Aging and bone health in Singaporean Chinese pre-menopausal and postmenopausal women

Maturitas 89 (2016) 16–21

Contents lists available at ScienceDirect

Maturitas journal homepage: www.elsevier.com/locate/maturitas

Aging and bone health in Singaporean Chinese pre-menopausal and postmenopausal women Victor Hng Hang Goh ∗ , William George Hart Curtin Medical School, Faculty of Health Sciences, Bldg 400, Curtin University, Kent Street, Bentley, WA 6102, Australia

a r t i c l e

i n f o

Article history: Received 12 January 2016 Received in revised form 4 April 2016 Accepted 6 April 2016 Keywords: Physical exercise Age Bone mineral density Percent body fat Singaporean Chinese women

a b s t r a c t Objective: The study evaluated relationships between menopausal statuses, hormone replacement therapy (HRT), body mass index (BMI), percent body fat (PBF), and exercise with osteoporosis and bone mineral density (BMD) in Singaporean women. Study design: This is a cross-sectional study. Main outcome measures: The spine BMD, and femoral neck BMD as well as the prevalence of osteoporosis are the main outcome measures studied. Results: Age, BMI, PBF and exercise intensity were independently associated with spine and femoral neck BMD. Women with higher BMI and lower PBF had higher BMD and lower prevalence of osteoporosis. Postmenopausal women without HRT had lower BMD and higher prevalence of osteoporosis while those on HRT had similar BMD and prevalence of osteoporosis as premenopausal women. Conclusion: This study shows that BMI and PBF are powerful predictors of BMD. Osteoporosis is sitespecific in the Singapore population, being higher in the femoral neck than in the lumbar spine. The bone status after menopause may not be worse than that dictated by age alone and both ERT and E/PRT could sustain the BMD to levels corresponding to those of women a decade younger. A strategy to improve bone health should include dieting and physical exercise program that focuses on selectively reducing fat mass and increasing lean mass. © 2016 Elsevier Ireland Ltd. All rights reserved.

1. Introduction In postmenopausal women, estrogen is a major factor in the pathogenesis of postmenopausal osteoporosis as the effects of declining estrogen levels in female skeletal health is well established in both human and animal models [1,2]. In addition to age and menopausal status, several other factors including weight bearing exercises and bodyweight are known to be associated to osteoporosis [3,4]. The varying associations of these factors in different populations may contribute to the wide range of prevalence of osteoporosis, from 10% to more than 50% for women above 50y old in different countries [5]. With a rapidly aging population, osteoporosis is becoming a major public health threat to the elderly, especially women [6]. As with the rest of the world, osteoporosis-associated hip and vertebral fracture rates have been rising in most Asian countries [7,8].

∗ Corresponding author. E-mail addresses: [email protected], [email protected] (V.H.H. Goh). http://dx.doi.org/10.1016/j.maturitas.2016.04.004 0378-5122/© 2016 Elsevier Ireland Ltd. All rights reserved.

Early preventive measures are required to address this huge health problem. Osteoporosis is also known as a silent disease, and its pathological changes remain unnoticed until a fracture occurs or when a bone scan is carried out [9]. Furthermore, osteoporosis apart from being age-related is gender-, bone site- and population-specific [10]. Therefore, the present study sought to evaluate in a sample of healthy community dwelling Singaporean Chinese women, the prevalence of osteoporosis, and how age, menopausal status, bodyweight, obesity and engagement in regular physical exercise are associated with the prevalence of osteoporosis and bone mineral density (BMD). A better understanding of the interrelationships of these factors on BMD will assist in the formulation of appropriate recommendations for aging women to delay or reduce the prevalence of osteoporosis. In addition, the study afforded the opportunity to compare the BMD of Singaporean with those for other Asian and American populations.

