The authors have no conflict of interest.
Racial Differences in Rate of Decline in Bone Mass in Older Men: The Baltimore Men's Osteoporosis Study†
Article first published online: 14 MAR 2005
Copyright © 2005 ASBMR
Journal of Bone and Mineral Research
Volume 20, Issue 7, pages 1228–1234, July 2005
How to Cite
Tracy, J. K., Meyer, W. A., Flores, R. H., Wilson, P. D. and Hochberg, M. C. (2005), Racial Differences in Rate of Decline in Bone Mass in Older Men: The Baltimore Men's Osteoporosis Study. J Bone Miner Res, 20: 1228–1234. doi: 10.1359/JBMR.050310
- Issue published online: 4 DEC 2009
- Article first published online: 14 MAR 2005
- Manuscript Accepted: 10 MAR 2005
- Manuscript Revised: 4 MAR 2005
- Manuscript Received: 14 OCT 2004
- bone densitometry
Older black men have higher adjusted BMD than older white men. Using data from a longitudinal cohort study of older men followed for a mean of 18.8 ± 6.5 (SD) months, we found that older black men have a higher rate of decline in femoral neck and total hip BMD and femoral neck BMAD than older white men.
Introduction: Older black men have higher adjusted BMD compared with older white men. The difference in BMD may be caused by having attained higher peak bone mass as young adults and/or having a slower rate of decline in bone mass as adults. There are few published longitudinal data on change in bone mass in older white men and no published data for older black men.
Materials and Methods: Three hundred forty-nine white men and 119 black men ⩾65 of age (mean age, 75 ± 5.7 and 72 ± 5.6 years, respectively) who participated in the longitudinal component of the Baltimore Men's Osteoporosis Study returned for a second visit after a mean of 18.8 ± 6.5 (SD) months and were not taking medications used to treat low bone mass at either visit. BMD was measured at the femoral neck and total hip by Hologic-certified technicians using a QDR 2000 at the baseline visit (V1) and QDR 4500 at the first follow-up visit (V2). Participants also completed self-administered and interviewer-administered questionnaires and underwent standardized clinic examinations. Bone mineral apparent density (BMAD) at the femoral neck was calculated as an estimate of volumetric BMD. Annual crude and multiple variable adjusted percent changes in BMD and BMAD were calculated.
Results: In univariate analyses, black men had lower percent decline in femoral neck and total hip BMD and femoral neck BMAD than white men. In addition, older age at baseline, lower baseline weight, current smoking, and lower baseline BMD were associated with greater percent decline per year in femoral neck BMD; older age at baseline, current smoking, and lower baseline BMD were associated with greater percent decline per year in total hip BMD; and older age at baseline and lower baseline femoral neck BMAD were associated with greater percent decline per year in femoral neck BMAD. Racial differences in bone loss persisted in multiple variable models that controlled for other factors associated with change in BMD and BMAD.
Conclusions: Older black men seem to lose bone mass at a slower rate than older white men. These differences in the rate of bone loss may account, in part, for the racial disparities in BMD and BMAD and risk of osteoporotic fractures among older men.
OLDER BLACK men have higher BMD at all sites compared with older white men.(1) The difference in femoral neck BMD, 0.11 g/cm2, approximates 1 SD and may account, in part, for the lower risk of hip fracture among older black compared with white men in the United States.(2–4) Higher BMD among older black men may reflect higher peak skeletal bone mass, a slower rate of bone loss, or both. Indeed, Bell et al.(5) showed that black men achieved a higher peak bone mass than white men.
There are few published studies of longitudinal change in BMD in older men, and all of these have been conducted in North American, Australian, or European white men.(6–14) Cauley et al.(15) recently compared the rate of decline in BMD between older black and white women participating in the Study of Osteoporotic Fractures. They found that both the absolute and percentage decline in hip BMD was about twice as great in white than black women. To study possible racial differences in the rate of change in BMD among older men, we established a cohort of older men who are being followed longitudinally.(1) Reported herein are analyses comparing absolute and percentage changes in hip BMD and femoral neck bone mineral apparent density (BMAD) between black and white men in this cohort over an average follow-up of 18 months. Factors associated with the rate of decline in hip BMD and BMAD are also reported.
