Dr Barrett-Connor serves on advisory committees for and receives research support from Amgen, Eli Lilly, and Merck. Dr Cauley is a consultant for Merck, Merck Speakers Bureau, Eli Lilly, and Novartis and receives research support from Merck, Eli Lilly, Pfizer, and Novartis. All other authors state that they have no conflicts of interest.
Race and Ethnic Variation in Proximal Femur Structure and BMD Among Older Men†
Version of Record online: 17 SEP 2007
Copyright © 2008 ASBMR
Journal of Bone and Mineral Research
Volume 23, Issue 1, pages 121–130, January 2008
How to Cite
Marshall, L. M., Zmuda, J. M., Chan, B. K., Barrett-Connor, E., Cauley, J. A., Ensrud, K. E., Lang, T. F., Orwoll, E. S. and for the Osteoporotic Fractures in Men (MrOS) Research Group (2008), Race and Ethnic Variation in Proximal Femur Structure and BMD Among Older Men. J Bone Miner Res, 23: 121–130. doi: 10.1359/jbmr.070908
- Issue online: 4 DEC 2009
- Version of Record online: 17 SEP 2007
- Manuscript Accepted: 12 SEP 2007
- Manuscript Revised: 22 MAY 2007
- Manuscript Received: 6 DEC 2006
- bone QCT;
- ethnic groups;
Femoral neck dimensions and vBMD from QCT were compared among 3305 black, Asian, Hispanic, and white men ≥65 yr of age. All had similar stature-adjusted mean femoral neck volume, but black and Asian men had thicker cortices and higher trabecular vBMD, which may increase bone strength.
Introduction: Hip fracture rates among elderly U.S. black and Asian men are lower than rates among white men. Structural characteristics or volumetric BMD (vBMD), which confer advantages for femoral neck bone strength, may vary by race/ethnicity. However, this topic has not been studied in detail.
Materials and Methods: In a cross-sectional study, dimensions and vBMD in the femoral neck and shaft were obtained from QCT scans among 3305 men ≥65 yr of age in the Osteoporotic Fractures in Men (MrOS) study. Femoral neck measures were cross-sectional area; integral, cortical, and medullary volumes and integral, cortical, and trabecular vBMD. Shaft measures were cross-sectional, cortical, and medullary areas and cortical vBMD. Self-reported race/ethnicity was classified as black, Asian, Hispanic, or white. We used multivariable linear regression models with adjustment for age, height, and body mass index to compare means of the outcome measures in black, Asian, and Hispanic men to those in whites.
Results: All groups had similar femoral neck integral volume. Among black and Asian men, mean cortical volume as a percent of integral volume was 6% greater, integral vBMD was 6-10% greater, and trabecular vBMD was 33-36% greater than means among whites. Shaft cross-sectional area was similar among blacks, but smaller among Asians, compared with whites. However, mean shaft cortical area was greater among blacks but similar among Asians and whites, resulting in mean cortical thickness being 5% greater among black and Asian men. Blacks also had greater mean cortical vBMD in both the femoral neck and shaft.
Conclusions: Black and Asian men ≥65 yr of age have features in the proximal femur that may confer advantages for bone strength. Specifically, greater cortical thickness and higher trabecular vBMD among black and Asian men could help explain the lower hip fracture rates in these populations. Discerning the mechanisms underlying these differences could provide advances for the prevention and treatment of osteoporosis.
