR 2
P
0.035
B0.001 B0.001
a Age, height, percentage of body fat, and total fat mass amount were not significant determinants of lumbar spine BMD. Weight and BMI were excluded from the analysis. BMD, bone mineral density; R 2 = (multiple correlation coefficient)2.
life cycle. Some factors have strong impact on BMD during a certain period of the life cycle, then have a reduced impact on BMD at other stages of life cycle. This concept was also supported by the present study, demonstrating that age at menarche has a strong impact on BMD in premenopausal women, while its effect disappeared after menopause. It is generally considered that body composition (lean and fat mass component) influences BMD. In the current study, it was hypothesized that the effect of overall adiposity on BMD differs between pre- and postmenopausal women. It was found that the total fat mass amount was one of the important determinants of BMD in postmenopausal women, but was a less important
Table 5 Outcome of stepwise multiple regression analysis in postmenopausal women (n =233)a Dependent and independent variable Lumbar spine BMD Years since menopause +Lean mass amount +Total fat mass amount +Height
R2
DR 2
P
0.183 0.234 0.260 0.273
0.005 0.026 0.013
B0.001 B0.001 B0.001 B0.001
a Percentage of body fat and age at menarche were not significant determinants of BMD. Weight and BMI were excluded from the analysis. BMD, bone mineral density; R 2 = (multiple correlation coefficient)2.
determinant in premenopausal women. It is of interest to determine why the effect of overall adiposity on BMD is more prominent in postmenopausal women than in premenopausal women. In a cross-sectional study such as the present study, it may be difficult to directly address the difference in the effect of adiposity on BMD between pre- and postmenopausal women. However, no difference in total fat mass amount between pre- and postmenopausal women suggests that the weight-bearing effect of fat mass may not be responsible for the difference in the effect of overall adiposity on BMD between the two groups. In postmenopausal women, aromatization of androstenedione produces estrone
Table 4 Relationships of lumbar spine BMD to baseline and anthropometric characteristics on univariate regression analysis in postmenopausal womena Dependent and independent variables
Lumbar spine BMD Age (years) Age at menarche (years) Years since menopause (years) Height (cm) Weight (kg) Body mass index Fat mass amount (kg) Percentage of body fat Lean mass amount (kg) a
Univariate regression
Standardized regression
Coefficient
95% CI
−0.009 −0.002 −0.010 0.011 0.007 0.011 0.008 0.004 0.010
−0.012, −0.005, −0.012, 0.007, 0.005, 0.004, 0.004, 0.001, 0.005,
ns, not significant; BMD, bone mineral density; CI, confidence interval.
−0.006 0.000 −0.007 0.015 0.010 0.018 0.010 0.007 0.020
Coefficient
P
−0.399 −0.104 −0.427 0.341 0.343 0.200 0.255 0.170 0.247
B0.001 ns B0.001 B0.001 B0.001 B0.01 B0.001 B0.01 B0.001
T. Douchi et al. / Maturitas 34 (2000) 261–266
which is the major estrogen in postmenopausal women [7]. Although circulating estrogen levels in postmenopausal women are lower than those in premenopausal women, circulating estrogen levels derived from aromatization in obese women appear to enhance BMD from the perspective of the estrogen threshold hypothesis [8]. In fact, there are several data indicating elevated BMD in postmenopausal women with obesity [1 – 5,9]. In premenopausal women with regular menstruation, the effect of estrogen from the extraglandular aromatization on BMD may be masked by greater amount of ovarian estrogens. The observation agrees with previous report [10], demonstrating that body composition has no direct influence on BMD in normal premenopausal women. However, it appears that, in women with prolonged duration of amenorrhea, circulating estrogen from extraglandular aromatization becomes an important preventing factor for bone mineral loss [10 – 12]. Drinkwater et al. [10] demonstrated that obesity becomes more important as the severity of menstrual irregularities increased. The observation disagrees with the report by Aloia et al. [13], demonstrating that adiposity does not play a major role in preventing bone loss. However, they did not separate postmenopausal women from premenopausal women in their analysis. In addition, their observation was mainly based on BMI, which is indirect method of estimating overall adiposity [14]. Finally, normal premenopausal women are prone to undergo fewer changes in BMD in response to various risk factors for bone mineral loss than postmenopausal women, because this group also has several factors inhibiting bone mineral loss that postmenopausal women do not have. These factors may include youth, functioning ovary, greater daily physical activity, healthier daily diet, greater muscle strength [15], and higher baseline BMD. Based on these results, it is concluded that the effect of overall adiposity on BMD is more prominent in postmenopausal women than in premenopausal women. It is speculated that the effect of aromatized estrogen on BMD becomes evident in the situation of profound and pro-
265
longed depletion of ovarian estrogens. In premenopausal women, however, the effect of estrogen from extraglandular aromatization on BMD may be masked by greater amount of ovarian estrogen.
