SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Most estimates of osteoporosis in older U.S. adults have been based on its occurrence in white women, even though it is known to affect men and minority women. In the present study, we used dual-energy X-ray absorptiometry measurements of femoral bone mineral density (BMD) from the third National Health and Nutrition Examination Survey (NHANES III, 1988–1994) to estimate the overall scope of the disease in the older U.S. population. Specifically, we estimate prevalences of low femoral BMD in women 50 years and older and explore different approaches for defining low BMD in older men in that age range. Low BMD levels were defined in accordance with an approach proposed by an expert panel of the World Health Organization and used BMD data from 382 non-Hispanic white (NHW) men or 409 NHW women ages 20–29 years from the NHANES III dataset. For women, estimates indicate 13–18%, or 4–6 million, have osteoporosis (i.e., BMD >2.5 standard deviations [SD] below the mean of young NHW women) and 37–50%, or 13–17 million, have osteopenia (BMD between 1 and 2.5 SD below the mean of young NHW women). For men, these numbers depend on the gender of the reference group used to define cutoff values. When based on male cutoffs, 3–6% (1–2 million) of men have osteoporosis and 28–47% (8–13 million) have osteopenia; when based on female cutoffs, 1–4% (280,000–1 million) have osteoporosis and 15–33% (4–9 million) have osteopenia. Most of the older U.S. adults with low femur BMD are women, but, regardless of which cutoffs are used, the number of men is substantial.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

OSTEOPOROTIC HIP FRACTURE remains a major public health concern in the United States. Although the etiology of osteoporotic fractures is multifactorial, bone mineral density (BMD) has been identified as one of the primary predictive risk factors.1–3 Based on this fact, an expert panel of the World Health Organization (WHO) recently proposed diagnostic criteria for osteoporosis based on bone density.4,5 The applicability of these criteria to groups other than white women is not certain; but osteoporosis is not only a disease of white women.6,7 For example, of the 281,000 hospital discharges for hip fracture among persons age 45 years and older in the U.S in 1994, 74,000, or 26%, were men (personal communication, Dr. W. Edward Bacon). The proportion of men may increase in the future because the hip fracture incidence rate in U.S. men appears to be going up over time, while rates in women may have plateaued.8,9

We recently estimated the prevalence of older U.S. women with low femoral BMD using the WHO diagnostic criteria in conjunction with femoral BMD data collected in the first 3-year national sample of the third National Health and Nutrition Examination Survey (NHANES III).10 We included estimates for nonwhite women from two minority groups with BMD values that fell below cutoffs based on white women. Since that time, the second 3-year national sample of NHANES III has been completed. In this paper, we update the estimates of osteoporosis in older U.S. women, including minority women, using the full 6 years of NHANES III. In addition, we explore the application of the WHO criteria to older U.S. men.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Data source

Estimates of low femoral bone density are based on data collected in NHANES III. The NHANES are conducted periodically by the National Center for Health Statistics, Centers for Disease Control and Prevention, to assess the health and nutritional status of the civilian noninstitutionalized population of the United States. NHANES III is a 6-year study (1988–1994) divided into two 3-year national probablity samples: phase 1 (1988–1991) and phase 2 (1991–1994). Both phases were designed to be separate national probability samples. The survey uses a stratified, multistage probability design to select the sample and has been described in detail elsewhere.11

All men and nonpregnant women age 20 years and older were eligible for bone densitometry unless they had fractured both hips previously. Bone mineral measurements were performed on 14,646 men and women age 20 years and older in the full survey. This represents 63% of the eligible selected sample, 78% of the eligible interviewed sample, and 88% of the eligible examined sample. The left hip was scanned unless there was a history of previous fracture or surgery; only 1% received a scan of the right femur. Because their inclusion did not alter prevalence estimates, those who received a scan of the right femur were included in the analyses.

Prevalences were estimated for men and women age 50 years and older because the predictive relationship between fracture and femur BMD has been studied exclusively in older persons.1–3 Bone mineral measurements were performed on 3176 older men in NHANES III, but 86, or 3%, were rejected for technical reasons after review, leaving 3090 with acceptable data. Of the 3379 women age 50 years and older who received scans, 68, or 2%, were rejected, leaving 3311 with acceptable BMD data. The race/ethnic composition of the sample of older persons with acceptable BMD data for men and women, respectively, was 1723 and 1880 non-Hispanic whites (NHW), 647 and 695 non-Hispanic blacks (NHB), and 625 and 600 Mexican Americans (MA). Race and ethnic categories were based on self-reported data using U.S. Census Bureau definitions. There were too few persons of other race and ethnic groups (n = 95 and 136 older men and women, respectively) to report prevalences separately for these groups, although they were included in estimates for the total population.