Hip fracture rates among older U.S. black, Asian, and Hispanic men are considerably lower than rates among white men,(1-9) although rates among Hispanic men may be increasing.(10) Areal BMD, an important fracture risk factor among men,(11-13) also varies according to race or ethnicity but does not always parallel patterns in hip fracture rates.(14) Total hip and femoral neck BMD are greater on average among black men than they are among white men of the same age range.(15-20) In contrast, hip BMD measures are reported to be lower or similar among Asian men(20-22) and similar among Hispanic men(20) compared with those among similarly aged white men. Although areal BMD by DXA is strongly correlated with bone strength,(23,24) it is unable to provide specific information on structural dimensions or BMD in the cortical and trabecular compartments that also contribute to bone strength and fracture resistance.(25)
Comparisons of skeletal features that confer bone strength between race and ethnic groups may yield insights about mechanisms that could contribute to lower hip fracture rates among older nonwhite men. However, information on this topic is limited.(26) Therefore, we obtained QCT scans among participants in the Osteoporotic Fractures in Men (MrOS) study, a large cohort of U.S. men ≥65 yr of age, to measure dimensions and volumetric BMD (vBMD) in the femoral neck and shaft. We compared distributions of the QCT measures among black, Asian, and Hispanic men to distributions among white men in the cohort.
MATERIALS AND METHODS
The MrOS study enrolled 5995 participants from March 2000 through April 2002 as previously described.(27,28) Briefly, recruitment occurred at six U.S. academic medical centers in Birmingham, AL; Minneapolis, MN; Palo Alto, CA; Pittsburgh, PA; Portland, OR; and San Diego, CA. Recruitment was accomplished primarily through targeted mailings based on motor vehicle registration, voter registration, and Veteran's Administration databases. Eligible participants were at least 65 yr of age, able to walk without assistance from another person, and had not had bilateral hip replacement surgery. Each study site enrolled ∼1000 men. Proportions of black, Asian, and Hispanic men enrolled at each study site were generally representative of those reflected in the local population of older men by U.S. Census data.(27,28) All participants gave written informed consent.
Information about demographic factors, body size, physical activity, health status, smoking history, alcohol use, and medical conditions was obtained by self-report at baseline. Race and ethnicity was collected for demographic purposes using a single question in a manner consistent with standards at the time.(29) Participants indicated their background as one or more of the following: American Indian or Alaska Native, Asian, African American or black, Hispanic or Latino, Native Hawaiian, or Other Pacific Islander, or white. All men responded, and 5908 (98.5%) marked only one item. Responses were classified into mutually exclusive “race/ethnicity” categories as Hispanic, black, Asian, white, or other. The category Hispanic includes 127 men, of whom 99 (78%) indicated Hispanic or Latino only, 23 (18%) additionally indicated white, and 5 indicated additionally American Indian/Alaska Native or Asian. The category “other” includes 71 men, of whom 59 (83%) marked at least two response items (other than Hispanic or Latino), 7 marked American Indian/Alaska Native only, and 5 marked Native Hawaiian/Other Pacific Islander only.
Cigarette smoking was classified as current, past, or never. Current alcohol consumption was computed as average drinks per week and categorized as nondrinker, 1 to <7/wk, and ≥7/wk. The physical activity scale for the elderly (PASE) provided an overall physical activity score based on exercise, leisure, and occupational activities.(30) History of arthritis, diabetes, hypertension, angina, myocardial infarction, stroke, cancer, or osteoporosis was classified if the participant reported ever being diagnosed by a physician with the specific condition. Fracture history was determined from self-report of the anatomic location and circumstances resulting in the event. Medications specifically to treat osteoporosis about which the men were queried were fluoride, sodium fluoride, calcitonin, Miacalcin, bisphosphonates (alendronate, clodronate, etidronate, ibandronate, pamidronate, risedronate, or tiludronate), raloxifene, or Evista. Men who indicated current or past use of any of these were classified as ever-users of osteoporosis medication.
Height (cm) was measured using a Harpenden stadiometer. Participants were weighed (kg) while wearing indoor clothing except shoes on balance beam or digital scales. Body mass index (BMI) was calculated as weight divided by the square of height in meters (kg/m2). Usual walking pace (m/s) was computed from the time for the participant to walk a 6-m course.
Areal BMD was obtained for the total hip and its subregions with fan-beam DXA (QDR 4500 W; Hologic) at all study sites. Participants were scanned according to standardized procedures.(20) Scanners were calibrated at baseline, and daily quality control scans showed no shifts in scanner performance at any site during the enrollment period.