MATERIALS AND METHODS
From July 2000 through July 2001, 694 men (503 white and 191 black) ⩾65 years of age were enrolled in the Baltimore Men's Osteoporosis Study (MOST) and completed the baseline examination. The methods of recruitment and inclusion and exclusion criteria have been published.(1) Briefly, volunteers were recruited from population-based listings of age-eligible male drivers in the Baltimore metropolitan area and surrounding counties. Men with bilateral total hip replacements, weight over 300 lb, or who were unable to give informed consent were excluded from the study. Baseline characteristics of the original MOST sample have also been reported.(1)
From July 2001 to July 2003, 542 (415 white and 127 black) men (80% of surviving participants) attended a follow-up visit during which BMD measurements of the femoral neck and total hip were obtained. The study sample included in this analysis comprised those 468 (349 white and 119 black) men who participated in both visits and were not taking medications that can be used to treat low bone mass and/or osteoporosis (bisphosphonates, calcitonin, fluoride, teriparatide, testosterone) at either visit.
The study was approved by the Institutional Review Boards of the University of Maryland Baltimore and the Maryland Veterans Affairs Health Care System. All participants gave written informed consent for participation in the study.
Measurement of bone mass
BMD was measured by DXA using a QDR-2000 (Hologic, Waltham, MA, USA) at the lumbar spine, proximal femur, total hip, and total body at the baseline visit (visit 1). BMD was measured by DXA using a QDR-4500 (Hologic) at the proximal femur and total hip at the follow-up visit (Visit 2). The DXA systems were calibrated daily to provide accurate BMC measurements in vivo using an anthropomorphic phantom. Precision error rates for both the QDR 2000 and QDR 4500 were 1% or less. All DXA measurements were performed by Hologic-certified technicians.
BMD measured by DXA is expressed as an areal density in g/cm2. Because there are differences in bone size by race,(1) the estimated volumetric BMD, the BMAD, was calculated at the femoral neck using published equations.(16,17)
To address potential variability in DXA measurements between the QDR 2000 and QDR 4500, the densitometers were cross-calibrated using data from repeated measures of hip and spine phantoms obtained on both devices; a human calibration study was not performed. The hip phantom was measured 20 times on each QDR machine without repositioning, and a total of 40 measurements were taken on each machine using two different spine phantoms (20 measures for each phantom). Student's t-tests were performed to determine whether phantom measurements differed significantly between the QDR 2000 and the QDR 4500. Constants were applied to DXA measurements where appropriate to adjust for statistically significant intermachine measurement variability and to make results for the QDR 4500 comparable with those of the QDR 2000 (Table 1). All subsequent analyses were performed using corrected values.
At both visits, participants completed both a self-administered and an interviewer-administered questionnaire. All interviews were conducted by a trained nurse examiner at the time of clinic examination. Information collected included (1) demographic characteristics, (2) lifestyle and habits, (3) current and previous medical conditions, (4) current medications, (5) participation in recreational physical activities, (6) personal fracture history, (7) family history of fracture, and (8) health-related quality of life measures. In addition, at the baseline visit only, participants completed the 1998 version of the Block Food Frequency Questionnaire.(18)
Anthropometric data obtained at each clinic examination included (1) measurement of weight without shoes and outer clothing using a standard balance beam scale and recorded to the nearest 0.1 kg, and (2) height without shoes using a Harpenden stadiometer recorded to the nearest centimeter. Participants' medications were also reviewed.
Student's t-tests and Pearson's χ2 tests of significance were computed to determine differences in demographic and clinical characteristics of men by race who returned for the first follow-up visit and those who did not to evaluate the potential for bias related to attrition.