Bone density measurements

BMD of the proximal femur was measured using dual-energy X-ray absorptiometry (DEXA) at five regions of interest. In this study, we present data on four regions: femoral neck, trochanter, intertrochanter, and total femur. Data for the fifth region, Ward's triangle, was not included because (1) it represents a calculated area of low bone density rather than a true anatomic region; (2) Ward's triangle data from the two phases of NHANES III are not directly comparable due to software changes that allowed the location of Ward's triangle to vary in phase 1, but fixed it midcervically in phase 2; and (3) it had a larger in vivo measurement error (5%) than the other regions (2–3%).12 Three densitometers (Hologic QDR 1000; Waltham, MA, U.S.A.), located in mobile examination centers, were used to obtain the measurements. A rigorous quality control (QC) program, including use of anthropomorphic phantoms and review of each QC and respondent scan at a central site, was used throughout the study to ensure data quality. QC results for the first phase of NHANES III have been published elsewhere12; QC results from the second phase were similar to those for phase 1.

Definition of low bone density

The WHO diagnostic criteria for osteopenia and osteoporosis4,5 were used to define low bone density among men and women age 50 years and older. This approach defines cutoff values using BMD data from a young adult reference group. The WHO criteria did not specify details about the reference group in terms of age, race, or gender. NHW men or women between 20 and 29 years of age were used as the reference group in the present study because there are prospective data indicating that bone loss occurs at the femur in women during their 30s.13 In addition, the International Committee for Standards in Bone Densitometry recently chose the 20–29 year age range as the reference range for standardizing cutoffs between different bone densitometry instruments (personal communication, Dr. Peter Steiger). Of the scans performed on the young white men (n = 388) and women (n = 415), all but six scans (1.5%), excluded for technical reasons in each group, were used for the reference group. For each of the four regions of interest, low bone density was defined as: (1) osteopenia: a BMD value between 1 standard deviation (SD) and 2.5 SD below the mean of white men or women age 20–29 years; and (2) osteoporosis: a BMD value >2.5 SD below the young reference mean.

These cutoff points were applied to minority men and women in this study as well as to NHWs because data on the predictive utility of femur BMD for hip fracture in nonwhites is lacking to date. In addition, estimates for men were made using cutoffs derived from both the young male and female reference groups, since currently there are insufficient data to identify clearly the most appropriate approach for men.

Data analysis

Sampling weights were used to calculate prevalence estimates and to account for oversampling and nonresponse to the household interview and physical examination. The sampling weights for phases 1 and 2 were based on the March 1990 and March 1993 Current Population Survey values for the civilian noninstitutionalized population, both adjusted for undercounts.14 All analyzes were performed using SUDAAN.15

Both unadjusted and age-adjusted prevalences of low BMD among men and women age 50 years and older were calculated by race and ethnic group. Unadjusted prevalences were used to calculate the number of men and women with low BMD in the U.S., while age-adjusted prevalences were used for comparison of prevalences between genders and race/ethnic groups, since bone density is related to age16 and the age structure of the different gender and race/ethnic groups varies in the U.S. population. Prevalences were age-adjusted using the direct method to the age distribution of the 1980 population.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Mean BMD, SDs, range of BMD, and the cut-off values corresponding to the WHO diagnostic criteria for low bone mass in the four regions of interest for the reference group of young white men and women are shown in Table 1.