Constraints on study resources limited the number of QCT scans that could be acquired. Thus, the MrOS study was designed such that the first 650 men and all nonwhite men enrolled at each site were referred for QCT scans of the hip and lumbar spine as part of their baseline visit. Referral of nonwhite men was made to enhance the statistical precision of race/ethnic comparisons. We obtained QCT scans for 3786 of the referred participants (63% of the MrOS cohort). Baseline characteristics of men referred and not referred for scans were comparable, except for a slightly greater proportion of nonwhite men among those referred.(31) Of men referred, 122 were ineligible for a hip scan because of hip replacement, and 1 refused the scan.
Details regarding acquisition of the baseline QCT scans have been described.(31) Briefly, hip scans were obtained using a standardized protocol. The pelvic region from the femoral head to 3.5 cm below the lesser trochanter was scanned at settings of 80 kVp, 280 mA, 3-mm slice thickness, and 512 × 512 matrix in spiral reconstruction mode. Calibration standards (Image Analysis) containing known hydroxyapatite concentrations were included with the participant in every scan. Of the 3663 hip scans, 102 (2.8%) were lost or corrupted during transfer to the central processing site, leaving 3561 available for image processing.
Femoral outcome measures from QCT
Image processing occurred centrally according to published methods(32,33) and without knowledge of the participants' race/ethnicity or other baseline characteristics. Images were calibrated from the native scanner Hounsfield Units (HU) to equivalent concentration (g/cm3) of calcium hydroxyapatite contained in the calibration standard, which appeared in the field of view in all scans. This step reduces variability in image attenuation across scanners, because all HU are scaled to materials of known density.(34)
QCT scans of the left proximal femur were reformatted along the neutral axis of the femoral neck. The periosteal boundary was determined with a region growing algorithm using thresholds to distinguish bone from adjacent soft tissue.(32) Using this boundary, the cross-sectional area along the neutral axis in each slice was plotted. The neck length was computed as the distance between the minimum and the maximum cross-sections. The femoral neck region of interest (ROI) was defined as the portion along the femoral neck length bounded medially at minimum cross-sectional area and laterally to a distance of 25% of the neck length toward the maximum cross-sectional area.
The following measures were obtained in the femoral neck ROI. The cross-sectional area (cm2) was computed as the area within the periosteal boundary at the minimum cross-section. Integral volume (cm3) was computed as the total volume of the ROI within the periosteal boundary. A trabecular volume was obtained by applying an erosion process to the integral volume to retain the same shape in a region fully contained within the medullary space. The cortical volume was defined by applying a threshold of 0.35 g/cm3 to all voxels between the periosteal boundary and the outer boundary of the trabecular volume. Medullary volume was computed by subtracting the cortical volume from the integral volume. The percent cortical volume was computed as cortical volume divided by integral volume times 100%. vBMD (g/cm3) was computed as the concentration of calcium hydroxyapatite averaged over all voxels in the integral, trabecular, or cortical volumes.
The femoral shaft ROI was centered 3 cm beneath the inferior aspect of the lesser trochanter. The following measures were obtained in a 10-mm transverse section reconstructed perpendicular to the neutral axis of the shaft. The cross-sectional area was defined as the area within the periosteal boundary of the shaft. The cortical area was defined by applying a threshold of 0.35 g/cm3 to all voxels within the shaft cross-section and computing the area delineated by those voxels. Medullary area was computed by subtracting the cortical area from the cross-sectional area. The percent cortical area was computed as cortical area divided by cross-sectional area times 100%. Cortical vBMD was computed as the concentration of calcium hydroxyapatite averaged from all voxels in the cortical area.
Of the 3561 hip scans available for processing, 134 (3.7%) failed because of insufficient number of images, interference from metal, calibration standard not visible, or unrecorded cause. Shaft measures could not be obtained in an additional 354 (9.9%) scans because the image did not extend below the lesser trochanter. We further excluded 44 men classified as “other” race/ethnicity. From the remaining 3383 men with femoral measures, we restricted analyses to the 3305 with complete data for the femoral neck outcomes (cross-sectional area; integral, cortical, and medullary volumes; percent cortical volume; integral, cortical, and trabecular vBMD) or to the 2936 with complete data for the shaft outcome measures (cross-sectional area, cortical area, medullary area; percent cortical area; cortical vBMD).
