JOURNAL OF BONE AND MINERAL RESEARCH Volume 7, Number 5, 1992 Mary Ann Liebert, Inc., Publishers

Bone Mineral Density in Elderly Men and Women: Results from the Framingham Osteoporosis Study MARIAN T. HANNAN, DAVID T. FELSON, and JENNIFER J . ANDERSON

ABSTRACT Our study investigated bone mineral density of the proximal femur and ultradistal and proximal radius in a population of elderly men and women. The Framingham study started in 1948, following a population-based sample for evaluation of cardiovascular risk factors and events. During the 20th biennial Framingham examination (1988-89) we conducted the Framingham osteoporosis study, measuring bone mineral density in the proximal femur and distal and proximal radius for 1154 study participants. Ages ranged from 68 to 98 years, with a mean age of 76 years. Bone mineral density was measured using Lunar SP2 and DP3 densitometers. This cross-sectional study evaluates mean bone mineral density measurements at each site by 5 year age intervals for men and women, testing for trends in bone density with age. Analyses were repeated adjusting for weight and height. Among the 446 and 708 women, bone mineral density of the femur and bone mineral content of the proximal radius were inversely and significantly related to age in both sexes and were considerably higher in men than women at all sites. The linear decline with age group in our cross-sectional study remained after multivariate adjustment for height and weight. The ultradistal radius showed no significant correlation with age for either sex. There were significant correlations between the bone measurements made at different sites for both men and women (range in r = 0.27-0.89). Cross-sectional curves of bone mineral density with age showed no significant differences in slope between males and females. In this cross-sectional study, we found that increased age was significantly associated with decreased bone mass in a linear and equivalent fashion for both men and women through the elderly years in the proximal femur and proximal radius.

INTRODUCTION has been extensively studied in peri- and postmenopausal women.('-'4) Loss in axial trabecular bone occurs in women before menopause, and then bone loss accelerates in the first 6-10 years after the menopause, perhaps slowing in later life. (3-8.12.15-18)Bone loss in males is less well defined and has not been well studied in men of either middle or elderly years. With the accurate delineation of the bone mass curve in women in and around the menopause, attention has shifted to bone mineral density among adults in their older years. The rate of change in bone mass is relevant to the risk of fracture faced by the elderly, since in older years adults are highly susceptible to relatively atraumatic frac-

B

ONE MINERAL DENSITY

tures associated with bone fragility. The absolute levels of bone mineral density may determine the susceptibility of elderly adults to fractures as they age. Furthermore, a recent study by Davis et of age-related bone loss rates found that the longitudinal and cross-sectional estimates were similar. Studies of bone mineral density that have included elderly subjects have generally studied a large age range without a specific focus on bone mass in the elderly. ( L - 4 ~ B - 1 0 ~ L 3 ~ 1 5 ~ 1 7 ~ 1As 9 - 2noted 1) by some of these studies, bone mineral loss in elderly years may slow [see especially Davis et and Hui et al.(31],and this may be modeled statistically by nonlinear regression models. Nonetheless, these models generally rely on sparse data about subjects in the oldest years.

Boston University Arthritis Center, Departments of Medicine, University Hospital, and Boston City Hospital, Massachusetts

547

548

HANNAN ET AL.