Table Table 1.. MEAN FEMORAL BONE MINERAL DENSITY (BMD) OF 20–29-YEAR-OLD NONHISPANIC WHITE MEN AND WOMEN, NHANES III 1988–94
Thumbnail image of

Updated prevalences of osteopenia and osteoporosis in women in the four femur regions using the WHO diagnostic criteria are shown by population in Table 2. Estimates of osteopenia in older women ranged from 37 to 50%, or 13–17 million women, while osteoporosis ranged from 13 to 18%, or 4–6 million. Prevalences of each condition were higher in the femoral neck region than in the other femoral regions. Age-adjusted prevalences of both osteopenia and osteoporosis were higher in NHWs than in NHBs; prevalences in MAs were similar or slightly lower than in NHWs. The ratio of age-adjusted prevalences in NHWs versus NHBs ranged from 1.5 to 2.8, depending on the femoral region and definition of low BMD used (data not shown). Corresponding ratios of prevalences for NHWs versus MAs ranged from 0.8 to 1.2.

Table Table 2.. PREVALENCE OF LOW FEMORAL BONE DENSITY IN NONINSTITUTIONALIZED U.S. WOMEN AGES 50+, NHANES III 1988–94
Thumbnail image of

Prevalences of osteopenia and osteoporosis in men of all races based on male versus female cutoff values are shown in Table 3. Prevalences of osteopenia were noticably higher when based on male cutoffs than female cutoffs, ranging from 38 to 47%, compared with 15–33%. However, prevalences of osteoporosis were only slightly higher when based on male versus female cutoffs.

Table Table 3.. PREVALENCE OF LOW FEMORAL BONE DENSITY IN NONINSTITUTIONALIZED U.S. MEN AGES 50+, USING MALE VS. FEMALE CUTOFF VALUES, ALL RACES, NHANES III 1988–94
Thumbnail image of

When age-adjusted prevalences in men and women are compared, the female-to-male ratios were 1.1–1.6 for osteopenia when prevalence in men was based on male cutoffs and 1.5–2.8 when based on female cutoffs (data not shown). For osteoporosis, the female-to-male ratios were 2.7–4.3 when using male cutoffs for men and 4–7.5 when using female cutoffs.

Patterns of osteopenia and osteoporosis by race or ethnicity in men using the two sets of cutoffs are compared with the patterns seen in women as shown in Fig. 1. The pattern for osteopenia is similar in men and women regardless of the cutoff used in men, with NHWs of both genders having the highest prevalences, followed by MAs, and NHBs having the lowest prevalences. The prevalences of osteoporosis were also greatest in NHW men regardless of the cutoff used. Prevalences of osteoporosis in men of either minority group were so small that they did not achieve statistical reliability, and so the pattern must be interpreted with caution. It suggests, however, that, unlike women, MA men may have a lower prevalence of osteoporosis than NHB men.

thumbnail image

Figure FIG. 1.. Age-adjusted prevalence of low femur neck BMD by race or ethnicity, ages 50+ years. NHW = NonHispanic white; NHB = NonHispanic black; MA = Mexican American.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

A considerable number of noninstitutionalized older citizens in the U.S. have low femoral bone density, as defined by WHO diagnostic criteria.4,5 For women, updated estimates indicate 4–6 million have osteoporosis and 13–17 million have osteopenia. For men, these numbers depend on whether male versus female cutoff values are used. When based on male cutoffs, 1–2 million men have osteoporosis and 8–13 million have osteopenia; when based on female cutoffs, 280,000–1 million have osteoporosis and 4–9 million have osteopenia. Regardless of which cutoffs are used for men, the numbers are substantial.

The updated prevalences of osteopenia in women are similar to those published previously using phase 1 data only, while the updated prevalences of osteoporosis are slightly lower.10 The latter is a result of the larger SDs and concomitantly lower osteoporosis cutoff values when based on the larger reference group of 409 young NHW women. By chance, the representative sample in phase 2 of the survey included a small number of young women with BMD values that exceeded the BMD range observed in phase 1. This wider BMD distribution contributed to the larger SD for the total sample of these young women. Because of the larger sample sizes, we believe the updated prevalence estimates are the most reliable estimates of osteoporosis and osteopenia in women.

The definitions of low BMD used in this study were based on the WHO diagnostic criteria.4,5 These guidelines represent an important step in describing the extent and characteristics of osteoporosis, but the panel was aware that the guidelines had limitations and would likely change with more experience. A major issue for the current study is the application of the guidelines to men, since they were developed for women. The WHO panel noted that the criteria likely should be modified for men, but there were insufficient data to do so. To firmly resolve the issue, prospective data on the absolute risk of fracture at any given BMD level in men must be obtained. In the absence of such data, the panel suggested either applying female cutoffs to men or using a more stringent criterion of 3–4 SD for osteoporosis if cutoffs were based on a male reference group.