Distributions of baseline characteristics according to race/ethnicity were evaluated with a one-way ANOVA for continuous variables or with χ2 or Fisher's exact tests as appropriate from contingency tables for categorical variables. We used multivariable general linear models (GLMs) to estimate least square (LS) means of each outcome variable in categories of race/ethnicity, with whites serving as the referent group. To build the multivariable models, we first examined differences in all femoral outcome measures by race/ethnicity after adjustment for 5-yr age groups and study site. Next, we examined whether differences persisted after accounting for the potentially confounding effects of body size by adding terms for height and weight or BMI to the models. Finally, we examined several other baseline variables as potential confounders including marital status, education, smoking status, alcohol consumption, physical activity, walking pace, and medical history. None of these variables confounded the associations of race/ethnicity with the femoral outcomes, so they were not included in the regression models. To determine if history of stroke, history of osteoporosis, or use of osteoporosis medications, which were all rare among the nonwhite men, affected the results, we refit our models after excluding men who reported these conditions. Because the LS means estimated in models without these men were virtually identical to the main results, we did not exclude men with these conditions from the analyses shown. We also found that the LS means in each race/ethnicity group were similar regardless of whether (1) age was fit as a continuous variable or in 5-yr groups and (2) BMI was fit instead of weight. Thus, our final regression model included adjustments for age (continuous), study site, current height, and BMI. Femoral neck volume measures were also additionally adjusted for the femoral neck length. All analyses were conducted with SAS statistical software (SAS Institute).
Proportions in the race/ethnic categories among men in the main MrOS cohort and among the 3786 men referred for QCT were comparable with, respectively, 4.1% and 5.6% black, 3.2% and 3.5% Asian, 2.1% and 2.6% Hispanic, 1.2% and 1.3% other, and 89.4% and 87.1% white. Similar proportions in each race/ethnic group remained after restricting the analytic group to men with femoral outcome measures (Table 1), with 88.5% classified as white.
Distributions of baseline characteristics among the race/ethnic groups are shown in Table 1. On average, whites and Asians were slightly older than blacks or Hispanics. Asians were more likely to be married and to be college graduates. White and Hispanic men more often reported their health to be excellent. Similar proportions had ever smoked. Blacks and Asians were more likely to report no current alcohol use. Asian men were shorter, weighed less, and had the lowest BMI. Compared with whites, Hispanics were shorter but of similar weight, and blacks were of similar height but greater weight, which resulted in greater BMI in these two groups. The PASE score(30) was least in blacks and greatest in the Hispanics. Walking pace was lower among blacks and Asians. There was substantial variation in medical history, with greater prevalence of arthritis and hypertension among blacks, diabetes among blacks and Hispanics, and osteoporosis, cancer, and medication use for osteoporosis among whites. History of nontrauma fracture was greatest among whites and Hispanics.
Comparison of femoral neck dimensions and BMD
Means of the femoral neck dimensions, areal BMD, and vBMD in each race/ethnic group are shown in Table 2. Differences in the adjusted mean femoral neck outcome measures among black, Asian, and Hispanic men as a percentage of the adjusted means among white men are shown in Figs. 1 and 2.
Black compared with white men
Among black men, the crude and adjusted means of the femoral dimensions were similar. Based on the adjusted means, femoral neck cross-sectional area was 3% smaller compared with that among whites, whereas integral volume did not differ significantly between the two groups. Mean cortical volume was 6% larger and mean medullary volume was 3% smaller among the black men, and the net effect was that mean percent cortical volume was 6% larger among black men. Crude and adjusted means of areal BMD from DXA and integral vBMD from QCT were significantly greater by 10% among blacks compared with whites. Notably, adjusted mean trabecular vBMD was 36% higher among blacks. Mean cortical vBMD was similar among the two groups before adjustment, but after adjustment was significantly greater among blacks.