V.H.H. Goh, W.G. Hart / Maturitas 89 (2016) 16–21

2. Subjects, materials and methods 2.1. Subjects This study was approved by the Institutional Review Board of the National University Hospital, Singapore and each volunteer gave her written informed consent. One thousand three hundred and twenty-six Singaporean Chinese women, aged between 29y and 71y, were included in the analyses. These were communitydwelling healthy individuals with no known history of major medical illnesses such as cancer, hypertension, thyroid dysfunction, diabetes, or cardiovascular events nor major sleep disorders including sleep apnea. None of the subjects had a history of major joint surgery, or bone fracture. None had been clinically diagnosed as having any form of osteoporosis. Subjects were not paid for participation. They represented the diverse spectrum of people in Singapore, ranging from those with low to high levels of education, working and non-working, and those in various types of occupations [11]. Their profiles were typical of Singapore, which is a highly urbanized city-state with no rural population. 2.2. Questionnaire Each subject answered a self-administered and investigatorguided questionnaire to collect demographic, previous medical history and lifestyle factors. The questionnaire also allowed participants to record their physical exercise and/or sport as a lifestyle habit for each week. Only engagement in the physical exercise and/or sport for at least six months was considered a lifestyle habit. Participants recorded either no exercise or up to 4 different types of exercise/sport in which they were engaged in per week. For each exercise/sport type, they stated the duration and frequency per week. For example, a participant could record walking for 30 min 5 times a week and playing tennis for 60 min once a week. 2.3. Exercise/sport intensity In order to normalize the different types of exercise/sport into a single common score, exercise intensity was calculated using the Metabolic Equivalent of Task (MET) of each exercise/sport type. A score was then calculated to denote the intensity of exercise/sport per week in MET minutes (METmin) by taking into account the duration of each exercise/sport episode and the frequency of the exercise/sport per week in accordance with the exercise guidelines [12]. For example, a participant reported that she walked for 30 min 5 times a week, played tennis for 60 min twice a week and did line dancing for 60 min once a week. Walking is assigned a MET of 3; tennis, a MET of 7; and line dancing, a MET of 5. Her total exercise/sport activities intensity per week was therefore 1590 METmin [(3 × 30 × 5) + (7 × 60 × 2) + (5 × 60 × 1)]. 2.4. Age groups (AgeGp) For comparisons, all women were divided into four age groups: AgeGp1: ≤40y; AgeGp2: 41-50y; AgeGp3: 51-60y; and AgeGp4: >60y.

17

for at least the past year (mainly on 0.625 mg Premarin (Pfizer, NY) ; MenoGp4: postmenopausal women who were on estrogen only replacement therapy (ERT) for at least the past year (mainly on progynova, (2 mg estradiol valerate, Schering AG, Berlin); and MenoGp5: postmenopausal women on estrogen/progestin hormone replacement therapy (E/PRT) for at least the past year (mainly Prempak C (0.625 mg Conjugated estrogen and 0.15 mg norgestrel, Pfizer Ltd., NY) and only two subjects were on Climen (2 mg of estradiol valerate and 1 mg cyproterone acetate, Schering AG, Berlin). 2.6. Body mass index As suggested in previous work, BMI is a good index of bodyweight normalized for height, but not of body fat [13]. The BMI was computed by taking bodyweight in kilograms divided by the square of height in meters. 2.7. Bone scan: and osteoporotic groups Each subject underwent a whole body scan, a lumber spinal scan at the L2-L4, and a scan of the hip (representing the femoral neck, shaft, and trochanter) using DXA (DPX-L, Lunar Radiation, Madison, WI, USA; software version 1.3z). The DXA scan was used for clinical management and has routinely been calibrated with the phantom. Total percent body fats (PBF), lumbar spine bone mineral density (SBMD) (average BMD of L2-L4) and femoral neck bone mineral density (FnBMD) were computed automatically by the DXA scanner. The T-scores for SBMD and FnBMD were computed with reference to the young reference values established for the local population using the DXA scanner. According to the WHO guidelines, a T-score >−1.00 is normal, while T-scores of <−2.50 are considered as osteoporosis [18]. Hence, the following 4 groups were classified as SOstGp1 (normal spinal BMD) and FnOstGp1 (normal femoral neck BMD) when the spine and femoral neck T-scores were >−1.00, SOstGp2 (osteoporosis of spine) and FnOstGp2 (osteoporosis of femoral neck) where the spine and femoral neck T-scores were <−2.50. 2.8. Statistical analysis Statistical analyses were performed using SPSS for windows version 19.0. Basic descriptive statistics, as well as multivariate linear comparison of means using the General Linear Model coupled with the Bonferroni as the Post-Hoc test for multiple comparisons for the bone parameters of SBMD, FnBMD, age, BMI, percent body fat, exercise and menopause groups were calculated. Where appropriate, the exercise intensity (METmin), BMI, and PBF and age were analyzed as covariates in the comparison groups. Linear regression analyses were carried out separately for SBMD and FnBMD using the stepwise method with age, METmin, PBF, and BMI and weighted for Menopause status. Cross-tab and Fisher Exact tests were used to assess the prevalence of spine and femoral neck osteoporosis in the different age, BMI, percent body fat, exercise intensity and menopause groups. 3. Results

2.5. Menopausal groups (MenoGp) All women who were still having their menstrual periods were classified as premenopausal women. All women were had ceased their menstrual periods for at least a year were classified as postmenopausal women. All women were also divided into five groups: MenoGp1: all premenopausal women; MenoGp2: postmenopausal women not on any hormone replacement therapy; MenoGp3: postmenopausal women on estrogen only replacement therapy (ERT)

Table 1 shows that both spine and femoral neck BMD were independently and negatively associated with age with FnBMD apparently more associated with age than SBMD (Table 1). Body mass index was highly and positively correlated with both SBMD and FnBMD, while percent body fat was highly but negatively correlated with both SBMD and FnBMD (Table 1). Exercise intensity (METmin) was positively associated only with FnBMD but not with SBMD (Table 1).