Our cross-sectional study can illustrate some of the issues to be addressed in determining the appropriate definitions of osteopenia and osteoporosis in men. For example, in our sample of older men, 1.5% had BMD values more than 3 SD below the young men's mean, which suggests that this criterion may be too stringent. When we used the 2.5 SD criterion, the prevalence of femoral osteoporosis in older men was similar regardless of the gender of the reference group. This is probably because the osteoporosis cutoff values fell in the tail of the BMD distribution of older men, where small differences in cutoff values will not result in adding or subtracting many individuals. Our estimates of osteoporosis in older men using either male or female cutoffs were similar in magnitude to the estimates of lifetime risk of hip fracture for men from other studies17; this type of similarity was cited by the WHO panel to support the validity of the guidelines for women.4,5 Until the appropriate prospective data on fracture risk in men are available, it seems reasonable to use the 2.5 SD criterion to estimate prevalence of osteoporosis among older men in the population. This approach for defining osteoporosis in a population does not, however, constitute a clinical diagnostic criteria for men.

An interim definition of osteopenia for the male population is not clear though, since prevalences of osteopenia depended on the gender of the reference group used to derive the cutoff values. When based on male cutoffs, the prevalences of osteopenia in older men were almost as high as those in older women, which may seem too high given the lower fracture rate in men. When considering how to define osteopenia in men, it may be pertinent to recall that this category was defined for women primarily to identify those in whom preventing further bone loss would be most useful.4,5 The gradient of risk between BMD and fracture is continuous,4,5 so men with osteopenia are likely to be at some increased risk of developing osteoporosis compared with men with higher BMD values. In addition, femoral BMD also declines progressively with age in men.18 Given this, an osteopenia category in men may be useful to help direct prevention efforts regardless of whether it is defined by male or female cutoffs.

Other issues may be important to consider in defining low bone mass for men. For example, uniform criteria for men and women have been used for other disease risk factors, such as hypercholesterolemia and hypertension. The assumption of similar mechanisms, levels of risk, and response to treatment in men and women is inherently appealing and also provides clinical simplicity. However, these considerations must be reconciled with factors that may favor gender-specific cutoffs. For example, male fracture cases have higher BMD levels than female fracture cases19–22; it is unclear, however, whether this finding reflects a truly higher fracture risk for a given BMD in men or simply that the entire male BMD distribution is higher than that of women. Fracture risk at a given BMD level might truly differ between genders because of differences in skeletal loads from the larger male body mass or gender differences in structure–resistance relationships.23

Alternatively, fracture risk in men could appear to differ because areal bone density, as measured by DEXA, does not completely account for the confounding effect of bone size.24 Young adult men and women have similar spinal volumetric bone density, as measured by computed tomography, but spinal areal bone density by DEXA is greater in men.23,24 At the femur, men have approximately 10% higher areal femoral BMD than women before adjusting for body size,16 but levels are similar after this adjustment, with the possible exception of the femoral neck.25–27 The current DEXA reports do not adjust BMD for body size, so gender-specific cutoff values may be necessary to account for this artifact even though men and women may have similar true volumetric bone density levels.

Another unresolved issue in using the WHO criteria is the appropriate cutoff values for minority men and women. We applied cutpoints based on young whites to the minority groups in this study, as well as previously,10 because the relationship between femur BMD and hip fracture risk has been studied exclusively in whites to date.1,2,3 Prevalences of osteoporosis for whites and blacks in our study are similar to estimated lifetime fracture risks for these groups.17 We also found a similar ratio of age-adjusted prevalences in whites versus blacks in our study as the white–black fracture ratio reported by others.7,28–31 Both of these findings provide some indirect support for our approach. The validity of applying white cutoffs to MAs is more difficult to assess, since very little data are available on hip fracture rates in U.S. Hispanics. The few studies done to date suggest that they have lower rates than whites.31–33 We found lower prevalences of osteopenia in MAs than in NHWs, but prevalences of osteoporosis in women were similar in the two groups (prevalences in men were too small to compare reliably). However, due to their higher mean BMD levels,16 cutoffs based on young MAs would have resulted in even higher prevalences in this group compared with NHW, which is less consistent with fracture patterns. Finally, fracture risk is also influenced by other, non-BMD factors such as hip axis length, muscle strength, or falls, which may differ among race or ethnic groups.21,34–39