References [1] Dawson-Hughes B, Shipp C, Sadowski L, Dallal G. Bone density of the radius, spine and hip in relation to percent of ideal body weight in postmenopausal women. Calcif Tissue Int 1987;40:310 – 4. [2] Ribot C, Tremollieres F, Pouilles JM, Bonneu M, Germain F, Lowret JP. Obesity and menopausal bone loss: the influence of obesity on vertebral density and bone turnover in postmenopausal women. Bone 1987;8:327 – 31. [3] Hassager C, Chritiansen C. Influence of soft tissue body composition on bone mass and metabolism. Bone 1989;10:414 – 9. [4] Shiraki M, Ito H, Fujimaki H, Higuchi T. Relation between body size and bone mineral density with special reference to sex hormones and calcium regulating hormone in elderly women. Endocrinol Jpn 1991;38: 343 – 9. [5] Douchi T, Yamamoto S, Nakamura S, Oki T, Maruta K, Nagata Y. Bone mineral density in postmenopausal women with endometrial cancer. Maturitas 1999;31: 165 – 70. [6] Douchi T, Yamamoto S, Nakamura S, Ijuin T, Oki T, Maruta K, Nagata Y. The effect of menopause in regional and total body lean mass. Maturitas 1998;29: 247 – 52. [7] Judd HL, Samonki IM, Frumar AM, Lagasse LD. Origin of serum estradiol in postmenopausal women. Obstet Gynecol 1982;59:680 – 6. [8] Barbieri RL. Hormone treatment of endometriosis: the estrogen threshold hypothesis. Am J Obstet Gynecol 1992;166:740 – 5. [9] Goulding A, Taylor RW. Plasma leptin values in relation to bone mass density and to dynamic biochemical markers of bone resorption and formation in postmenopausal women. Calcif Tissue Int 1998;63:456 – 8. [10] Drinkwater BL, Bruemner BS, Chestnut CH III. Menstrual history as a determinant of current bone density in young athletes. J Am Med Assoc 1990;263:545 – 8. [11] Warren MP, Brook-Cunn J, Fox RP, Lancelot C, Newman D, Hamilton WG. Lack of bone accretion and amenorrhea: evidence for a relative osteopoenia in weight-bearing bones. J Clin Endocrinol Metab 1991;72: 847 – 53. [12] Biller BMK, Baum HBA, Rosenthal DI, Saxe VC, Charpie PM, Klibanski A. Progressive trabecular os-
266
T. Douchi et al. / Maturitas 34 (2000) 261–266
teopenia in women with hyperprolactinemic amenorrhea. J Clin Endocrinol Metab 1992;75:692–7. [13] Aloia JF, McGowan DM, Vaswani AN, Ross P, Cohn SH. Relationship of menopause to skeletal and muscle mass. Am J Clin Nutr 1991;53:1378–83. [14] Ley CJ, Lees B, Stevenson JC. Sex- and menopause-asso-
.
ciated changes in body-fat distribution. Am J Clin Nutr 1992;55:950 – 4. [15] Douchi T, Kosha S, Kan R, Nakamura S, Oki T, Nagata Y. Predictors of bone mineral loss in patients with ovarian cancer treated with anticancer agents. Obstet Gynecol 1997;90:12 – 5.