Asian compared with white men
Among Asian men, the crude means for the dimensions and BMD measures tended to smaller than the adjusted means, except for the integral and cortical vBMD measures, which were similar before and after adjustment. Crude mean femoral neck dimensions were smaller among Asian compared with white men, but the adjusted means for cross-sectional area, integral volume, cortical volume, and medullary volume did not differ significantly. Unadjusted mean percent cortical volume did not differ significantly among Asians compared with whites, but adjusted mean percent cortical volume was nearly 4% greater among Asians. Crude mean femoral neck areal BMD was lower among Asian compared with white men, but the adjusted mean did not differ significantly between the two groups. In contrast, both the crude and adjusted means of the integral and trabecular vBMD measures were larger among Asian compared with white men, with adjusted mean integral vBMD and mean trabecular vBMD being, respectively, 6% and 33% higher. Neither crude nor adjusted mean cortical vBMD differed significantly between the two groups.
Hispanic compared with white men
Among Hispanic men, the crude means for the dimension measures were smaller than the adjusted means, except for percent cortical volume; the unadjusted and adjusted mean BMD measures were similar. Unadjusted means of dimensions were lower among the Hispanic compared with white men, except for percent cortical volume. However, after adjustment none of the femoral dimensions or BMD measures differed significantly between the two groups.
Comparison of femoral shaft dimensions and BMD
Means of the shaft dimensions and cortical vBMD in each race/ethnic group are shown in Table 3. Differences in the adjusted mean shaft outcome measures among black, Asian, and Hispanic men as a percentage of the adjusted means among white men are shown in Fig. 3.
Among blacks, the crude and adjusted means of the shaft measures were similar. Compared with whites, black men had similar mean cross-sectional area, but 5% greater adjusted mean cortical area, 10% smaller adjusted mean medullary area, and 5% greater adjusted mean percent cortical area. Among Asian men, the adjusted means for cross-sectional, cortical, and medullary area were larger than the unadjusted means, but similar for percent cortical volume and cortical BMD. Compared with white men, the Asian men had 6% smaller adjusted mean cross-sectional area, similar adjusted mean cortical area, 15% smaller adjusted mean medullary area, and 5% greater adjusted mean percent cortical area. Shaft cortical vBMD was ∼3% greater among black and Asian men compared with white men. Femoral shaft dimensions and cortical vBMD among Hispanic men did not differ significantly from those among whites.
Assessment of possible bias caused by missing data
There was substantial missing outcome data because of errors of acquisition or processing failures. Scans from black men were more likely to be affected. The proportion of scans with missing femoral neck data were 17% among blacks and 7-8% among whites, Asians, and Hispanics; the proportions with missing shaft data were 27% among blacks and 11-15% among whites, Asians, and Hispanics. The reason that scans for black men were more frequently affected is that acquisition errors and processing failures were more common in scans from two of the enrollment sites, and proportion of black men at these two sites was 66%, whereas the proportion of whites was 34%, Asians was 45%, and Hispanics was 12%.
If differences in the outcome measures between black and white men in the two study sites most affected by missing data were not comparable to differences between the two groups in the remaining study sites, then results of the primary analysis would be biased. To assess this possibility, we repeated several analyses after excluding data from the two study sites (data not shown). The crude means and SDs of the femoral neck and shaft outcome measures among the race/ethnic groups in the remaining four study sites were nearly identical to those shown in Tables 2 and 3. In statistical models rerun with the restricted data, the direction and magnitude of the differences in the femoral outcome measures between blacks and whites did not differ substantially from those reported in Tables 2 and 3. Precision of some results was reduced, especially for the shaft measures, because of the smaller sample size. Nevertheless, mean differences between black and white men in femoral neck cortical volume, percent cortical volume, cross-sectional area, integral vBMD, trabecular vBMD, and areal femoral neck BMD remained statistically significant, and mean differences in shaft medullary area and percent cortical area were nearly significant (p = 0.06 for both). The similarity of results from the primary and the restricted analyses suggests that disproportionate missing data among black men is unlikely to have introduced a systematic bias.
