1 Introduction

In the elderly, osteoporosis and related proximal femur fracture is becoming a major health problem as the number of older people in the world increases.¹ Osteoporotic proximal femur fracture has potential serious implications for morbidity, mortality disability, impairment in quality of life as well as subsequent increasing hospital and social costs²Population aging has become a serious challenge that the world is facing, reliable data show that 11% of the world's population is over 60 years of age, with a projected increase to 22% of the population by 2050,³ Osteoporosis will assume even greater significance for health care. So a comprehensive understanding of the epidemiology of proximal femur osteoporosis will be indispensable for appropriate health care prevention and planning of osteoporosis and osteoporotic fracture in the developing country.

China is a developing country with a population of 1.3 billion, the environment, socioeconomic levels, and culture of which are very different from the other places in the world. Although there are much research of osteoporosis of proximal femur in other countries, the studies in China are not enough.^4–6

According to the definition given by World Health Organization (WHO), osteoporosis is defined as a bone mineral density of 2.5 more standard deviations (SD) below the normal race-and sex-matched reference population's mean peak bone mineral density, (PBMD).⁷ Using dual-energy X-ray absorptiometry (DXA) to measure bone mineral density (BMD) has been proved to be an effective method for diagnosing osteoporosis and assessing the risk of osteoporotic proximal femur fracture.⁸ Although a few studies have focused on accumulating proximal femur BMD reference data, similar databases in the Chinese population are very poor. Establishing more bone mineral density (BMD) reference data base and monitoring its change are essential for the prevention, diagnosis, and treatment of osteoporosis and related proximal femur fracture in China.The goal of our study was to establish BMD reference database of a Chinese female population and compare it with Lebanese⁹ and USA white women,¹⁰ for providing information on the prevalence of osteoporosis of proximal femur in China.

2 Materials and methods

3 Results

3.2

3.2 Reference database of bone mineral density of proximal femur

Table 2 presents the reference data for the BMD values, grouped according to age of the Chinese women and the conversion of measured BMD values into standardized BMD (SBMD). The peak BMD occurred in the age range 30–39 years for the femoral neck and Ward's triangle, and 40–49 years for the Trochanter. Between the peak BMD values and those of the age group 70–79 years at the same sites, overall declines were observed of 25.2% for the femoral neck, 23.1% for the trochanter, and 38.1% for Ward's triangle(p < 0.001) The BMD values for femoral neck, trochanter and Ward's triangle site were converted to SBMD using a published conversion formula,¹¹ which was proved effective.^13,14

Table 2 Reference data for bone mineral density (BMD, g/cm²) at various proximal femur sites.

Age	Number	Femoral neck	Trochanter	Ward's triangle
BMD
20–29	26	0.810 ± 0.154	0.648 ± 0.167	0.638 ± 0.180
30–39	93	0.858 ± 0.118	0.657 ± 0.107	0.698 ± 0.127
40–49	306	0.847 ± 0.124	0.661 ± 0.104	0.655 ± 0.131
50–59	407	0.775 ± 0.125	0.609 ± 0.109	0.570 ± 0.123
60–69	393	0.690 ± 0.112	0.550 ± 0.099	0.490 ± 0.107
70–79	281	0.642 ± 0.116	0.508 ± 0.104	0.432 ± 0.110
SBMD
20–29	26	0.804 ± 0.152	0.644 ± 0.165	0.635 ± 0.177
30–39	93	0.851 ± 0.117	0.653 ± 0.106	0.694 ± 0.125
40–49	306	0.841 ± 0.123	0.657 ± 0.103	0.651 ± 0.129
50–59	407	0.769 ± 0.124	0.606 ± 0.107	0.568 ± 0.121
60–69	393	0.686 ± 0.110	0.547 ± 0.097	0.489 ± 0.105
70–79	281	0.639 ± 0.114	0.506 ± 0.102	0.432 ± 0.108

Fig. 1a shows the age-related declines in T-scores for BMD at the various proximal femur sites. The mean T-scores for BMD of the femoral neck, trochanter, and Ward's triangle were −1.831, −1.471, – 2.094 respectively in subjects of 70–79 years.