Other issues noted previously in our application of the guidelines to older women in the U.S., such as accounting for the individual's age in addition to BMD when considering risk,40–42 still remain to be resolved. In addition, as noted previously,10 the young white male and female reference groups in our study had a lower mean BMD and higher SD than the male and female references group provided by the densitometer manufacturer,43 so that fewer individuals will be identified with low bone mass than if the manufacturer's cutoff values are used. Finally, it should be noted that the absolute BMD values presented in this paper, including the cutoff values, are densitometer-specific and cannot be compared directly with BMD results from other DEXA instruments without a conversion factor for the femur.

Limitations of this study include potential nonresponse bias in the sample and the exclusion of institutionalized people from the NHANES III sample frame. Nonresponse bias in the sample who came to the mobile exam centers in NHANES III is reduced to some extent by a nonresponse adjustment factor included in the calculation of the sample weights.11,14 After applying these weighting adjustments, differences between examinees and nonexaminees were minor.14,44,45 However, about 12% of those who came to the mobile exam centers did not have usable BMD data, and this nonresponse is not addressed by the sample weight adjustments. A nonresponse bias study for phase 1 of NHANES III found differences in weight, self-reported health status, and vitamin supplement use between those with bone density data compared with the examined sample as a whole.45 We explored nonresponse bias among the 20- to 29-year-old NHW women and in individuals age 50+ years to assess the possible impact on means and prevalences. Nonrespondents in the young female reference group weighed more than respondents, but mean total femur BMD did not differ greatly when imputed data for nonrespondents (based on their body weight) were added to the dataset (i.e., 0.941 g/cm2 based on actual and imputed data vs. 0.937 g/cm2 based on actual data only). Likewise, differences in weight, self-reported health status, and age between respondents and nonrespondents age 50+ years did not appear to impact greatly the prevalence estimates (e.g., ∼16% of women and 2–4% of men had total femur osteoporosis whether based on actual and imputed data or on the actual data alone). Studies to assess the effect of missing BMD data in phase 2 are continuing as more data become available from NHANES III. It is also important to note that the prevalence estimates in this study pertain only to the noninstitutionalized U.S. population. Institutionalized persons may have lower bone mass,46,47 so the prevalence of low bone mass in the total U.S. population is probably higher than our estimates.