Although bone mineral density in women has been well at a nursing home did not receive bone mineral densitomestudied, men have been understudied. However, a seventh ter measurements. These were essentially the only subjects of all vertebral compression fractures and a fourth to a missed, as fewer than 2% of the participants who came to fifth of all hip fractures occur in men.('*) New studies have the Framingham study clinic for examination refused bone included parallel assessments of men and women to deter- density measurement. mine whether their bone loss curves were similar. (1*.19.21.*3) As part of the routine biennial examination, which inFurthermore, studies of bone mineral density have cluded assessment of cardiovascular risk factors and end points, medical history, and physical examination, we almost all been studies of vo1unteers.'1.3.6~7~10,11~17~19~z4~ Those few studies that have attempted t o comprehensively measured bone mineral density in the proximal femur and evaluate population-based groups have often excluded sub- the proximal radius. At that time, because of the length of jects with any risk factors known to affect bone mineral the combined examination, we asked subjects to return for density, such as estrogen or thiazide use. (3.6,8,9.1524)These further bone mineral density examination involving the restrictions may not bias study estimates relative to popu- distal radius. Lumbar spine measurement was not a focus lation means for subjects in their middle years, but among of this study because of problems with measurement of elders the exclusion of subjects with various exposures and this site in the elderly. Specifically, aortic calcificathe requirement for subjects to be volunteers may result in tions, vertebral osteophytes, and facet joint osteoarthritis biases. Elderly subjects who are active volunteers for stud- all may confound bone density measurement. In addition, ies and who have no other ongoing medical problems that wedge fractures can spuriously increase bone density. might affect bone mineral density are a minority of the X-rays of the spine to identify vertebral compression fracelderly population. Only the Kuakini osteoporosis study tures were not performed at this examination. Bone mineral density measurement was performed using attempted to systematically examine a population-based sample of ambulatory elderly subjects, and they studied a Lunar SP2 single-photon absorptiometer (using only the arm and 0s calcis sites.(zs)Indeed, few studies source) for the proximal radius and ultradistal radius sites have attempted to measure bone mineral density at multi- and a Lunar DP3 dual-photon absorptiometer (using a ple sites in their subjects. Recent work suggests that among 153Gdsource) for the proximal femur site (Lunar Radiation the elderly, the bone mineral density of appendicular sites Corporation, Madison, WI). Standard positioning was comprised primarily of trabecular bone (e.g., the ultradis- used, including medial rotation of the femur to obtain a tal radius) may remain relatively stable with age,(6.1*.15) clear scan of the femoral neck. For the arm and femur whereas bone densities in other sites (e.g., the femur), cor- sites, the right side was scanned except in cases of prior fracture or joint replacement, when the contralateral side relate negatively with age. We examined bone mineral density in an ambulatory was scanned. Femoral densities were measured at three population-based group of elderly subjects, the surviving sites: femoral neck, trochanter, and Ward's triangle. An members of the Framingham heart study cohort. This anatomic phantom (Hologic) of the hip was used to test group, which in 1948 was a one-third sample of Framing- day-to-day reproducibility and to determine whether a ham residents aged 30-60 years, has been followed for over change in the gadolinium source resulted in a shift in mea40 years. Now elderly, they participated in a bone mineral surement. We found no shift or any change in day-to-day density study during a biennial examination 20 (1988-89), variability with change in radioactive source. In addition, during which bone mineral density was measured at distal young normal subjects were used to assess reproducibility and proximal radius and proximal femur. The Framing- across the 2 year study at all sites. The coefficient of variation of the femur measurements ham cohort, in addition to being population based (not composed primarily of volunteers or those selected due to (n = 68), using the bone phantom ranged from 1.49% disease status), includes large numbers of elderly men as (trochanter) and 2.85% (Ward's triangle) to 4.15% well as women, allowing some insight into the distribution (femoral neck). Using three young normal controls across of bone mass for both older men and women. The age and a 2 year period, the coefficient of variation ranged from sex distribution of the Framingham cohort survivors is 2.65% (femoral neck) and 2.80% (trochanter) to 4.16% close to the estimated 1990 population distribution of (Ward's triangle). The proximal radius coefficient of variation was 3.94% using young normal controls, and the elders for the town of Framingham.~z6) ultradistal reliability was 5.73%. These values are consistent with coefficients of variation reported by other reMETHODS searchers, (5,73.1227) although the ultradistal radius value is The Framingham heart study began in 1948 with the pri- somewhat higher than that reported by others. mary goal of longitudinally evaluating the risk fractors for heart disease. The Framingham study cohort has been exA nalysis amined every 2 years. Since the inception of the cohort 40 years ago, over half the original cohort members have For each sex, we looked at the average bone mineral died. Of 1402 surviving subjects taking part in biennial ex- density measurement at each site according to 5 year age amination 20, 1164 (83%) participated in the bone mineral groups. We then tested whether a linear association existed density study, including several nursing home residents between bone mineral density and age within each sex and who were ambulatory. Because of the lack of machine mo- at each site. We examined the relation of age with bone bility, the 213 participants who were examined at home or mineral density using a simple linear model and also using

549

BONE DENSITY IN THE ELDERLY quadratic, cubic, and exponential functions of age. We did not find that these additional tests yielded any significant improvement in the fit of the model compared with a simple linear model. In fact, models of bone mineral density with cubic, quadratic, and exponential functions of age were all nonsignificant. These analyses were repeated after adjustment for typical adult weight (kilograms) and height (centimeters) as assessed by measured weight and height at biennial examination 1 (1948-51) to examine the possible effects of age on bone density independent of body mass. We used typical adult weight and height since the rapidity of bone loss may depend to a certain extent on weight or height loss, activity, and other age-related factors. The relation of bone mineral density to weight loss, height loss, and other agerelated factors will be examined in future analyses. Finally, we examined the correlations of bone mineral density at the three femoral sites and bone mineral content of the arm and wrist sites.