Primary statistical analyses were performed on data available from men that participated in both the baseline visit (visit 1) and the first follow-up visit (visit 2) and were not taking medications that can be used to treat low bone mass and/or osteoporosis at either visit. Student's t-tests and Pearson's χ2 tests of significance were computed to determine differences in demographic and clinical characteristics of these men by race. Statistical significance was inferred when p ⩽ 0.05.
Change in BMD and BMAD between the baseline and follow-up visit was quantified in the following manner: (1) absolute change in BMD and BMAD in g/cm2 and g/cm3, respectively, per year and (2) percent change in BMD and BMAD in g/cm2 and g/cm3, respectively, per year.
Simple linear regression analyses were performed to examine the association between race, age, baseline weight, weight change between visits, current smoking, and baseline BMD or BMAD and the absolute and percent change in BMD and BMAD, respectively. Multiple variable linear regression analyses were performed to identify factors associated with percent change in BMD and BMAD. All variables examined in the univariate analyses were included in the initial multiple variable models. A backward, stepwise elimination procedure was used, and main effects were retained in the models when they were associated with the outcome or when their interaction term with race was associated with the outcome with p < 0.10. All statistical analyses were performed using Stata 7 Special Edition software (Stata Corp., College Station, TX, USA).
The mean time between the baseline and first follow-up visit was 18.8 ± 6.5 (SD) months. White men were significantly more likely to complete both visits than black men: 415 (82.5%) of 503 versus 127 (66.5%) of 191 (p < 0.05). The white men who participated in both visits were on average younger and slightly taller at baseline; among the black men, those who participated in both visits were on average slightly taller and had higher quadriceps strength at baseline (Table 2). There were no significant differences in femoral neck or total hip BMD or femoral neck BMAD at baseline between those who participated in both visits compared with those who only the completed the baseline visit among the race-specific groups.
Table 2 also provides selected demographic and clinical characteristics of the men who attended both the baseline and follow-up visits by race. White men were significantly older and more likely to report current drinking and engaging in regular physical activity during the previous year than black men. Black men, on the other hand, had greater average weight loss between visits and were more likely to report current smoking than white men. Baseline femoral neck, total hip, and lumbar spine BMD and femoral neck BMAD were all significantly higher at baseline among black than white men. Black men were more likely to self-report having diabetes mellitus and a history of thyroid disorder and less likely to report a history of cancer and prostate enlargement than white men. Black men were more likely to be taking thiazide diuretics than white men; there were no differences in the proportion of men who were taking antiepiletic drugs, β blockers, or glucocorticoids by race.
Racial differences in bone loss
White men had greater absolute and percent decline in BMD at the femoral neck and total hip and greater absolute and percent decline in BMAD at the femoral neck than black men. Femoral neck BMD declined by an average of 0.017 ± 0.033 g/cm2 in white men, whereas black men lost an average of 0.010 ± 0.029 g/cm2 (p < 0.05). The rate of decline in femoral neck BMD among white men was ∼2.1 ± 3.7% per year compared with 1.1 ± 3.3% per year among black men (p = 0.01; Fig. 1). Total hip BMD decline by an average of 0.010 ± 0.028 g/cm2 in white men, whereas total hip BMD among black men was basically unchanged (+0.001 ± 0.034 g/cm2 between visits 1 and 2; p = 0.007). The rate of decline in total hip BMD among white men was 0.8 ± 2.8% per year for white men compared with basically no change (+0.05 ± 3.4% per year) for black men (p = 0.007; Fig. 2). Femoral neck BMAD declined by an average of 0.019 ± 0.039 g/cm3 in white men, whereas black men lost an average of 0.008 ± 0.017 g/cm3 (p = 0.12). The rate of decline in femoral neck BMAD among white men was ∼3.6 ± 7.9% per year compared with 1.9 ± 4.6% per year among black men (p = 0.03; Fig. 3).
