Race/ethnic differences in bone mineral densities in older men - pdf attached
Download the PDF here|
Osteoporosis International Nov 2010
"When compared with US Caucasian men, age-adjusted mean BMD measures at all three BMD sites were 8-20% higher among Afro-Caribbean and 6-12% higher among African-American men. Hip BMD was similar among US Caucasian and Hispanic men, but spine BMD was 3% lower among Hispanic men. Hip and spine BMD values were 3-5% lower among US Asian, 7-10% lower among Hong Kong Chinese, and 8-14% lower except femoral neck among Korean men compared to US Caucasians. The differences shown above were statistically significant (p<0.001) or nearly significant (p=0.057 for femoral neck in Asian men) except for spine BMD in Hispanic or Asian men (Table 2; Fig. 1)."
H.-S. Nam1, 2, M.-H. Shin3, J. M. Zmuda2, P. C. Leung4, E. Barrett-Connor5, E. S. Orwoll6, J. A. Cauley2 Contact Information and Osteoporotic Fractures in Men (MrOS) Research Group
(1) Department of Preventive Medicine, Research Institute for Medical Sciences, Chungnam National University College of Medicine, Daejeon, South Korea
(2) Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, 130 DeSoto Street, Crabtree Hall A524, Pittsburgh, PA 15261, USA
(3) Department of Preventive Medicine, Chonnam National University Medical School, Gwangju, South Korea
(4) Jockey Club Centre for Osteoporosis Care and Control, Chinese University of Hong Kong, Prince of Wales Hospital, Hong Kong, China
(5) Department of Family and Preventive Medicine, University of California San Diego, La Jolla, CA, USA
(6) Bone and Mineral Unit, Oregon Health and Sciences University, Portland, OR, USA
Contact Information J. A. Cauley
BMD was compared across race/ethnic groups. There were substantial race/ethnic differences in BMD even within African or Asian origin. Additional adjustment for body size greatly attenuated or reversed the differences between US Caucasian men vs Asian men. It illustrates the role of body size on the difference between these groups.
There is insufficient epidemiologic information about men's bone mineral density (BMD) levels across race/ethnic groups and geographic locations.
In a cross-sectional design, we compared BMD in older men across seven race/ethnic groups in four countries. Femoral neck, total hip, and lumbar spine BMD were measured in men (age 65 to 78 years) from the Osteoporotic Fractures in Men (MrOS) Study (4,074 Caucasian, 208 African-American, 157 Asian, and 116 Hispanic men in USA), Tobago Bone Health Study (422 Afro-Caribbean men), MrOS Hong Kong Study (1,747 Hong Kong Chinese men), and the Namwon Study (1,079 South Korean men). BMD was corrected according to the cross-site calibration results for all scanners.
When compared with US Caucasian men, Afro-Caribbean and African-American men had, respectively, 8-20% and 6-11% higher age-adjusted mean BMD at all three bone sites. Hip BMD was similar in US Caucasian and Hispanic men, US Asian, Hong Kong Chinese, and Korean men had 3-14% lower BMD at all bone sites except femoral neck in Korean men. Additional adjustment for weight and height greatly attenuated or reversed the differences between US Caucasian men vs Asian men including US Asian, Hong Kong Chinese, and South Korean men. Among Asian groups, Korean men had higher femoral neck BMD and lower total hip BMD.
These findings show substantial race/ethnic differences in BMD even within African or Asian origin and illustrate the important role of body size on the difference between Asian men and others.
Osteoporosis and fractures are important health problems in older men [1, 2]. The lifetime risk of experiencing an osteoporotic fracture in Caucasian men over the age of 50 is similar to the lifetime risk of developing prostate cancer . Mortality after an osteoporotic fracture is greater in older men compared to older women [3, 4]. Considering demographic trends leading to greater numbers of older men in both developed and developing countries, the societal burden of osteoporosis in men is a major international health concern.
Many studies in US people reported that hip fracture rates among older African-American, Asian, and Hispanic men are lower than rates among Caucasian men [5-11]. Several population studies have reported that African-American men have higher bone mineral density (BMD) than US Caucasian and Hispanic men at major weight-bearing sites such as femoral neck and lumbar spine [12-15]. Age-related cross-sectional declines in BMD have been shown to be significantly steeper among US Hispanic men than African-American or US Caucasian men [14, 15]. These race/ethnic differences in BMD could contribute to the lower risk of fracture in African-American men when compared to Caucasian and Hispanic men. However, the evidence of difference in BMD between US Hispanic and Caucasian men is not consistent [13-15], and the difference between Caucasian and Asian men is also inconclusive [13, 16, 17].