Difference in the effect of adiposity on bone density between pre- and postmenopausal women Tsutomu Douchi *, Shinako Yamamoto, Toshimichi Oki, Kuninori Maruta, Riki Kuwahata, Hedeki Yamasaki, Yukihiro Nagata Department of Obstetrics and Gynecology, Faculty of Medicine, Kagoshima Uni6ersity, 8 -35 -1 Sakuragaoka, Kagoshima 890 -8520, Japan Received 24 August 1999; accepted 29 November 1999
Abstract Objecti6es: Elevated bone mineral density (BMD) in obese women is partially attributable to the higher circulating estrogen levels derived from extraglandular aromatization in adipose tissue. However, it remains unclear whether there is an effect of overall adiposity on BMD in both pre- and postmenopausal women. The difference in the effect of overall adiposity on BMD between pre- and postmenopausal women was investigated. Materials and methods: Subjects were 296 premenopausal women with regular menstruation and 233 postmenopausal women. Age, age at menarche, years since menopause (YSM, in postmenopausal women), weight, height, and body mass index were recorded. Total fat mass amount, lean mass amount, and percentage of body fat were measured by whole body scanning with dual-energy X-ray absorptiometry (DEXA). Lumbar spine BMD (L2 – L4) was measured by DEXA. In each group, significant determinants of BMD were investigated using univariate and stepwise multiple regression analysis. Results: In postmenopausal women, YSM, lean mass amount, total fat mass amount, and height were significant determinants of BMD (R 2 =0.273, P B0.001). In premenopausal women, only two variables including lean mass amount and age at menarche were significant determinants of lumbar spine BMD (R 2 = 0.110, PB 0.001), but total fat mass amount and percentage of body fat were not significant determinants of BMD. Conclusion: The effect of overall adiposity on BMD is more prominent in postmenopausal women than in premenopausal women. © 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Adiposity; Women; Bone mineral density
1. Introduction Postmenopausal osteoporosis is an important public health issue in Western countries. There * Corresponding author. Tel.: +81-99-2755422; fax: + 8199-2650507.
are many factors which affect bone mineral density (BMD). Estrogen is one of the principal determinants of BMD. It is well known that obese women have elevated BMD [1–4]. This may be partially attributable to the effect of increased serum estrogen levels derived from extraglandular aromatization of androgen in adipose tissue and
0378-5122/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 8 - 5 1 2 2 ( 9 9 ) 0 0 1 1 4 - 0
262
T. Douchi et al. / Maturitas 34 (2000) 261–266
the increased weight-bearing effect of fat mass on BMD. However, it remains unclear whether there is an effect of overall adiposity on BMD in both pre- and postmenopausal women. In the present study, we investigated the difference in the effect of overall adiposity on BMD between pre- and postmenopausal women.
2. Materials and methods Two hundred and ninety-six premenopausal and 233 postmenopausal Japanese women were recruited at the Department of Obstetrics and Gynecology, Kagoshima University Hospital between July 1996 and March 1999. These women were undergoing screening for cervical uterine cancer and ovarian cancer, and were in apparent good health by medical clinical evaluation. All of the postmenopausal women had experienced natural menopause. The subjects were judged to have entered menopause when they had not menstruated for at least 12 months prior to the investigation. Exclusion criteria were ovarian tumor, postmenopausal hormone replacement therapy either previously or currently, alcohol consumption, tobacco users, and endurance physical training. Women on diet or loosing or gaining weight were also excluded from the protocol. None of the subjects took any medications likely to affect BMD during the study. Baseline characteristics including age, age at menopause, years since menopause (YSM), age at menarche, weight, height, and body mass index (BMI) were recorded for each subject. BMI was calculated as weight (kg) divided by height squared (m2). Lumbar spine BMD (L2–L4), total fat mass amount, and the percentage of body fat were measured by dual-enery X-ray absorptiometry (DEXA, QDR 2000/W, Hologic, Waltham, MA). Baseline characteristics, anthropometric characteristics, and lumbar spine BMD were compared between the two groups. In each group, correlations between BMD and variables were investigated. DEXA measurements were performed between 09:00 and 12:00 h with a total body scanner and results were evaluated by the same examiner. This equipment uses switched pulsed stable dual-en-
ergy radiation with 70 and 140 kV. The machine performs serial transverse scans from head to toe at 1.2-cm intervals, providing a pixel size of 1.9 × 1.2 mm. The radiation dose is 0.05–0.15 mGy. The reproducibilities of total body fat mass amount and lean mass amount measurements were determined in 12 women, each of whom was measured twice at a 1-week interval. The coefficient of variation (CV) of total body fat mass measurements was B 3.0% [5]. CV for the lean mass measurements was B 4.5% [6]. All recordings were performed by the same experienced investigator. The examiner was blinded to the study status. Informed written consent was obtained in accordance with institutional guidelines. The study was also conducted in accordance with provisions of the Declaration of Helsinki. Intergroup comparisons were made using Student t test. On regression analysis, the dependent variable was lumbar spine BMD, while independent variables included other variables. On stepwise multiple regression analysis, however, weight and BMI were excluded from the independent variables. A 95% confidence interval (CI) was used to describe the strength of association. PB 0.05 was considered significant.