RESULTS Of the 1164 subjects who participated in the Framingham osteoporosis study, 448 were male and 716 were female (Table 1). The age of subjects ranged from 68 to 98 years, with a mean age of 76 years. Most subjects who obtained femur and proximal radius scans returned for evaluation of ultradistal radius. Table 1 also shows the characteristics of those subjects on whom bone densitometry was not performed. These 213 subjects were seen at home or in a nursing home and thus did not come to the Framingham clinic for examination 20 and could not have bone mineral density measured. This group was older and had a slightly greater proportion of women, an expected result since these traits are characteristic of a nursing home and homebound population.

Femoral bone mineral densities by sex and age group and bone mineral content in the radius sites by sex and age group are shown in Table 2. As seen in Table 2, bone mineral density and bone mineral content were considerably higher in men than women at all sites and generally declined linearly with age group. The number of subjects was greatest in the age group 68-74 and was much less in more elderly years. In both men and women, the femoral bone mineral density and bone mineral content of the proximal radius fell linearly with age (Table 3). This age trend was not significant for bone mineral content in the ultradistal radius in either sex. The significant decrease in bone mineral density with age in the femur and proximal radius sites remained after multivariate adjustment for weight and height. The rates of decline in bone mass were quite similar for men and women at all sites (see Fig. 1). In fact, tests to determine whether these slopes were different between sexes showed no significant differences at any sites. All the bone density measurements at the different sites in our study were statistically significantly correlated (p < 0.oOOl) for men and women (Table 4). Within the femur sites, correlation coefficients ranged from 0.67 to 0.89; the arm sites had coefficients ranging from 0.27 to 0.52. The ultradistal radius site was the area with the lowest correlation (0.27) with the densities in the other areas under evaluation, and densities for the femoral sites had the largest correlation coefficients.

DISCUSSION Bone mineral density falls linearly through the elderly years in the proximal femur and proximal radius. The decrease with age may be less for the ultradistal radius. Bone mineral density loss curves showed no significant differences in slope between men and women.

TABLE1. CHARACTERISTICS OF FRAMINGHAM PARTICIPANTS WHOHADBONEDENSITOMETRY VERSUSTHOSEWHODID NOTHAVEBONEDENSITOMETRY

070 Female Ages 68-74 75-79 80-84 85-89 90-98 Age (mean f SD, range) Height (mean f SD, range), cm Weight (mean f SD, range), kg

Bone densitometry (clinic examinees), N = 1164

No bone densitometry (nursing home or home examinees), N = 213

N%

N %

716 61.5

141 66.2

533 45.8 349 30.0 195 16.8 66 5.1 21 1.8 76.1 f 5.19, 68-98 163.19 f 10.18, 129-193 70.04 f 14.15, 36.4-149.6

76 35.7 59 27.7 39 18.3 22 10.3 17 8.0 78.1 f 6.88 68-96 NAa,b NA NA NA ~~

aNA, not applicable.

bparticipants and those not participating in the bone densitometry examination had similar heights and weights at examination 1.

H A N N A N ET AL.

550

TABLE 2. MEANBONEMINERAL CONTENT OR DENSITY AT RADIUSAND FEMURSITESBY SEXA N D AGE Radius mean BMC f SD (g/cm) Age

N

Proximal

N

Proximal femur mean BMD f SD (g/cm2)