Most epidemiologic reports on race/ethnic differences in men's BMD are limited to US race/ethnic groups. To extend our knowledge about race/ethnic difference in BMD, we collected datasets from one US  and three non-US bone health studies [19-21] and compared older men's mean BMD, respectively, across seven race/ethnic groups: US Caucasian, US Hispanic, US Asian, African-American, Afro-Caribbean, Hong Kong Chinese, and South Korean.
We compared hip and spine BMD in men of seven race/ethnic groups and five countries. Our results indicate that there are substantial differences in age-adjusted BMD across race/ethnic groups and countries. In age-adjusted analysis, total hip BMD distributed across Five strata: Afro-Caribbean men had the highest level; African-American men in the second; US Caucasian and US Hispanic in the third; US Asian and Hong Kong Chinese in the fourth; and Korean men had the lowest level. Although age-related change in osteophytic calcification might affect spine DXA measures, similar patterns were observed for lumbar spine BMD as well as femoral neck except for Korean men. Unlike total hip BMD, femoral neck BMD among Korean men was similar to Caucasian men.
Identification of the BMD differences across race/ethnicity and geography has important implication for understanding geographic variability in fracture risk. In general, hip BMD is strongly associated with the risk of nonvertebral fracture in older men [29, 30]. Differences in age-adjusted BMD among Asian groups are consistent with the wide variability in fracture rates across Asian countries in the Asian Osteoporosis Study (AOS) . The reported hip fracture rate among Korean men aged 70 to 79 (325 per 105 men in 2004)  is slightly higher than Hong Kong Chinese men in AOS and is compatible with the difference in total hip BMD among both groups in our study. However, total hip BMD across some race/ethnic groups in our study is not compatible with previous reports [5-11] showing that fracture rates are lower in US Hispanic and Asian men than in Caucasian men. This paradox in Asian men may be in part attributable to more favorable hip geometry (the shorter hip axis length and smaller neck shaft angle)  and bone structure (greater cortical thickness and trabecular volumetric BMD)  among this group than Caucasian men. In addition to these factors, different fall rates  across race/ethnic groups can be involved in that paradox.
The differences in BMD depend both on genetic and environmental factors across countries and race/ethnic groups . The environmental factors include social factors, as well as lifestyle factors, that could influence BMD within each community. For example, the prominent differences in total hip BMD between Korean and other Asian groups suggest differences in lifestyle and social factors in part. As shown in Table 1, the lower amount of calcium intake in Korean men may contribute to the lower total hip BMD: The difference in total hip BMD between Korean and Hong Kong Chinese men was smaller after adding dietary calcium intake into the regression model including age, weight, and height as covariates. Furthermore, the Korean men experienced relative nutritional deficits during Korean War (1950-1953) in their childhood or adolescence and might have led to lower peak skeletal mass . However, the possible genetic influence on the difference among Asian groups should also be considered.
Consistent with the report of Hill et al. , Afro-Caribbean men had 10-11% higher hip BMD than African-American men. Hill et al.  suggested two possible explanations for higher BMD in Afro-Caribbean men: Firstly, the proportion of European admixture (25%) among African-American men is more than in Tobago (6%); secondly, Tobago people have more weight-bearing activities due to the lack of industrialization than US people. As shown in Table 2, there was no change of the difference in BMD among both African origin groups before and after additional adjustment for lifestyle factors including walking. Considering this, it is thought that the proportion of European admixture is more responsible for the difference than weight-bearing activities.
The difference in BMD between US Caucasian men vs Asian groups may be explained to a great extent by body size [13, 16], although additional factors may also contribute. Body size has two kinds of implications for BMD. First, it has weight-bearing effects. The range of weight is quite different between Asian and non-Asian groups. Second, height and weight may in part correct for the confounding effect caused by bone size difference between both groups. In previous studies [16, 17], bone mineral apparent density (BMAD) measurements have been used to correct for the differences in bone size. However, recent evidence  suggests that BMAD may not address bone size differences appropriately when race/ethnic groups differ in body size. Moreover, there has been no evidence that estimates of BMAD improve fracture prediction more than using BMD .
US Hispanic and US Caucasian men had similar total hip BMD regardless of body size. Travison et al.  also showed the similarity in femoral neck BMD between both race/ethnic groups, but NHANES III reported 4.9-5.8% higher femoral neck BMD at age 60-69 and 70-79 in Hispanic men than White men. The lack of clear-cut Hispanic-White differences in BMD may reflect the diversity among Hispanic subpopulations due to differences in admixture and acculturation .
There are several limitations to our study. Firstly, due to the smaller number of US Hispanic and US Asian men, we had limited power to find statistically significant differences between these groups and Caucasian men.