In this cross-sectional study among 3305 men ≥65 yr of age with QCT from the MrOS cohort, we observed that elderly black and Asian men have features in the proximal femur that may yield distinct advantages in bone strength. Because of variation in the size of the cortical and medullary dimensions, black and Asian men had thicker cortices in both the femoral neck and shaft than did whites. Greater cortical thickness would be expected to increase bone strength by dispersing more cortical bone outward from the center and conferring resistance to bending and twisting forces.(25) Black and Asian men had greater femoral neck integral vBMD, apparently as a result of greater cortical thickness and higher trabecular vBMD. Higher trabecular vBMD may indicate greater trabecular thickness, at least among the black men,(35) although greater trabecular number or material density could also pertain. Greater cortical thickness and trabecular vBMD among Asian men could explain at least part of the paradox of lower hip fracture rates than whites despite their similar or lower areal BMD. Our observations suggest possible reasons that hip fracture rates are lower among older U.S. black and Asian men, although the contribution of race/ethnic variation in femoral structure to fracture risk must be established.
Black men were of similar height as whites, weighed more, had higher BMI, and had greater femoral neck areal BMD regardless of adjustment for body size. Our data agree with reference data from the Third National Health and Nutrition Examination Survey (NHANES III) in which mean femoral neck BMD among black men 60-69, 70-79, and ≥80 yr of age was 8%, 8%, and 10%, respectively, higher than in whites of the same age. (17) In our study, the percent differences were 13%, 10%, and 8%, respectively, for the same age groups. Others have also reported that areal femoral neck BMD is 11-15% greater among black compared with white men ≥65 yr of age.(19,20) Differences of a similar or slightly greater magnitude are reported among middle-aged(16,18) and young adult(15) black and white men. Prevalence of osteoporosis and osteopenia defined by BMD T-score at the femoral neck is also lower among U.S. black compared with white men.(36)
Femoral neck BMD among the black men appears to reflect potentially advantageous traits revealed by QCT, including greater bone density in both the cortical and trabecular compartments, larger cortical volume, and greater cortical thickness. As in the femoral neck, blacks had thicker shaft cortices. However, shaft cross-sectional area was similar among black and white men, in contrast to what we observed in the femoral neck. Shaft cortical BMD was also greater than that in whites. Our data are consistent with a study among South African men of iliac crest histomorphometry which demonstrated that blacks had significantly higher bone volume to total volume (BV/TV) ratio and greater trabecular thickness than did whites.(35) Greater cortical thickness has also been reported among older black compared with white women.(37) Higher cortical vBMD suggests the possibility of reduced cortical porosity which increases with age and tends to reduce measured cortical vBMD.(38)
Asian men were smaller in stature, had lower unadjusted femoral neck BMD by DXA, and smaller unadjusted bone size. After adjustment for body size, however, areal femoral neck BMD did not differ significantly between Asian and white men, a finding reported by others.(20-22) In contrast to results among black men, areal femoral neck BMD among Asian men did not reflect that integral vBMD, trabecular vBMD, and cortical thickness from QCT were all greater than in white men. This discrepancy supports the argument that areal BMD comparisons between race groups may be confounded by bone and body size.(26,39) In the femoral shaft, Asian men had smaller cross-sectional area compared with whites but thicker cortices in the presence of a smaller medullary area, as well as greater shaft cortical vBMD. Femoral neck data from our study differs from a comparison of older white and Chinese men from Australia, in which it was reported that DXA-derived measures of femoral neck cross-sectional area, endocortical diameter, and cortical thickness were significantly smaller among the Chinese men.(22) However, different scan modalities and image processing techniques could contribute to these discrepancies.