18

V.H.H. Goh, W.G. Hart / Maturitas 89 (2016) 16–21

Table 1 Linear regressions using the step-wise method of variables with Spine Bone Mineral Density (SBMD) and Femoral neck Bone Mineral Density (FnBMD). Variables

Age (y) BMI (kg/m2 ) PBF (%) METmin

Spine Bone Mineral Density (SBMD)

Femoral Neck Bone Mineral Density (FnBMD)

Standardized Coefficients Beta

p-value

Standardized Coefficient Beta

p-value

−0.249 0.505 −0.424 0.006

<0.001 <0.001 <0.001 NS

−0.261 0.472 −0.309 0.049

<0.001 <0.001 <0.001 0.0470

Table 2 Young reference (20–30y women)values for SBMD and FnBMD of different population. Spine BMD (g/cm2 )

Femoral Neck BMD (g/cm2 )

Singaporean [35]

1.098 ± 0.156

0.942 ± 0.195

Non-Hispanic White [36]

1.047 ± 0.110

0.858 ± 0.120

Japanese [36]

1.006 ± 0.115

0.803 ± 0.107

Thai [36]

0.957 ± 0.110

0.814 ± 0.098

Korean [28]

0.959 ± 0.127

0.771 ± 0.127

Table 2 shows that the both SBMD and FnBMD of Singaporean Chinese women aged 20–30y were higher than the young reference values for Non-Hispanic White, Japanese, Korean and Thai women (Table 2). Overall, the prevalence of osteoporosis for lumbar spine (2.5%) was lower than corresponding prevalence for the femoral neck (12.5%) (Table 3). The prevalence of spine osteoporosis was significantly increased after age 50y, and an age “dose-response” relationship was noted, with prevalence in the >60y age group being significantly higher than corresponding prevalence in the >50y age group (Table 3). The prevalence of femoral neck osteoporosis, however, only increased significantly after age 60y (Table 3). Both SBMD and FnBMD were age-dependent. After adjusting for BMI, PBF, exercise intensity (METmin) and weighted for menopausal status, both SBMD and FnBMD were significantly lower only after age 50y by 7.2% and 5.5%, respectively, when compared to the younger, ≤40y age group, and were significantly and further lowered, by 13.2% and 12.9%, respectively in the >60y age group (Table 3). Table 4 shows the influence of the menopausal status of women on the prevalence of oesteoporosis. After adjusting for age, BMI, PBF and METmin, both SBMD and FnBMD in postmenopausal women (MenoGp2) who were not on any hormone replacement therapy (HRT), were significantly lower, by 9.2% and 3.4%, respectively, when compared to premenopausal women (MenoGp1) (Table 4). It is interesting to note that the average SBMD and FnBMD of postmenopausal women without any HRT (MenoGp2) were ageappropriate, that is the SBMD and FnBMD were similar to those in their age group of 51–60y (Table 3). On the other hand, both SBMD and FnBMD of postmenopausal women who were on estrogen replacement therapy (ERT), either on premarin (MenoGp3) or estradiol valerate (MenoGp4) were not significantly different from the premenopausal women (MenoGp1)(Table 4). The SBMD of postmenopausal women on estradiol valerate (MenoGp4) was significantly higher than corresponding value in postmenopausal women not on any HRT (MenoGp1, Table 4). Postmenopausal women on E/PRT (MenoGp5), on the other hand, had significantly lower SBMD than those in premenopausal women (MenoGp1, Table 4). The SBMD and FnBMD in postmenopausal women either on estrogen alone or on estrogen/progestogen combination HRT (MenoGp3, MenoGp4 & MenoGp5) were at levels equivalent to those in women at least a decade younger (AgeGp2, Table 3). The menopausal status-dependent differences in SBMD and FnBMD were also reflected in prevalence of osteoporosis in