Factors associated with decline in BMD and BMAD
Potential factors that were evaluated included race, age, height, weight, weight change between visits, smoking status, and selected co-morbid conditions and medications as well as BMD and BMAD at baseline.
Results of selected univariate analyses are shown in Table 3. White race, older age, lower baseline weight, and current smoking were all associated with a greater rate of decline in both absolute and percent change per year in femoral neck BMD; lower femoral neck BMD at baseline was associated with greater percent decline in femoral neck BMD. White race, older age, current smoking, and baseline total hip BMD were associated with greater rate of decline in both absolute and percent change per year in total hip BMD. White race and older age were both associated with greater decline in percent change per year in femoral neck BMAD; baseline femoral neck BMAD was associated with both the absolute change and percent change between visits. Neither weight change between visits nor height at baseline were associated with either absolute or percent change per year in any of these outcomes.
Adjusted differences in BMD and BMAD by race
Racial differences in percent change per year in femoral neck BMD were no longer statistically significant after adjusting for age, baseline weight, weight change, current smoking, and baseline femoral neck BMD (p = 0.07; Fig. 1). In this model, older age and current smoking remained significant independent predictors of percent decline in femoral neck BMD; baseline femoral neck BMD was not a significant predictor of percent change in femoral neck BMD. Similarly, racial differences in percent change per year in total hip BMD were no longer statistically significant after adjusting for age, baseline weight, weight change, current smoking, and baseline total hip BMD (p = 0.09; Fig. 2). In this model, older age and current smoking remained significant independent predictors of percent decline in total hip BMD; baseline total hip BMD was not a significant predictor of percent change in total hip BMD.
Racial differences in percent change per year in femoral neck BMAD did remain significant after adjusting for age, baseline weight, weight change, current smoking, and baseline femoral neck BMAD (p < 0.001; Fig. 3). In this model, older age, lower baseline weight, and higher baseline femoral neck BMAD were significant predictors of percent decline in femoral neck BMAD.
This study found that the decline in bone mass, measured either as areal BMD or estimated as volumetric BMAD, was greater in older white than in older black men. These results extend the results of our(1) and other cross-sectional studies(5,19–22) that found that older black men have higher BMD than older white men. This is the first study that reports cross-sectional differences in BMAD by race in older men. These racial differences in BMD and BMAD and the rate of decline in BMD and BMAD in old age may explain, in part, the lower rate of fractures seen in older black than white men.(2–4)
In this study, BMD declined at both the femoral neck and total hip among this cohort of older men. This finding is consistent with inferences drawn from age-related changes in cross-sectional studies,(1,19,20) as well as findings from prospective studies.(6–14) The percent decline in both femoral neck and total hip BMD was much greater in this study than in others; this may reflect not only characteristics of the older men but also the fact that different densitometers were used to measure BMD at each visit in this study. The men included in this analysis had an older mean age, were less likely to report current physical activity, and were more likely to report current smoking than men included in other studies.(6–8,14) These three factors are all related to greater rates of decline in percent hip BMD. Nonetheless, whereas the magnitude of the decline in BMD is greater than that reported in prior studies, the finding of a decline over time is consistent with findings from the prior studies.
Older age was associated with greater absolute and percent decline in femoral neck BMD and BMAD and total hip BMD in this cohort of older men, even after adjustment for potential confounding variables. This finding is consistent with findings from previous prospective studies in white men.(6–14) Furthermore, in data from the Study of Osteoporotic Fractures, the rate of decline in total hip BMD not only increased with advancing age in white women,(15,23) but was also almost 4-fold greater in black women ⩾75 years of age compared with black women 65-74 years of age.(15) Hence, progressive declines in BMD occur in both race and sex groups with advancing age.
In this study, lower baseline weight and current smoking were both associated with greater absolute and percent decline in femoral neck and total hip BMD. These findings are largely consistent with findings from previous prospective studies.(6–14) We did not find, however, an association between current physical activity or current alcohol consumption and decline in either femoral neck or total hip BMD. Results of previous longitudinal studies have been inconsistent for associations between these variables and decline in BMD; this may reflect differences in types and intensity of physical activity and frequency and volume of alcohol consumed, respectively, in different cohorts of older men. As noted above, there was generally a low level of current physical activity among participants of this study.