3. Results Table 1 presents baseline characteristics, anthropometric characteristics, and lumbar spine BMD of the two groups. Mean height, weight, lean mass amount, and lumbar spine BMD were significantly higher in premenopausal women than in postmenopausal women (PB 0.05), while percentage of body fat was significantly higher in postmenopausal women (PB 0.001). Mean BMI and total fat mass amount did not differ between the two groups. Table 2 presents the relationship of lumbar spine BMD to baseline characteristics and anthropometric characteristics on univariate regression analysis in premenopausal women. Height, weight, lean mass amount, and fat mass amount were positively correlated with lumbar spine BMD (PB 0.01), while age at menarche was in-
T. Douchi et al. / Maturitas 34 (2000) 261–266
versely correlated with BMD (P B 0.001). Table 3 presents the outcome of stepwise multiple regression analysis in premenopausal women. Only two variables including lean mass amount and age at menarche were significant determinants of lumbar spine BMD (R 2 =0.110, P B0.001). These relations were independent one from the other. Table 4 presents the relationship of lumbar spine BMD to baseline characteristics and anthropometric characteristics on univariate regression analysis in postmenopausal women. Height, weight, BMI, percentage of body fat, lean mass amount, and fat mass amount were positively correlated with lumbar spine BMD (P B 0.01),
263
while age and YSM were inversely correlated with BMD. Table 5 presents the outcome of stepwise multiple regression analysis in postmenopausal women. YSM, lean mass amount, total fat mass amount, and height were the significant determinants of lumbar spine BMD (R 2 = 0.273, PB 0.001). These relations were independent one from the other.
4. Discussion Factors that affect BMD do not always generally influence BMD throughout all stages of the
Table 1 Baseline characteristics, anthropometric characteristics, and lumbar spine BMD of the two groupsa
n Age (years) Age at menarche (years) Years since menopause Height (cm) Weight (kg) Body mass index (kg/m2) Total fat mass amount (kg) Percentage of body fat (%) Lean mass amount (kg) Lumbar spine BMD (g/cm2)
Premenopausal women
Postmenopausal women
P
296 38.99 8.7 (20–55) 13.09 1.4 (11–19) – 155.6 9 5.7 (140–170) 54.8911.0 (35–85) 22.8 94.3 (14–41) 17.9 97.7 (5.2–52) 31.397.5 (9.8–41.7) 34.59 4.8 (23.8–51.5) 1.09 9 0.16 (0.76–1.55)
233 62.09 7.4 (45–75) 14.891.7 (11–20) 11.9 97.8 (1–26) 150.895.5 (135–165) 53.0 9 8.1 (34–87) 23.2 9 3.3 (16–38) 18.4 9 6.0 (5.2–48) 34.197.3 (8.2–55.2) 32.3 9 4.4 (19.9–44.9) 0.84 9 0.18 (0.45–1.42)
– – B0.001 – B0.001 B0.05 ns ns B0.001 B0.001 B0.001
a Data are expressed as mean 9S.D. or n. Numbers in parentheses indicate ranges; BMD, bone mineral density; ns, not significant.
Table 2 Relationships of lumbar spine BMD to baseline and anthropometric characteristics on univariate regression analysis in premenopausal womena Dependent and independent variables
Lumbar spine BMD Age (years) Age at menarche (years) Height (cm) Weight (kg) Body mass index (k/m2) Fat mass amount (kg) Percentage of body fat (%) Lean mass amount (kg) a
Univariate regression
Standardized regression
Coefficient
95% CI
−0.001 −0.024 0.005 0.004 0.006 0.004 0.002 0.009
−0.003, −0.037, 0.002, 0.002, 0.002, 0.001, −0.000, 0.005,
0.002 −0.012 0.008 0.003 0.010 0.006 0.005 0.010
ns, not significant; BMD, bone mineral density; CI, confidence interval.