women. As expected, postmenopausal women not on any HRT (MenoGp2) have significantly higher prevalence of spine and femoral neck osteoporosis when compared to premenopausal women (MenoGp1) (Table 4). On the other hand, postmenopausal women on estrogen alone HRT (MenoGp3 and MenoGp4) did not have significantly different prevalence of spinal and femoral neck osteoporosis when compared to corresponding prevalence in premenopausal women (MenoGp1)(Table 4). However, postmenopausal women on estrogen/progestogen combination HRT had significantly higher prevalence of spine and femoral neck osteoporosis when compared to corresponding prevalence in premenopausal women (MenoGp1, Table 4). 4. Discussion The present study showed that besides age, the two other important independent determinants of BMD are body mass index and total percent body fat. Regular exercise per se of sufficient intensity contributes to a small extent to the BMD. Together with age, BMI and PBF account for much of the variations seen in SBMD and FnBMD in the women in the present study. In agreement with earlier study, BMI is a powerful predictor of BMD and has a dose-response effect on both SBMD and FnBMD [14,15]. Its potentiating effect appears to exceed that of the negative association of age and percent body fat. An increase of 5 kg/m2 is reflected in an increase of 12.5% of SBMD and a 9.9% increase in FnBMD, as opposed to a decrease of 6.4% and 7.8% respectively of SBMD and FnBMD seen by a decade increase in age, from 50y to 60y. There are conflicting results concerning the association of body fat and BMD, with some reports suggesting that high adiposity is detrimental, while others beneficial to BMD [16,17]. The conflicting results arose, in part, due to the choice of the indices for adiposity. Several studies have made used of BMI, waist/hip ratio, fat mass as the index of obesity [15,18,19]. In a previous study, it was shown that BMI is more an index of bodyweight than of adiposity; when compared to DXA derived percent body fat (PBF) and waist circumference, waist/hip and waist/height ratio were not accurate indices of adiposity [13]. The use of these indices for adiposity, therefore would lead to varying levels of misclassification thereby marring the real association of body fat with BMD. The observed negative correlation of percent body fat with BMD is in contradiction with two earlier studies in which DXA-derived percent body fat was used, but they showed positive association of fat mass with BMD [16,17]. The discrepancy in results with the two latter studies may be a statistical problem. When both lean mass and fat mass were analyzed together in the multi-linear regression analysis, a collinearity problem had ensued giving fat mass a false positive association with BMD. It has been amply shown that regular physical exercise is protective of BMD and can reduce fracture risk [20]. A more sedentary lifestyle may lead to lower BMD [21]. It is therefore clear that exercise may be able to mitigate the causative effect of aging on osteopenia and osteoporosis. The mechanism for physical activityrelated increases in BMD is unclear and remains to be elucidated. The mechanical stimulation of physical exercise could modulate

V.H.H. Goh, W.G. Hart / Maturitas 89 (2016) 16–21

19

Table 3 Number (percent) of spine and femoral neck osteoporosis and SBMD and FnBMD in the 4 age groups. AgeGp1: <40y (n = 181)

AgeGp2: 41–50y (n = 533)

AgeGp3: 51–60y (n = 479)

AgeGp4: >60y (n = 133)

SOstGp1

161

443

281

48

SOstGp2

1 (0.6%)

0 (0%) Gp2vGp3,Gp4 (<0.001, <0.001)

18 (3.8%) Gp3Gp1,Gp4 (0.033, 0.009)

14 (10.5%)

FnOstGp1

78

235

136

16

FnOstGp2

11 (6.1%)

31 (5.8%) Gp2vGp3,Gp4 (<0.001, <0.001)

75 (15.7%) Gp3vGp1,Gp4 (<0.001, <0.001)

49 (36.8%) Gp4vGp1 (<0.001)

SBMD(g/cm2 )

1.053 ± 0.015

1.032 ± 0.008

0.977 ± 0.005 Gp1,2vGp3 (<0.001, <0.001)

0.914 ± 0.009 Gp1,2,3vGp4 (<0.001, <0.001, <0.001)

FnBMD(g/cm2 )

0.830 ± 0.011

0.817 ± 0.005 (−1.5%)

0.784 ± 0.004 (−5.5%) Gp1,2vGp3 (<0.001, <0.001)

0.723 ± 0.007 (−12.9%) Gp1,2,3vGp4(<0.001, <0.001, <0.001)

The prevalence of osteoporosis was tested for significance using the Fisher Exact Test. Comparisons of SBMD and FnBMD among the 4 groups were carried out using the multilinear regression analyses with age, METmin, BMI, PBF as covariates. SOstGp1 = normal spine BMD (T-score >−1.0), OstGp2 = osteoporotic spine BMD (T score <−2.5). FnOstGp1 = normal femoral neck BMD (T-score >−1.0), OstGp2 = osteoporotic femoral neck BMD (T score <−2.5).