The greater rate of decline in BMD in white men may be related to higher rates of bone turnover in white men. In studies in which bone histomorphometry of the iliac crest was performed, bone formation was reduced in normal black men compared with white men.(24) Consistent with a lower rate of bone turnover in black men is the finding of decreased serum osteocalcin levels in black men and women compared with white men and women.(25,26) Of interest, we found racial differences in average daily dietary intake of both calcium and vitamin D. Higher intake of both calcium and vitamin D should, however, be associated with lower levels of bone turnover. We have not measured bone turnover markers or vitamin D metabolites (25-hydroxy and 1,25-dihydroxyvitamin D) in these men.
This study has several strengths. It is one of the first studies to compare rates of bone loss in a well-characterized cohort of community dwelling older white and black men. BMD was measured with excellent quality control in the majority of survivors. We controlled for a number of important factors that have been reported to be associated with bone loss. Several potential limitations, however, need to be acknowledged. The most important limitation of this study is the use of different densitometers, albeit from the same manufacturer, to measure BMD between visits 1 and 2. Whereas we did perform a study to determine the intra- and intermachine variability using standard phantoms and make corrections for differences between the machines in measurement of BMD using these phantoms, we did not do this cross-calibration study using human subjects of appropriate age and sex. It is important to recognize that, whereas these procedures using phantoms are standard when transferring data to newer versions of densitometers from the same manufacturer, they may not fully address differences in BMD between visits that are related to machine variability. However, given that all participants were measured on the QDR 2000 at baseline and the QDR 4500 at the follow-up visit, and that corrections were used for differences in mean values obtained with the same phantoms on these machines, we have no reason to suspect that measurement variability would introduce a bias in the analyses of racial comparisons or in the relationship of risk factors for decline in BMD.
Additional limitations included the following: first, as noted previously, participants were recruited from a roster of licensed drivers and are relatively healthy, ambulatory older men with low levels of physical disability.(1) Second, older black men were significantly less likely to return for a follow-up visit than older white men. This racial difference in attrition raises the possibility of bias caused by a loss of less healthy older black men who might have had a greater rate of decline in BMD had they returned for follow-up. Black men who did not participate in the follow-up component of this longitudinal study were more frail at their baseline visit, as characterized by shorter height and lower quadriceps strength. However, there were no significant differences in baseline BMD between the men who did and did not return for follow-up visits. Third, we excluded data on 74 men (66 white, 8 black) who were taking medications that can be used to treat low bone mass and/or osteoporosis at either study visit from these analyses. The men who were taking these bone-active medications had significantly lower femoral neck and total hip BMD and were more likely to have osteoporosis (defined as femoral neck BMD T score of −2.5 or less) at their baseline visit.(27) Hence, the results of these analyses are relevant to older men not taking these bone-active medications. We did not exclude men who were taking other medications such as thiazide diuretics, β blockers, glucocorticoids, and antiepileptics that can potentially influence not only BMD but also fracture risk.
The results of this study are similar to those reported by Cauley et al.(15) in older white and black women who participated in the Study of Osteoporotic Fractures and suggest that the processes involved in skeletal remodeling in older adults may be similar in men and women. Furthermore, these differences in rate of bone loss, measured either as areal or volumetric BMD, may account for the racial disparities in osteoporotic fractures among older men and women.
This study was supported by grants from the Department of Veterans Affairs and the Arthritis Foundation, Maryland Chapter and University of Maryland General Clinical Research Center Grant M01 RR165001, General Clinical Research Centers Program, National Center for Research Resources, NIH.