Hispanic men were shorter than white men, had similar weight but higher BMI, and similar measures of structure and BMD in the femoral neck and shaft. Our observations are consistent with NHANES III data, in which age-specific percent differences in mean femoral neck BMD between Hispanic and white men ≥60 yr of age were modest, ranging from 1.6% to 4.9%.(17) In our study, the differences were −2.6% to 3.8% for the same age groups. Femoral neck integral and trabecular vBMD were slightly greater among Hispanic than white men, and these differences were of borderline significance. Thus, the small number of Hispanic men in the cohort may have limited the statistical power to detect differences in comparison with whites.
There are several potential limitations to these analyses. First, this is a cross-sectional study, so we are unable to study antecedents of the observed race/ethnic differences in femoral dimensions and vBMD. Second, although we used a standard definition of race/ethnicity,(29) heterogeneity within each group is likely. We could not distinguish ancestral origins or nationality among the men. Moreover, concepts of race and ethnicity include biologic, environmental, and cultural differences,(14,40) which this study was not designed to evaluate and which we are unable to address. Third, missing femoral outcome measures occurred more often in scans from black men. However, results from extensive additional analyses indicated that bias caused by missing data was an unlikely explanation for the observed differences between black and white men. Although there are challenges to the use of QCT in multisite studies, we attempted to minimize sources of variability by using a standardized acquisition protocol, processing all scans in a single laboratory, and calibrating each image to the same calibration standard.(34) Furthermore, to account for potential variability from the use of different scanners at each study site, we controlled for study site as a proxy for scanner in all analyses. Finally, the MrOS cohort includes older men, so additional studies are needed to determine the extent of variation by race/ethnicity in these femoral measures among young adults and women.
This study has several strengths. We had the opportunity to directly measure structural dimensions and vBMD with QCT and compare these measures with femoral neck BMD from DXA in a large cohort of older men. The extensive baseline data enabled us to accurately control for body size and determine that the observed race/ethnic differences in the femoral outcome measures were not caused by confounding by lifestyle, medical history, or several other factors. Importantly, data from our study seems to provide a reasonable representation of the magnitude of differences in areal femoral neck BMD among older U.S. men. Mean femoral neck BMD by DXA is slightly higher among black, Hispanic, and white men in our study than means reported for men of similar ages and race in NHANES III,(28) but the age- and race-specific percentage differences for blacks and Hispanics compared with whites is comparable in both cohorts.(17) BMD reference data are not available in NHANES III for Asian men. The likelihood that our results apply to older U.S. men is strengthened by observations that means for height, weight, and BMI among black, Hispanic, and white men studied here are nearly identical to the race-specific means from the NHANES 1999-2000 reference data for men ≥60 yr of age.(41)
Overall, the presence of differences between black, Asian, and white men in femoral structure and vBMD suggests that there are corresponding race/ethnic differences in periosteal and endosteal modeling or remodeling with aging.(26) Further study of femoral neck morphology in human populations using comparable methods will enhance our knowledge of mechanisms that result in skeletal differences among aging black, Asian, and white men. Such information may yield advances for the prevention and treatment of osteoporosis.
The Osteoporotic Fractures in Men (MrOS) Study is supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS), the National Institute on Aging (NIA), and the National Cancer Institute (NCI) through Grants U01 AR45580, U01 AR45614, U01 AR45632, U01 AR45647, U01 AR45654, U01 AR45583, U01 AG18197, and M01 RR000334. The authors thank all the participants in the MrOS study for continued participation, Michael M Bliziotes, MD, for insightful editorial comments, and Lori C Lambert, MA, for assistance with the analytic approach.
- 142001 Race, ethnicity, and osteoporosis. MarcusR, FeldmanD, KelseyJ (eds.) Osteoporosis, 2nd ed., vol. 1. Academic Press, San Diego, CA, USA, 569–584., ,
- 29Anonymous 1997 Revisions to the standards for the classification of federal data on race and ethnicity. Fed Reg 62: 58781–58790.
- 412005 Anthropometric Reference Data for Children and Adults: U.S. Population, 1999-20002. Advance Data From Vital and Health Statistics. National Center for Health Statistics, Hyattsville, MD, USA., , ,