Table 4 Number (percent) of spinal and femoral neck osteoporosis and SBMD and FnBMD in the five menopause groups. MenoGp1 (n = 729)

MenoGp2 (n = 431)

MenoGp3 (n = 61)

MenoGp4 (n = 16)

MenoGp5 (n = 89)

Age (y)

44.8 ± 0.22

55.6 ± 0.28 Gp2vGp1 (<0.001)

55.5 ± 0.71 Gp3vGp1 (<0.001)

54.2 ± 1.42 Gp4vGp1 (<0.001)

56.2 ± 0.60 Gp5vGp1 (<0.001)

SOstGp1

621

264

40

12

50

SOstGp2

4 (0.5%)

28 (6.5%) Gp2vGp1 (<0.001)

1 (1.6%)

0 (0%)

3 (3.4%) Gp5vGp1 (0.032)

FnOstGp1

304

96

14

8

25

FnOstGp2

52 (7.1%)

101 (23.4%) Gp2vGp1 (<0.001)

8 (13.1%)

2 (12.5%)

13 (14.6%) Gp5vGp1 (0.021)

SBMD(g/cm2 )

1.061 ± 0.006

0.963 ± 0.008 Gp2vGp1 (<0.001)

1.017 ± 0.019 (5.6%)

1.071 ± 0.036 (13.5%) Gp4vGp2 (0.027)

0.999 ± 0.016 (3.7%) Gp5vGp1 (0.006)

FnBMD(g/cm)

0.813 ± 0.005

0.785 ± 0.006 Gp2vGp1 (0.009)

0.806 ± 0.013 (2.7%)

0.847 ± 0.025 (7.9%)

0.806 ± 0.011 (2.7%)

MenoGp1 = premenopausal women, MenoGp2 = postmenopausal women not on HRT. MenoGp3 = postmenopausal women on ERT (premarin only HRT), MenoGp4 = postmenopausal women on ERT (estradiol valerae), and MenoGp5 = postmenopausal women on E/PRT (mainly 0.625CEE, 0.15 Norgestrel, Prempak—C). The prevalence of osteoporosis was tested for significance using the Fisher Exact Test. Comparisons of SBMD and FnBMD among the 5 groups were carried out using the multilinear regression analyses with age, METmin, BMI, PBF as covariates. SOstGp1 = normal spine BMD (T-score >−1.0), OstGp2 = osteoporotic spine BMD (T score <−2.5). FnOstGp1 = normal femoral neck BMD (T-score >−1.0), OstGp2 = osteoporotic femoral neck BMD (T score <−2.5).

bone remodelling resulting in increased mineralization of the bone [22]. The positive moderating effect of physical exercise on BMD was further shown in a longitudinal study [20]. However, the present study showed that regular physical exercise was associated with higher femoral neck but not lumber spine BMD. It is also possible that physical exercise may have greater association with BMD via its proxy effects of increasing lean mass and reducing body fat, two powerful determinants of BMD which the present study has clearly shown. Apart from age which is not modifiable, promotion of better bone health must take into account these three important and modifiable determinants of BMD. Therefore, an effective strategy for improving bone health, employing both dieting and regular exercise, must focus on ensuring not just weight loss, but concurrently, increase in lean mass and loss of fat mass. The effective physical exercise regime may include sufficient aerobic routine to improve stamina and loose fat, but also weight-bearing exercise to increase strength and lean mass. Many studies have shown ethnicity is associated with BMD [23,24]. The present study also showed that there are distinct differences in SBMD and FnBMD in different populations with Singaporean Chinese women having higher BMD than Non-Hispanic White and other Asian populations as shown by others [25,26].

However, not all studies have adjusted with the relevant co-variates especially with BMI, percent body fat and exercise intensity. It is unclear whether the international variations in BMD would diminish had these adjustments been made. There is conflicting evidence; in a previous study, ethnic differences in some bone parameters were eliminated when analyses were adjusted for weight [18]. However, in another study, ethnic difference in BMD was noted even after adjusting for many other factors including age, weight, height, but not concurrently with percent body fat [26]. Therefore, further studies on cross-population comparisons of BMD must take into account all the important covariates including age, BMI, percent body fat, exercise intensity in order to have a clear picture of the true association of ethnicity with BMD. The present study showed that the prevalence of osteoporosis in women above age 50y in highly urbanized Singapore is sitespecific, with femoral neck osteoporosis (20.3%) about 4 times those of lumbar spine osteoporosis (5.2%). Lumbar spine osteoporosis was comparable, while femoral neck osteoporosis was higher than those in American women (9.9% and 7.7%, respectively) [27] and lower than the very high prevalence of spinal and femoral neck osteoporosis, 30.1% and 23.1% respectively in Korean women above 50y [28] and other population groups [5]. Reasons for the low rates