- 12003 Racial differences in bone mineral density in older men. J Bone Miner Res 18: 2238–2244., , , , ,
- 21990 Hip fracture incidence among the old and very old: A population-based study of 745,435 cases. Am J Public Health 80: 871–873., , , , ,
- 31994 Racial differences in fracture risk. Epidemiology 5: 42–47., , , , , ,
- 41999 Fracture risk in the U.S. Medicare population. J Clin Epidemiol 52: 243–249., , ,
- 51995 Demonstration that bone mineral density of the lumbar spine, trochanger, and femoral neck is higher in Black than in White young men. Calcif Tissue Int 56: 11–13., , ,
- 61994 Progressive loss of bone in the femoral neck in elderly people: Longitudinal findings from the Dubbo osteoporosis epidemiology study. BMJ 309: 691–695., , , ,
- 71998 Risk factors for increased bone loss in an elderly population: The Rotterdam Study. Am J Epidemiol 147: 871–879., , , , , ,
- 81999 Determinants of bone loss in elderly men and women: A prospective population-based study. Osteoporos Int 10: 384–391., , , , ,
- 92000 Risk factors for longitudinal bone loss in elderly men and women: The Framingham Osteoporosis Study. J Bone Miner Res 15: 710–720., , , , , ,
- 102002 Bone density and bone-related biochemical variables in normal men: A longitudinal study. J Gerontol 57A: M385–M391., , , , ,
- 112003 Weight loss: A determinant of hip bone loss in older men and women. The Rancho Bernardo Study. Am J Epidemiol 18: 1132–1138.,
- 122003 Effects of dietary nutrients and food groups on bone loss from the proximal femur in men and women in the 7th and 8th decades of age. Osteoporos Int 14: 418–428., , , , , , , ,
- 132003 Longitudinal association between sex hormone levels, bone loss, and bone turnover in elderly men. J Clin Endocrinol Metab 88: 5327–5333., , , , , , , , , ,
- 142004 Modifiable predictors of bone loss in older men: A prospective study. Am J Prev Med 26: 436–442., , ,
- 152005 Longitudinal study of changes in hip bone mineral density among Caucasian and African-American women. J Am Geriatr Soc 53: 183–189., , , , , , ,
- 161992 New approaches for interpreting projected bone densitometry data. J Bone Miner Res 7: 137–145., ,
- 172004 Apparent bone mineral density estimated from DXA in healthy men and women. Osteoporos Int 15: 295–300.,
- 182001 Invited commentary: Another perspective on food frequency questionnaires. Am J Epidemiol 154: 1103–1104.
- 191997 Prevalence of low femoral bone density in older U.S. adults from NHANES III. J Bone Miner Res 12: 1761–1768., , , , , , , ,
- 201998 Updated data on proximal femur bone mineral levels of U.S. adults. Osteoporos Int 8: 468–489., , , , , , ,
- 211995 Ethnic differences in regional bone density, hip axis length, and lifestyle variables among healthy Black and White men. J Bone Miner Res 10: 782–787., , ,
- 221999 An investigation of ethnic differences in bone mineral, hip axis length, calcium metabolism and bone turnover between West African and White adults living in the United Kingdom. Ann Hum Biol 26: 229–242., , , ,
- 231995 Hip and calcaneal bone loss increase with advancing age: Longitudinal results from the Study of Osteoporotic Fractures. J Bone Miner Res 10: 1778–1787., , , , , , , ,
- 241988 Diminished rates of bone formation in normal Black adults. N Engl J Med 319: 1698–1701.,
- 251994 Reference data for bone mass, calciotropic hormones, and biochemical markers of bone remodeling in older (55–75) postmenopausal white and black women. J Bone Miner Res 9: 1267–1276., , , , , ,
- 262001 Secondary hyperparathyroidism and bone turnover in elderly Blacks and Whites. J Clin Endocrinol Metab 86: 3801–3804., ,
- 272004 Bone-active medications increase bone mineral density in older men. Proceedings of the 10th Annual Meeting of the International Society for Clinical Densitometry, Miami, FL, USA, January 28–31, 2004., , , ,