Close

Fig. 1 Age-related declines in T-scores for BMD at proximal femur of Chinese women and changes in proximal femur SBMD among Chinese, Lebanese, and USA white women. a. Age-related declines in T-scores for BMD at proximal femur of Chinese women. b. Age-related change in femur neck SBMD among three populations. c. Age-related change in trochanter SBMD among three populations. d. Age-related change in Ward’s triangle SBMD among three populations.

Export to PPT

4 Discussion

The purposes of our study were to determine the normal BMD of proximal femur in Chinese women and whether Chinese BMD values at various measurement sites of proximal femur were sufficiently different from USA white and Lebanese women BMD values to warrant separate Chinese women proximal femur BMD reference data.

This report outlines proximal femur BMD values from a large group of ambulatory women in China with exclusion for extremes of body weight, chronic disease and taking medications known to affect skeletal metabolism. Institutionalized or non-ambulatory subjects were not measured. Previous studies had indicated that even if the subjects only came from health organizations and advertisements, the mean values of BMD were still similar to those random selected. When the exclusion criteria of our study were compared with the USA white women from the R. B. Mazess' study, our study selected subjects with strict exclusion criteria recommended by Norland Company and used in most studies of normal BMD reference ranges (i.e., subjects without disease or medication known to affect bone health). The R. B. Mazess' study¹⁰ is a compilation from different measuring centers and different sampling frames, including all subjects, without exclusions. Thus, we repeated analysis of the BMD values of all subjects (without exclusions) and compared this with the BMD values of the selected subjects in our study; we find the exclusion of subjects did not alter the overall results.

The results of our cross-sectional study support the use of the conventional age interval of peak BMD as the basis for adult BMD reference values. There is little diminution of BMD values of most sites in the 30–49 years old women compared with young women (20–29years). In fact, longitudinal studies demonstrate decrease of BMD in premenopausal women.^15,16 In our results, femur neck BMD decreased by 0.1% annually in younger women, and trochanter BMD was stable in younger women (20–49years). At the more trabecular Ward's triangle site, decline of BMD in younger women was more obvious. There was apparent BMD diminution of Ward's triangle in 30–49 year age groups, which was 0.4% annually. We did not specifically examine the influence of menopause on the results, but bone changes in several years after menopause are more rapid than in subsequent periods, as previous research indicated.¹⁷ Our research demonstrated that femoral bone loss accelerated in menopause and continues after 60 years old but slow down a little. In older age women the bone loss of Ward's triangle is the most serious, which has the lowest BMD. Opposite, the bone loss of trochanter is lightest, which has the highest BMD. Generally, BMD increases with age until a peak is reached and then decreases. In women, the decrease is particularly evident beginning at menopause. Our study showed that bone mass rapidly decreased in women beginning at age 50, in concordance with previous studies.^18–20 Also, similar to the results shown that the greatestrate of bone loss was at Ward's triangle in Chinese females.^21,22The peak BMD occurred in subjects aged 30–39 years at this site, which was earlier than the peak at trochanter. A possible reason is that Ward's triangle is primarily composed of trabecular bone, the density of which is known to decrease earlier than cortical bone.^23,24 Indeed, as described in other report,²⁵ bone fast loss after 50 years old may be attributable to the change of ratio of androgen to estrogen, which decreases in women.²⁶

Among Chinese women, peak BMD of femoral neck and Ward's triangle all occurred at age 30–39 years but 10 years earlier in USA white and Lebanese women. At the trochanter, peak BMD in Chinese and Lebanese women occurred at 40–49 years, which were later 20 years than in USA white women. Hammoudeh et al. found that the peak BMD of the femur neck and trochanter for the Qatari women was reached in the age group 40–49 years.²⁷ Similarly, studies of Sri Lanka females also found that peak BMD of femur neck occurred at 50 years.²⁸

These differences on the age of peak BMD might be associated with the difference in such factors as genetics, nutrition, lifestyle factors, and physical activity.^29–31 We found body weight profoundly influenced the proximal femur BMD of Chinese women. Some earlier studies have made the same point.^32,33The effect of weight was smaller in our sample of women over age 50 years. In multiple regression that included age, height and weight, the effect of per year, cm and kg were 1:2:6 at neck,2:4:6 at trochanter and 3:2:3 at ward's triangle in younger years age group, and 7:2:3 at neck,5:0.2:5 at trochanter and 7:1:2 at ward's triangle in older years age group. So that patients with low body weight, and hence low BMD, could be misconstrued as exceptionally “abnormal” unless an adjustment for body weight was considered. This indicates the potential importance of considering body weight in patient evaluation. Our results also show that body weight is an important factor, as well as age, in postmenopausal BMD and in the frequency of low BMD in Chinese women.