20

V.H.H. Goh, W.G. Hart / Maturitas 89 (2016) 16–21

of osteoporosis of both the spine and the femoral neck as compared to other populations are unclear. Accelerated bone loss has been suggested as a hallmark of the menopause syndrome [29]. However, the present study showed that after adjusting for the confounding factors of BMD including age, intensity of physical exercise, and importantly BMI and percent body fat, postmenopausal women not on HRT had SBMD and FnBMD and rates of osteoporosis of the spine and femoral neck similar to the age group of 51–60y that they belong to. Therefore, contrary to the earlier suggestion, an accelerated bone loss was not evident in postmenopausal women and that the loss of bone mass in postmenopausal women could be accounted for by age alone. On the other hand, the present study shows that both ERT or E/PRT are beneficial to BMD, they have the potential to sustain the BMD to levels equivalent to those of women a decade younger. The results confirmed earlier suggestions that both estrogen alone and estrogen/progestin hormone replacement therapy are beneficial to bone health [29,30]. The results also indicate that estrogen alone HRT may be slightly more beneficial to bone health than estrogen/progestin HRT. Further, the results of the present study suggest that those on estradiol valerate has slightly better benefit to BMD than those of conjugated equine estrogens. However, the differential benefits of estrogen alone and estrogen/progestin HRT are difficult to elucidate from results of the present study. A review and another study suggested that the combined estrogen/progestin HRT in postmenopausal women has beneficial effects on bone, and generally, that progestin does not moderate to any extent the protective action of estrogens on bone [14,20]. In addition, from the meta-analysis by Doren et al. (2003) [31], what is clear is that all estrogens irrespective of the route of administration are effective in maintaining BMD. Our results concurred with this observation. Furthermore, although not statistically significant, it was noted that the improvement of SBMD of women on ERT was better than for FnBMD, an observation also noted in several meta-analyses [32]. The use of either estrogen alone or estrogen/progestin combination HRT is protective of bone health in older women. However their uses must be weighed against the potential adverse impact of HRT on other health compartments, especially that of the cardiovascular compartment. The risk-benefit of HRT is complex and its use requires the consideration of the presence of risk factors of other diseases [33]. One of the strengths of the present study is the large number of normal healthy community dwelling women without any major illnesses who had participated in the study. The use of BMI as an index of bodyweight and the use of DXA scan as an index of percent body fat had allowed the clarification of the independent associations of bodyweight and body fat with bone density. The use of the METmin as the index of the intensity of physical exercise/sport allowed the normalizing of physical activities and their frequency and duration to a single index. However, a limitation of this computed exercise intensity is that it reflects mainly the aerobic component and has lesser indication of the weight-bearing component which may be more relevant to bone health. In summary, this study shows clearly a distinctive difference between the association of body mass index and the percent body fat. Both BMI and PBF are respectively powerful positive and negative predictors of BMD consistent with some earlier studies [14,34]. These two determinants together with age and exercise intensity may account for much of the variations in BMD observed. For any cross-ethnic comparisons of BMD, therefore, these factors must be taken into account in order to reflect more clearly the true ethnic variations. Osteoporosis is clearly site-specific in the Singapore population and various determinants including age, BMI, PBF and exercise intensity appeared to affect SBMD and FnBMD differently. The bone status after menopause may not be worse than that dictated by age alone and both ERT and E/PRT could sustain the BMD

to levels corresponding to those of women a decade younger. Based on findings of the present study, a strategy to maintain or improve bone health should include dieting and physical exercise program. The diet plan should be focus on selectively reducing fat mass and not lean mass. The exercise program should include an aerobic component to improve stamina and reduce body fat, and also a weight bearing component to improve strength and increase lean mass. This comprehensive program will not only be beneficial to bone but also cardiovascular health. Contributors to the paper The project was designed, executed and data collected and analyzed by Professor Victor Goh. Professor William Hart had provided valuable input in interpretations and final draft of the paper. Declaration of interest Hng Hang Victor Goh and William George Hart declare that they have no conflict of interest. This study was supported, in part, by funds from the Academic Research Fund of the National University of Singapore, Singapore. Ethics approval Approval for the project was granted by the Ethics Committee of the National University Hospital, Singapore. Funding Funding for this project was provided by the National University of Singapore under the Academic Research Funds. Acknowledgments We acknowledge the technical assistance from staff of the Endocrine Research and Service Laboratory of the Department of Obstetrics and Gynaecology, National University of Singapore, Singapore. This study was designed, conducted and data collected while Prof. Victor H. H. Goh was at the Department of Obstetrics and Gynaecology, National University of Singapore, Singapore. Prof. William G. Hart was intimately involved in the interpretation of findings, drafting and critical revision of the article for submission. References [1] B.L. Riggs, S. Khosla, L.J. Melton 3rd, A unitary model for involutional osteoporosis: estrogen deficiency causes both type I and type II osteoporosis in postmenopausal women and contributes to bone loss in aging men, J. Bone Miner. Res. 13 (1998) 763–773. [2] R. Recker, J. Lappe, K.M. Davies, R. Heaney, Characterization of perimenopausal bone loss: a prospective study, J. Bone Miner. Res. 15 (2000) 1965–1973. [3] H. Orimo, T. Nakamura, T. Hosoi, M. Iki, K. Uenishi, N. Endo, H. Ohta, M. Shiraki, T. Sugimoto, T. Suzuki, S. Soen, Y. Nishizawa, H. Hagino, M. Fukunaga, S. Fujiwara, Japanese 2011 guidelines for prevention and treatment of osteoporosis—executive summary, Arch. Osteoporos. 7 (2012) 3–20. [4] R. Rizzoli, M.L. Bianchi, M. Garabedian, H.A. McKay, L.S. Moreno, Maximizing bone mineral mass gain during growth for the prevention of fractures in the adolescents and the elderly, Bone 46 (2010) 294–305. [5] P. Pothiwala, E.M. Evan, K.M. Chapman-Novakoske, Ethnic variation in risk for osteoporosis among women: a review of biological and behavioral factors, J. Women’s Health 15 (2006) 709–719. [6] National Institutes of Health, Bone Health Information. Osteoporosis and related bone diseases From the Centers for Disease Control and Prevention. Prevalence of disabilities and associated health conditions among adults—United States 1999, JAMA 285 (2001) 1571–1572. [7] C.H. Bow, E. Cheung, C.L. Cheung, S.M. Xiao, C. Loong, C. Soong, K.C. Tan, M.M. Luckey, J.A. Cauley, S. Fujiwara, A.W. Kung, Ethnic difference of clinical vertebral fracture risk, Osteoporos. Int. 23 (2012) 879–885.