Coefficient
P
−0.029 −0.222 0.181 0.269 0.177 0.182 0.104 0.274
ns B0.001 B0.001 B0.001 B0.01 B0.01 ns B0.01
T. Douchi et al. / Maturitas 34 (2000) 261–266
264
Table 3 Outcome of stepwise multiple regression analysis in premenopausal women (n = 296)a Dependent and independent variable Lumbar spine BMD Lean mass amount +Age at menarche
R2
0.075 0.110
DR 2
P
0.035
B0.001 B0.001
a Age, height, percentage of body fat, and total fat mass amount were not significant determinants of lumbar spine BMD. Weight and BMI were excluded from the analysis. BMD, bone mineral density; R 2 = (multiple correlation coefficient)2.
life cycle. Some factors have strong impact on BMD during a certain period of the life cycle, then have a reduced impact on BMD at other stages of life cycle. This concept was also supported by the present study, demonstrating that age at menarche has a strong impact on BMD in premenopausal women, while its effect disappeared after menopause. It is generally considered that body composition (lean and fat mass component) influences BMD. In the current study, it was hypothesized that the effect of overall adiposity on BMD differs between pre- and postmenopausal women. It was found that the total fat mass amount was one of the important determinants of BMD in postmenopausal women, but was a less important
Table 5 Outcome of stepwise multiple regression analysis in postmenopausal women (n =233)a Dependent and independent variable Lumbar spine BMD Years since menopause +Lean mass amount +Total fat mass amount +Height
R2
DR 2
P
0.183 0.234 0.260 0.273
0.005 0.026 0.013
B0.001 B0.001 B0.001 B0.001
a Percentage of body fat and age at menarche were not significant determinants of BMD. Weight and BMI were excluded from the analysis. BMD, bone mineral density; R 2 = (multiple correlation coefficient)2.
determinant in premenopausal women. It is of interest to determine why the effect of overall adiposity on BMD is more prominent in postmenopausal women than in premenopausal women. In a cross-sectional study such as the present study, it may be difficult to directly address the difference in the effect of adiposity on BMD between pre- and postmenopausal women. However, no difference in total fat mass amount between pre- and postmenopausal women suggests that the weight-bearing effect of fat mass may not be responsible for the difference in the effect of overall adiposity on BMD between the two groups. In postmenopausal women, aromatization of androstenedione produces estrone
Table 4 Relationships of lumbar spine BMD to baseline and anthropometric characteristics on univariate regression analysis in postmenopausal womena Dependent and independent variables
Lumbar spine BMD Age (years) Age at menarche (years) Years since menopause (years) Height (cm) Weight (kg) Body mass index Fat mass amount (kg) Percentage of body fat Lean mass amount (kg) a
Univariate regression
Standardized regression
Coefficient
95% CI
−0.009 −0.002 −0.010 0.011 0.007 0.011 0.008 0.004 0.010
−0.012, −0.005, −0.012, 0.007, 0.005, 0.004, 0.004, 0.001, 0.005,
ns, not significant; BMD, bone mineral density; CI, confidence interval.
−0.006 0.000 −0.007 0.015 0.010 0.018 0.010 0.007 0.020
Coefficient
P
−0.399 −0.104 −0.427 0.341 0.343 0.200 0.255 0.170 0.247
B0.001 ns B0.001 B0.001 B0.001 B0.01 B0.001 B0.01 B0.001
T. Douchi et al. / Maturitas 34 (2000) 261–266
which is the major estrogen in postmenopausal women [7]. Although circulating estrogen levels in postmenopausal women are lower than those in premenopausal women, circulating estrogen levels derived from aromatization in obese women appear to enhance BMD from the perspective of the estrogen threshold hypothesis [8]. In fact, there are several data indicating elevated BMD in postmenopausal women with obesity [1 – 5,9]. In premenopausal women with regular menstruation, the effect of estrogen from the extraglandular aromatization on BMD may be masked by greater amount of ovarian estrogens. The observation agrees with previous report [10], demonstrating that body composition has no direct influence on BMD in normal premenopausal women. However, it appears that, in women with prolonged duration of amenorrhea, circulating estrogen from extraglandular aromatization becomes an important preventing factor for bone mineral loss [10 – 12]. Drinkwater et al. [10] demonstrated that obesity becomes more important as the severity of menstrual irregularities increased. The observation disagrees with the report by Aloia et al. [13], demonstrating that adiposity does not play a major role in preventing bone loss. However, they did not separate postmenopausal women from premenopausal women in their analysis. In addition, their observation was mainly based on BMI, which is indirect method of estimating overall adiposity [14]. Finally, normal premenopausal women are prone to undergo fewer changes in BMD in response to various risk factors for bone mineral loss than postmenopausal women, because this group also has several factors inhibiting bone mineral loss that postmenopausal women do not have. These factors may include youth, functioning ovary, greater daily physical activity, healthier daily diet, greater muscle strength [15], and higher baseline BMD. Based on these results, it is concluded that the effect of overall adiposity on BMD is more prominent in postmenopausal women than in premenopausal women. It is speculated that the effect of aromatized estrogen on BMD becomes evident in the situation of profound and pro-
265
longed depletion of ovarian estrogens. In premenopausal women, however, the effect of estrogen from extraglandular aromatization on BMD may be masked by greater amount of ovarian estrogen.