BMD values measured by different DXA manufactures cannot be compared directly because of technical differences in the devices. However, we were able to compare BMD values through converting the measured BMD values into standardized BMD (SBMD). The BMD values for femoral neck, Ward's triangle, and trochanter site were converted to SBMD using a published conversion formula of Hui et al.,¹¹ which was proved valid by studies.^13,14

When our data were compared with USA white and Lebanese women reference data after converted to SBMD. Changes in SBMD with age in Chinese women generally mirrored the pattern established for USA white and Lebanese women, although mean SBMD values in Chinese were generally lower. In the perimenopausal and postmenopausal decades (age 50–79 years), the decrease of all sites' SBMDs was similar to USA white values. Not only Chinese women began adult life with lower SBMD than Lebanese populations, but also there appeared to be a more dramatic SBMD decrease in the femur in postmenopausal years compared with Lebanese women. Those maybe reflect a biological difference of three populations in the rate of bone loss.

Compared with USA white women,¹⁰ SBMD values at all age groups were lower in Chinese women than in USA white women, and data on height and weight of all age group were also smaller values in Chinese women than those in USA white women. Although the body size (weight, and BMI) of Chinese women was smaller than that of Lebanese women in all age groups, among Chinese women aged 30–49 years, the SBMD of femur neck was higher than that in Lebanese women. The peak SBMD values of Chinese women were higher than those of Lebanese women for femur neck. the results suggest that it is difficult to explain BMD differences only in terms of body size, although it might explain the Asians had smaller bones, thicker and denser cortices, and more plate-like trabeculae than Caucasians.³⁴

Even though Chinese women had a higher peak BMD values than Lebanese in femur neck, bone loss was more rapid after the age of 50 years than in USA white and Lebanese women in all proximal femur sites, as determined by comparison with the respective reference databases. From the peak values to the values in the age group 70–79 years, the differences in BMD values for femoral neck, trochanter, and Ward's triangle were about 5%, 14%, and 18% lower than in USA white females and about 6%, 19%, and 20% lower than in Lebanese females, respectively. Thus, observed differences in bone loss may, at least in part, be caused by genetic differences as well as nutritional and lifestyle factors. There was study shown that birth weight had an effect on adolescent bone mass, but less than later growth and BMI in childhood and at adolescence,and underweight was consistently associated with lower BMC and aBMD z-scores.³⁵ China has experienced rapid socioeconomic growth with prominent nutrition transformations since the 1970s. Older Chinese (>40 years) were, when they were children, exposed to a relative nutritional deficit during the Chinese civil war, subsequent great Chinese famine and the cultural revolution (1942–1977), whereas younger Chinese (<40 years) have not had such an experience. The rapid socioeconomic improvement in China has also resulted in earlier age at menarche, fewer children, less breastfeeding, more abortions and later age at menopause. Previous studies observed that late menarche,³⁶ long-term breastfeeding³⁷ and early menopause³⁸ were associated with bone loss. It is therefore possible that such changes are also manifested in the different BMD values of younger and older generations of Chinese, and similarly that the environmental exposure of younger Chinese now resembles that of their USA white and Lebanese counterparts. However, since our study is cross sectional, it may not reflect the long term environmental factors' effects.

Regression analyses showed that weight was a significant predictor of female proximal femur BMD for both the premenopausal and postmenopausal decades. Many other studies have shown the effect of body weight on BMD.³⁹ Chinese women averaged 6.3 kg below Lebanese women and 5.5 kg below USA white women in body weight, a difference which is almost commensurate with the observed BMD difference of Chinese with Lebanese and USA white women in femoral neck and trochanter (0.4–0.5% per kg). The effect of weight on BMD of ward's triangle ranged from 0.2% to 0.3% per kilogram. There was only a modest effect (0.2–0.3% per kg) of weight on ward's triangle of Chinese women. It was not surprising, because those measures of body size are associated with BMD to a greater extent in more weight-bearing sites than in less weight-bearing sites.⁴⁰ Chinese women averaged 157.4 cm in height compared with 155.8 cm for Lebanese and 162.0 cm for USA white women. The effect of height on BMD was generally less than that of weight, and when height and weight were entered into a multiple regression equation, height was generally not significant (0.1–0.2% per cm in most sites). Body mass index (BMI) showed a pattern similar to weight in predicting proximal femur BMD in Chinese women, and was a significant predictor of Chinese female BMD at all proximal femur measurement sites.