V.H.H. Goh, W.G. Hart / Maturitas 89 (2016) 16–21 [8] T. Suzuki, Risk factors for osteoporosis in Asia, J. Bone Miner. Metab. 19 (2001) 133–141. [9] F. Hannafon, M.P. Cadogan, Recognition and treatment of postmenopausal osteoporosis, J. Gerontol. Nurs. 40 (3) (2014) 10–14. [10] A.C. Looker, H.W. Wahner, W.L. Dunn, M.S. Calvo, T.B. Harris, S.P. Heyse, C.C. Johnston Jr., R. Lindsay, Updated data on proximal femur bone mineral levels of US adults, Osteoporos. Int. 8 (1998) 468–489. [11] V.H.H. Goh, T.Y.Y. Tong, H.P.P. Mok, B. Said, Interactions among age adiposity, bodyweight, lifestyle factors and sex steroid hormones in healthy Singaporean Chinese men, Asian J. Androl. 9 (2007) 611–621. [12] US Department of Health and Human Services. Physical Activity Guidelines for American (PGA) http://www.health.gov/PAguidelines/guidelines/ appendix.aspx (accessed 05.04.15). [13] V.H.H. Goh, C.F. Tain, T.Y.Y. Tong, H.P.P. Mok, M.T. Wong, Are BMI and other anthropometric measures appropriate as indices for obesity? A study in an Asian population, J. Lipid Res. 45 (2004) 1892–1898. [14] G. Barrera, D. Bunout, V. Gattas, M.P. de la Maza, L. Leiva, S. Hirsch, A high body mass index protects against femoral neck osteoporosis in healthy elderly subjects, Nutrition 20 (2004) 769–771. [15] J.T. Lloyd, D.E. Alley, W.G. Hawkes, M.C. Hochberg, S.R. Waldstein, D.L. Orwig, Body mass index is positively associated with bone mineral density in US older adults, Arch. Osteoporos. 9 (March) (2014) 175, http://dx.doi.org/10. 1007/s1657.014-0-0175-2 (accessed 25.10.15). [16] H. Lu, X. Fu, X. Ma, Z. Wu, W. He, Z. Wang, D.B. Allison, S.B. Heymsfield, S. Zhu, Relationships of percent body fat and percent trunk fat with bone mineral density among Chinese black, and white subjects, Osteoporos. Int. 22 (2011) 3029–3035. [17] X. Fu, X. Ma, H. Lu, W. He, Z. Wang, S. Zhu, Associations of fat mass and fat distribution with bone mineral density in pre- and postmenopausal Chinese women, Osteoporos. Int. 22 (2011) 113–119. [18] M. Conradie, M.M. Conradie, M. Kidd, S. Hough, Bone density in black and white South African women: contribution of ethnicity, body weight and lifestyle, Arch. Osteoporos. 9 (2014) 193. [19] P.L.S. Yang, Y. Lu, C.M. Khoo, M.K.S. Leow, E.Y.H. Khoo, A. Teo, Y.S. Lee, S. Das De, Y.S. Chong, P.D. Gluckman, E.S. Tai, K. Venkataraman, C.M.A. Ng, Associations between ethnicity, body composition, and bone mineral density in a Southeast Asian population, J. Clin. Endocrinol. Metab. 98 (11) (2013) 4516–4523. [20] J. Wee, B.Y. Sng, L. Shen, C.T. Lim, G. Singh, S. Das De, The relationship between body mass index and physical activity levels in relation to bone mineral density in pre-menopausal and postmenopausal women, Arch. Osteoporos. 8 (162) (2013), http://dx.doi.org/10.1007/s11657-013-0162-z. [21] S. Ye, A. Song, M. Yang, X. Ma, X. Fu, S. Zhu, Duration of television viewing and bone mineral density in Chinese women, J. Bone Miner. Metab. 32 (2014) 324–330.