References [1] Dawson-Hughes B, Shipp C, Sadowski L, Dallal G. Bone density of the radius, spine and hip in relation to percent of ideal body weight in postmenopausal women. Calcif Tissue Int 1987;40:310 – 4. [2] Ribot C, Tremollieres F, Pouilles JM, Bonneu M, Germain F, Lowret JP. Obesity and menopausal bone loss: the influence of obesity on vertebral density and bone turnover in postmenopausal women. Bone 1987;8:327 – 31. [3] Hassager C, Chritiansen C. Influence of soft tissue body composition on bone mass and metabolism. Bone 1989;10:414 – 9. [4] Shiraki M, Ito H, Fujimaki H, Higuchi T. Relation between body size and bone mineral density with special reference to sex hormones and calcium regulating hormone in elderly women. Endocrinol Jpn 1991;38: 343 – 9. [5] Douchi T, Yamamoto S, Nakamura S, Oki T, Maruta K, Nagata Y. Bone mineral density in postmenopausal women with endometrial cancer. Maturitas 1999;31: 165 – 70. [6] Douchi T, Yamamoto S, Nakamura S, Ijuin T, Oki T, Maruta K, Nagata Y. The effect of menopause in regional and total body lean mass. Maturitas 1998;29: 247 – 52. [7] Judd HL, Samonki IM, Frumar AM, Lagasse LD. Origin of serum estradiol in postmenopausal women. Obstet Gynecol 1982;59:680 – 6. [8] Barbieri RL. Hormone treatment of endometriosis: the estrogen threshold hypothesis. Am J Obstet Gynecol 1992;166:740 – 5. [9] Goulding A, Taylor RW. Plasma leptin values in relation to bone mass density and to dynamic biochemical markers of bone resorption and formation in postmenopausal women. Calcif Tissue Int 1998;63:456 – 8. [10] Drinkwater BL, Bruemner BS, Chestnut CH III. Menstrual history as a determinant of current bone density in young athletes. J Am Med Assoc 1990;263:545 – 8. [11] Warren MP, Brook-Cunn J, Fox RP, Lancelot C, Newman D, Hamilton WG. Lack of bone accretion and amenorrhea: evidence for a relative osteopoenia in weight-bearing bones. J Clin Endocrinol Metab 1991;72: 847 – 53. [12] Biller BMK, Baum HBA, Rosenthal DI, Saxe VC, Charpie PM, Klibanski A. Progressive trabecular os-
266
T. Douchi et al. / Maturitas 34 (2000) 261–266
teopenia in women with hyperprolactinemic amenorrhea. J Clin Endocrinol Metab 1992;75:692–7. [13] Aloia JF, McGowan DM, Vaswani AN, Ross P, Cohn SH. Relationship of menopause to skeletal and muscle mass. Am J Clin Nutr 1991;53:1378–83. [14] Ley CJ, Lees B, Stevenson JC. Sex- and menopause-asso-
.
ciated changes in body-fat distribution. Am J Clin Nutr 1992;55:950 – 4. [15] Douchi T, Kosha S, Kan R, Nakamura S, Oki T, Nagata Y. Predictors of bone mineral loss in patients with ovarian cancer treated with anticancer agents. Obstet Gynecol 1997;90:12 – 5.