In conclusion, the number of older U.S. adults with low femur BMD using the WHO criteria is substantial. Most of these are women, but the number of older U.S. men with osteopenia or osteoporosis of the femur is not small, regardless of whether these are defined with male or female cutoffs. Much remains to be learned about BMD and fracture risk in men, but it is likely that men with osteopenia and osteoporosis would benefit from taking steps to prevent further bone loss.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We wish to acknowledge Mrs. Lana Walters for her sustained dedication in processing the NHANES III bone density data, Dr. W. Edward Bacon for providing additional analyses of hip fracture discharge data from the National Hospital Discharge Survey, and Dr. Charles Slemenda for his thoughtful suggestions to improve the manuscript.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Melton LJ III, Atkinson EJ, O'Fallon WM, Wahner HW, Riggs BL 1993 Long-term fracture prediction by bone mineral assessed at different skeletal sites J Bone Miner Res 8:12271233.
  • 2
    Black DM, Cummings SR, Genant HK, Nevitt MC, Palermo L, Browner W 1992 Axial and appendicular bone density predict fracture in older women J Bone Miner Res 7:633638.
  • 3
    Cummings SR, Black DM, Nevitt MC, Browner W, Cauley J, Ensrud K, Genant HK, Palermo L, Scott J, Vogt TM for the Study of Osteoporotic Fractures Research Group 1993 Bone density at various sites for prediction of hip fractures Lancet 341:7275.
  • 4
    Kanis JL, Melton LJ, Christiansen C, Johnston CC, Khaltaev N 1994 The diagnosis of osteoporosis J Bone Miner Res 9:11371141.
  • 5
    World Health Organization, 1994 Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. Technical Report Series No. 843. WHO, Geneva, Switzerland.
  • 6
    Orwoll ES, Klein RF 1995 Osteoporosis in men Endocr Rev 16:87116.
  • 7
    Luz Villa M, Nelson L, 1996 Race, ethnicity and osteoporosis. In: MarcusR, FeldmanD, KelseyJ, (eds.) Osteoporosis. Academic Press, San Diego, CA, U.S.A., pp. 435448.
  • 8
    Bacon WE 1996 Secular trends in hip fracture occurrence and survival: Age and sex differences J Aging Health 8:538553.
  • 9
    Melton LJ, O'Fallon WM, Riggs BL 1987 Secular trends in the incidence of hip fractures Calcif Tissue Int 41:5764.
  • 10
    Looker AC, Johnston CC, Wahner HW, Dunn WL, Calvo MS, Harris TB, Heyse SP, Lindsay RL 1995 Prevalence of low femoral bone density in older U.S. women from NHANES III J Bone Miner Res 10:796802.
  • 11
    National Center for Health Statistics 1994 Plan and Operation of the third National Health and Nutrition Examination Survey, 1988–94. Vital Health Stat 1(32):DHHS Publ. No. (PHS) 94–1308. NCHS, Hyattsville, MD, U.S.A.
  • 12
    Wahner HW, Looker A, Dunn WL, Walters LC, Hauser MF, Novak C 1994 Quality control of bone densitometry in a national health survey (NHANES III) using three mobile examination centers J Bone Miner Res 9:951960.
  • 13
    Slemenda C, Longcope C, Peacock M, Hui S, Johnston CC 1996 Sex steroids, bone mass and bone loss: A prospective study of pre-, peri-, and postmenopausal women J Clin Invest 97:1421.
  • 14
    Mohadjer L, Montaquila J, Waksberg J, Bell B, James P, Flores-Cervantes I, Montes M, 1996 National Health and Nutrition Examination Survey III. Weighting and estimation metholodgy. WESTAT Inc., Rockville, MD, U.S.A.
  • 15
    Shah BV, Barnwell BG, Hunt PN, LaVange LM, 1991 SUDAAN, User's Manual Release 5.50. Research Triangle Institute, Research Triangle Park, NC, U.S.A.
  • 16
    Looker AC, Wahner HW, Dunn WL, Calvo MS, Harris TB, Heyse SP, Johnston CC, Lindsay RL 1995 Proximal femur bone mineral levels of US adults Osteoporos Int 5:389409.
  • 17
    Cummings SR, Black DM, Rubin SM 1989 Lifetime risks of hip, Colles', or vertebral fracture and coronary heart disease among white postmenopausal women Arch Intern Med 149:24452448.
  • 18
    Jones G, Nguyen T, Sambrook P, Kelly PJ, Eisman JA 1994 Progressive loss of bone in the femoral neck in elderly people: Longitudinal findings from the Dubbo osteoporosis epidemiology study Br Med J 309:691695.
  • 19
    Greenspan SL, Myers ER, Maitland LA, Kido TH, Krasnow MB, Hayes WC 1994 Trochanteric bone mineral density is associated with type of hip fracture in elderly J Bone Miner Res 9:18891894.
  • 20
    Karlsson MK, Johnell O, Nilsson BE, Sernbo I, Obrant KJ 1993 Bone mineral mass in hip fracture patients Bone 14:161165.
  • 21