Secondly, since South Korean subjects were from one area in South Korea, BMD value of this group could be biased from the general Korean populations. However, our South Korean group is very similar in major characteristics to the same aged group from the Korea NHANES, a national health survey. The absolute difference is only 1.1 cm in height, 0.1 kg in weight, and 0.2% in the proportion of current smokers between the Namwon Study and Korea NHANES 2007, and 0.2 drinks per week in alcohol consumption between this study and Korea NHANES 2005, respectively. Therefore, the Korean men's mean BMD in this study is thought to be similar to the national value.
Thirdly, the manufacturer of the DXA scanner for Korean men was different than that for other race/ethnic groups. Lunar scanners are likely to overestimate the nominal BMD, while Hologic scanners underestimate it [39, 40]. To remove this bias, we used sBMD  in the cross-calibration procedure, which is specific for scanner manufacturer. Cross-calibration for Korean scanner was done by the quality assurance group who had also calibrated the MrOS scanners and the Hong Kong and Tobago scanners. Correction factors were systematically applied to each scanner. In spite of this procedure, femoral neck BMD results in Korean men compared to other race/ethnic groups were not consistent to those at other bone sites.
Lastly, we could not adjust for sun exposure factors such as latitude, urban/rural area, and outdoor activity, but we hope to measure serum 25-hydroxyvitamin D levels for all ethnic groups in a future study.
Our findings show substantial race/ethnic differences in BMD even within men of African or Asian origin and illustrate the important role of body size on the difference between Asian men and others.
Demographic and lifestyle characteristics
Table 1 shows the demographic and lifestyle characteristics across race/ethnic groups. Compared with US Caucasian men, US Hispanic, US Asian, and Hong Kong Chinese men had similar mean age; Afro-Caribbean and Korean men were slightly younger. African-American men had similar weight and height compared to US Caucasian men. Afro-Caribbean and US Asian men weighed less and were shorter than US Caucasian men. Hong Kong Chinese and Korean men weighed less and were shorter than other race/ethnic groups. BMI was lower in all Asian ethnic groups. The range of BMI differed across race/ethnic groups. Few men (1.4% to 5.1%) among the three Asian ethnic groups were obese (BMI>30 kg/m2). The two non-US Asian groups had a higher proportion (11.5% to 12.7%) of men with BMI<20. On the other hand, a substantial proportion (22.0% to 33.2%) of men among US Caucasian, African-American, Afro-Caribbean, and US Hispanic groups were obese, and few men (0.6% to 2.6%) had low body weight (BMI<20 kg/m2).
Current smoking was highest among Korean men and lowest among US men, but more than 50% of all men except Afro-Caribbeans reported past smoking. Korean men also reported much greater alcohol consumption compared to other groups. Most men reported walking at least five times per week, especially among Hong Kong men. Dietary calcium intake was greatest in US Caucasian men and lowest in Korean men. Compared with US Caucasian men, Korean, Hong Kong Chinese, and African-American men were more likely to report their health status as fair or poor (Table 1).
Differences in BMD among race/ethnic groups
Table 2 shows the crude and adjusted mean BMD at the femoral neck, total hip, and lumbar spine. Figures 1 and 2 present percentage differences in the adjusted mean BMD at each site among race/ethnic groups compared with Caucasian.
When compared with US Caucasian men, age-adjusted mean BMD measures at all three BMD sites were 8-20% higher among Afro-Caribbean and 6-12% higher among African-American men. Hip BMD was similar among US Caucasian and Hispanic men, but spine BMD was 3% lower among Hispanic men. Hip and spine BMD values were 3-5% lower among US Asian, 7-10% lower among Hong Kong Chinese, and 8-14% lower except femoral neck among Korean men compared to US Caucasians. The differences shown above were statistically significant (p<0.001) or nearly significant (p=0.057 for femoral neck in Asian men) except for spine BMD in Hispanic or Asian men (Table 2; Fig. 1).
After additional adjustment for weight and height, differences in mean BMD at each site between Caucasian men vs African-American men or Afro-Caribbean men persisted. However, this adjustment greatly attenuated the differences in BMD between US Caucasian men and Asian ethnic groups such as US Asian, Hong Kong Chinese, and Korean men (Table 2; Fig. 2). Afro-Caribbean men had higher adjusted BMD at all sites than African-American men. Among Asian groups, US Asian and Hong Kong Chinese men had similar BMD at hip sites, but Korean men had higher BMD at femoral neck and lower BMD at total hip. Hong Kong Chinese men had lower spine BMD than other Asian groups.
Additional adjustment with lifestyle factors showed only minor changes (-0.1-0.9% point) in the percentage differences between Caucasian men vs each race/ethnic group except those at hip sites between Caucasian men vs Korean men (1.9% point; Table 2).