The present study established prevalence rates of osteoporosis in a Chinese population. Our values were used to calculate the conventional “osteoporosis” threshold at −2.5 SD which was 0.563 g/cm2,0.401 g/cm2,0.382 g/cm2for femur neck, trochanter and ward's triangle respectively. Of the 1081 women in our sample between 50 and 79 years of age, about 6.7–17.4% were below the conventional −2.5 SD threshold of osteoporosis, and after standardized using Chinese population of 2012 about 5.6–14.9%. The percentage was low (2.9–5.7%) in the immediate postmenopausal decade; it was about 7.1–17.0% in women 60–69 years, and 14.2–34.9% in women 70–79 years. The prevalence rate at Ward's triangle in women age 50–59 was about 2.8% more than at neck and 4.5% more than at trochanter, respectively, whereas after the age of 70 years, the prevalence more rapidly increased at Ward's triangle than at neck and trochanter. The prevalence rate at Ward's triangle in women was 8.9% more than at neck and 20.7% more than at trochanter. The use of neck BMD as a standard region has been considered. Our results showed that the percentage of women below −2.5 SD at this site was 2.9%, 10.2%, and26.0% at 50–59, 60–69, and 70–79 years, respectively. These values are intermediate to the lower values for ward's triangle BMD and the higher values for trochanteric BMD. The prevalence of femur neck abnormality with age of Chinese women was comparable to the USA white and Lebanese women. The prevalence of osteoporosis at the femur neck is lower in Chinese women than in their USA white age-matched counter-parts, whereas higher than Lebanese women. The Lebanese women aged 50–79 years have a very lower prevalence of abnormality (2 versus 11.6%) than their Chinese peers.

There are several limitations of the present study. Our data were cross-sectional in nature and thus provide a survey of current BMD values by age but only an estimate of actual mean BMD changes through time. The subjects were recruited from advertisements and through health organizations and therefore were not randomly selected. This might result in selection bias, which might have influenced the results. These two limitations hold for nearly all DXA reference data. The study subjects more came from urban regions, which might have caused regional differences in the proximal femur BMD values and the anthropometric measurements. Thus, our results could be somewhat limited when applied to a general population. We used WHO guidelines for determining the prevalence of osteoporosis, even though these guidelines were established for Caucasian women and may not be entirely applicable to other ethnic groups, Finally, the number of 20–39 years age groups were quite small (n = 26 and 93) and thus may not be entirely representative of the relevant Chinese female population. Clearly, more investigation of relevant Chinese female subjects should be needed. Despite these limitations, our study was based on a large scale Chinese female population, and the sample size was sufficiently large, so oporosis in the Chinese population.

5 Conclusion

In conclusion, our study is based on a large sample population and thus provides DXA-based BMD reference data for the proximal femur. The results will aid in determining the prevalence of osteoporosis in China. The BMD of Chinese subjects tends to mirror rates of loss established for USA white and Lebanese subjects. However, Chinese subjects have somewhat lower mean BMD at proximal femoral sites than subjects included in the USA white and Lebanese reference data. The PBMD of Chinese women happened later than USA white and Lebanese. The loss of BMD in Chinese women accelerates in menopause and continues after 60 years old but slow down a little. Weight was a significant predictor of BMD at all sites in all age, and was more important than height, which was generally not significant when included with weight in multiple regression analyses. BMI had a same effect as weight. Age was also a significant predictor of BMD at all sites in older age group, but not in younger age group. Lower BMD values for Chinese subjects didn't always result in a substantially higher prevalence of osteoporosis, which higher than Lebanese but lower than USA white. Results suggest that Chinese women BMD values of proximal femur may be sufficiently different from USA white and Lebanese values to warrant a separate reference sample with which to compare individuals for the purpose of diagnosing osteoporosis. However, the practical consequences of our study in relation to proximal femur fracture rates of the Chinese women, secondary to BMD in the osteoporotic range, remain to be established.

Funding information

This research did not receive any specific grant from funding agencies in the public, commercial or not-profit sectors.

Compliance with ethical standards.