21

[22] A.G. Robling, A.B. Castillo, C.H. Tumer, Biomechanical and molecular regulation of bone remodelling, Annu. Rev. Biomed. Eng. 8 (2006) 455–498. [23] J.A. Cauley, L.Y. Lui, K.E. Ensrud, J.M. Zmuda, K.L. Stone, M.D. Hochbeg, S.R. Cunnings, Bone mineral density and the risk of incident non-spinal fractures in black and white women, JAMA 293 (17) (2005) 2102–2108. [24] E. Barrett-Connor, E.S. Siris, L.E. Wehren, P.D. Miller, T.A. Abbott, M.L. Berger, A.C. Santora, L.M. Sherwood, Osteoporosis and fracture risk in women of different ethnic groups, J. Bone Miner. Res. 20 (2) (2005) 185–194. [25] M.A. Paggiosi, C.C. Glueer, C. Roux, D.M. Reid, D. Felsenberg, R. Barkmann, R. Eastell, International variation in proximal femur bone mineral density, Osteoporos. Int. 22 (2011) 721–729. [26] H.S. Nam, S.S. Kweon, J.S. Choi, J.M. Zmuda, P.C. Leung, L.Y. Lui, D.D. Hill, A.L. Patrick, J.A. Cauley, Racial/ethnic differences in bone mineral density among older women, J. Bone Miner. Metab. 31 (2013) 190–198. [27] A.C. Looker, L.J. Melton III, L.G. Borrud, J.A. Shepherd, Lumber spine bone mineral density in US adults: demographic patterns and relationship with femur neck skeletal status, Osteoporos. Int. 23 (2012) 1351–1360. [28] E.J. Park, I.W. Joo, M.-J. Jang, Y.T. Kim, K. Oh, H.J. Oh, Prevalence of osteoporosis in the Korean population based on Korea National Health and Nutrition Examination Survey (KNHANES) 2008–2011, Yonsei Med. J. 55 (4) (2014) 1049–1057. [29] M.R. Sowers, H. Zheng, G.A. Greendale, R.M. Neer, J.A. Cauley, J. Ellis, S. Johnson, J.S. Finkelstein, Changes in bone resorption across the menopause transition: effects of reproductive hormones, body size and ethnicity, J. Clin. Endocrinol. Metab. 98 (7) (2013) 2854–2863. [30] J.A. Cauley, J. Robbins, Z. Chen, S.R. Cummings, R.D. Jackson, A.Z. LaCroix, M. LeBoff, C.E. Lewis, J. McGowan, J. Neuner, M. Pettinger, M.L. Stefanick, J. Wactawski-Wende, N.B. Watts, Women’s Health Initiative Investigators: effects of estrogen plus progestin on risk of fracture and bone mineral density: the Women’s Health Initiative randomized trial, JAMA 290 (2003) 1729–1738. [31] M. Doren, J.A. Nilsson, O. Johnell, Effects of specific post-menopausal hormone therapies on bone mineral density in post-menopausal women: a meta-analysis, Human Reprod. 18 (8) (2003) 1737–1746. [32] J.H. Thijssen, Overview on the effects of progestins on bone, Maturitas 46 (Suppl. 1) (2003) S77–S87. [33] T.J. de Villiers, J.C. Stevenson, The WHI: the effect of hormone replacement therapy on fracture prevention, Climacteric 15 (3) (2012) 263–266. [34] T. Douchi, S. Yamamoto, T. Oki, K. Maruta, R. Kuwahata, H. Yamasaki, Y. Nagata, Difference in the effect of adiposity one bone density between preand postmenopausal women, Maturitas 34 (2000) 261–266. [35] S.T. Lee, J.C.H. Goh, S.L. Low, K. Bose, J. Musculoskelet. Res. 1 (1997) 41–46. [36] S. Namwongprom, S. Rojnastein, A. Mangklabruks, S. Soontrapa, C. Wongboontan, B. Ongphiphadhanakul, Importance of ethnic base standard references for the diagnosis of osteoporosis in Thai women, J. Clin. Densitometr.: Assessm. Skelet. Health 15 (3) (2012) 295–301.