    Nguyen T, Sambrook P, Kelly P, Jones G, Lord S, Freund J, Eisman J 1993 Prediction of osteoporotic fractures by postural instability and bone density Br Med J 307:11111115.
  • 22
    Greenspan SL, Myers ER, Maitland LA, Resnick NM, Hayes WC 1994 Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly JAMA 271:128133.
  • 23
    Orwoll E, Klein RF, 1996 Osteoporosis in men: Epidemiology, pathophysiology, and clinical characterization. In: MarcusR, FeldmanD, KelseyJ, (eds.) Osteoporosis. Academic Press, San Diego, CA, U.S.A., pp. 745784.
  • 24
    Genant HK, Gluer CC, Lotz JC 1994 Gender differences in bone density, skeletal geometry, and fracture biomechanics Radiology 190:636640.
  • 25
    Kelly TL 1990 Bone mineral density reference database for American men and women J Bone Miner Res 5:S249 (abstract).
  • 26
    Krall EA, Dawson-Hughes B 1993 Heritable and life-style determinants of bone mineral density J Bone Miner Res 8:19.
  • 27
    Eckert KL, Marcus R, Bachrach LK 1993 Bone mineral acquisition in healthy adolescents J Bone Miner Res 8:S264 (abstract).
  • 28
    Kellie SE, Brody JA 1990 Sex-specific and race-specific hip fracture rates Am J Public Health 80:326328.
  • 29
    Farmer ME, White LR, Brody JA, Bailey KR 1984 Race and sex differences in hip fracture incidence Am J Public Health 74:13741380.
  • 30
    Griffin MR, Ray WA, Fought RL, Melton LJ 1992 Black–white differences in fracture rates Am J Epidemiol 136:13781385.
  • 31
    Silverman SL, Madison RE 1988 Decreased incidence of hip fracture in Hispanics, Asians, and Blacks: California hospital discharge data Am J Public Health 78:14821483.
  • 32
    Bauer RL, Diehl AK, Barton SA, Brender J, Deyo RA 1986 Risk of postmenopausal hip fracture in Mexican Americans Am J Public Health 76:10201021.
  • 33
    Bauer RL 1988 Ethnic differences in hip fracture: A reduced incidence in Mexican Americans Am J Epidemiol 127:145149.
  • 34
    Wahner HW, Looker A, Walters L, Dunn WL 1994 Racial and gender differences in the shape of the proximal femur J Bone Miner Res 9 (Suppl 1):S401 (abstract).
  • 35
    Faulkner KG, Cummings SR, Black D, Palermo L, Glüer CC, Genant HK 1993 Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures J Bone Miner Res 8:12111217.
  • 36
    Peacock M, Turner CH, Liu G, Manatunga K, Timmerman L, Johnston CC 1995 Better discrimination of hip fracture using bone density, geometry and architecture Osteoporos Int 5:167173.
  • 37
    Villa ML, Marcus R, Ramirez-Delay R, Kelsey JL 1995 Factors contributing to skeletal health of postmenopausal Mexican-American women J Bone Miner Res 10:12331242.
  • 38
    Cummings SR, Cauley JA, Palermo L, Ross PD, Wasnich RD, Black D, Faulkner KG 1994 Racial differences in hip axis lengths might explain racial differences in rates of hip fracture Osteoporos Int 4:226229.
  • 39
    Ortiz O, Russell M, Daley T, Baumgartner R, Waki M, Lichtman S, Wang J, Pierson RN, Heymsfield SB 1992 Differences in skeletal muscle and bone mineral mass between black and white females and their relevance to estimates of body composition Am J Clin Nutr 55:813.
  • 40
    Melton LJ, Kan SH, Wahner HW, Riggs BL 1988 Lifetime fracture risk: An approach to hip fracture risk assessment based on bone mineral density and age J Clin Epidemiol 41:985994.
  • 41
    Ross PD, Wasnich RD, Davis JW 1990 Fracture prediction models for osteoporosis prevention Bone 11:327331.
  • 42
    Black DM, Cummings SR, Melton LJ 1992 Appendicular bone mineral and a woman's lifetime risk of hip fracture J Bone Miner Res 7:639646.
  • 43
    Favus MJ, 1993 Bone density reference data. In: FavusMJ, (eds.) Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism, 2nd Ed. Raven Press, New York, NY, U.S.A., pp. 426430.
  • 44
    Ezzati T, Khare M 1992 Nonresponse adjustments in a national health survey., 1992 Proceedings of the American Statistical Association, Section on Survey Research Methods. American Statistical Association, Alexandria, VA, U.S.A., pp. 339344.
  • 45
    Mohadjer L, Bell B, Waksberg J, 1994 National Health and Nutrition Examination Survey III. Accounting for item nonresponse bias. WESTAT Inc., Rockville, MD, U.S.A.
  • 46
    Rudman D, Drinka PJ, Wilson CR, Mattson DE, Scherman F, Cuisinier MC 1994 Relations of endogenous anabolic hormones and physical activity to bone density and lean body mass in elderly men Clin Endocrinol (Oxf) 40:653661.
  • 47
    Greenspan SL, Myers ER, Kiel DP, Hayes WC, Resnick NM 1995 Hip fracture in the nursing home: The role of bone density and fall severity J Bone Miner Res 10 (Suppl 1):S362 (abstract).