Ultradistal

N

Neck

Trochanter

Ward’s triangle

Men

68-74 75-79 80-84 85-89 90-96

221 125 71 21 8

1.06 f

0.17 0.18 0.20 0.14 0.22

173 97 52 I5 4

1.93 f 1.91 f 1.82 f 1.91 f 1.90 f

0.44 0.48 0.36 0.44 0.41

216 122 71 21 7

0.89 f 0.86 + 0.83 f 0.79 f 0.74 f

310 219 122 44 13

0.69 f 0.13 0.64 f 0.13 0.62 f 0.12 0.60fO.11 0.59 f 0.12

247 180 87 30 5

1.09 f 0.29 1.07 f 0.32 1.06 f 0.29 1.08f0.29 1.26 f 0.31

302 220 122 43 11

0.73 f 0.12 0.71 f 0.12 0.67 f 0.10 0.67ziz0.10 0.64 f 0.13

1.15 f 1.14 f 1.13 f 1.13 f

0.14 0.15 0.14 0.12 0.15

0.86 f 0.85 f 0.83 f 0.81 f 0.77 f

0.17 0.14 0.14 0.21 0.11

0.70 f 0.69 f 0.63 f 0.64 f 0.53 f

0.17 0.18 0.16 0.13 0.19

Women

68-74 75-79 80-84 85-89 90-96

TABLE 3. CHANGEIN BMD AND BMC PER YEAROF AGE AND

0.64 f 0.13 0.63 f 0.14 0.59 f 0.11 0.61f0.12 0.51 f 0.10

vo CHANGEPER

Men Change in g/cm’/year f SD Femur Neck Trochanter Ward’s triangle

-0.006 -0.004

-0.006

f f f

0.0014b 0.0015b 0.0016b

Change in g/cm/year f SD Radius Proximal Ultradistal

-0.003

f

-0.006

f

0.0016C 0.0049

0.58 f 0.13 0.56 f 0.13 0.52 f 0.10 0.52ztz0.10 0.48 f 0.16

YEAR^

Women

Change in % BMD per year f SD

-0.69% f 0.15 -0.45% f 0.17 -0.88% f 0.23 Change in % BMC per year f SD

-0.26% -0.30%

f f

0.14 0.25

Change in g/cm’/year f SD

-0.005 -0.003

f f -0.005 f

O.ooo9b 0.001Ob O.ooo9b

Change in g/cm/year f SD

-0.006 -0.002

f

0.0009b 0.0026

Change in 9% BMD per year f SD

-0.68% f 0.12 -0.53% f 0.15 -0.94% f 0.16 Change in Yo BMC per year f SD -0.88% f 0.13

-0.16%

f

0.24

aRegression coefficients and standard errors from regression of bone mass loss rate as a function of age (ages 68-98). Percentage loss rates figured by dividing the loss rate by the mean bone mineral density value for each sex. bp value < 0.05 for linear trend. cp value = 0.07 for linear trend.

The bone mineral densities in our study are similar to re~ the . ~radius ~ ~ re~ ~ sults reported by ~ t h e r ~ , ‘ ~ .and sults especially are similar to those of a longitudinal study by Davis et al.,1151 suggesting that longitudinal and crosssectional data d o not differ much among the elderly. Yano et al.,(25)Mazess et al.,11*,19) and Pocock et al.Il6) report similar age-specific bone mineral densities at the radius and femur sites as found in our study; however, their studies did not include subjects over age 80. A number of authors note that a linear relationship best fit the bone mass and age data for the proximal radius and distal ra~ ~ U S ~ l ~ . 9 . 1 5 . 1 7Davis . ~ 5 . ~ et ~ l al .(I5) present slopes of bone loss rates of the proximal and ultradistal radius for women over 70 in the longitudinal component of their study that were similar to those we found in older women in our cross-sectional study (Table 3). Our results of the means for the femoral neck, trochanter, and Ward’s triangle are similar to those of Mazess et al.‘12.19) for both men and women. Although Mazess et al. reported relative stabilization of bone mineral density of the trochanter and Ward’s

triangle after age 70, our study, with greater numbers in ~the~ older ~ ~ age ~ ~groups ~ * ~of ~the~ elderly, ~ ) showed a continued decrease in mean bone mineral density with increased age. Several studies discuss the differential rates of bone loss at different body sites in the elderly.ca.1’~14-1s.’4~z7~34~ For example, the elderly experience more rapid loss of bone from femoral sites than spine sites, perhaps because spinal bone loss occurred at an earlier age.(z4,z51Although bone loss appears to differ with age, prior studies that included elderly subjects focused on subjects 50 years and older, not on the elderly. Distal trabecular bone loss in the elderly may level with age. Bone mineral loss in the elderly appears to continue with age in the sites composed primarily of cortical bone. This study is limited in several regards. First, the study is cross-sectional and did not evaluate individual change in bone mass across time. Thus, individual bone loss rates are not known for our study participants; however, cross-sectional data may describe longitudinal bone loss with age if the bone densities of the younger age groups are similar to

BONE DENSITY IN THE ELDERLY

551

B

~

1.5

1

2.0

1.5

.',

1.5

m

" -. 0

n m

I

r,

II

(1

~- -

. (

j

I

a5

87

Age

90

95

0 0 I

65

70

75

a)

fll

90

I S

1

Age

FIG. 1. Linear associations of age with femoral bone mineral densities and proximal radius bone mineral content for men and women: (A) radial BMC; (B) femoral neck BMD; (C) trochanteric BMD; (D) Ward's triangle BMD. Dotted lines indicate 95% confidence intervals around the regression lines.

H A N N A N ET AL.

552

TABLE4. CORRELATION COEFFICIENTS OF RADIUSAND FEMURSITESAND AGE FOR MENAND WOMEN^

Men Age Proximal radius Ultradistal radius Femoral neck Trochanter Ward’s triangle Women Age Proximal radius Ultradistal radius Femoral neck Trochanter Ward’s triangle

Age

Proximal radius

Ultradistal

Neck

Trochanter

Ward’s triangle

1

-0.088

-0.065 0.436 1

-0.213 0.295 0.415

-0.126 0.278 0.429 0.725

-0.180 0.271

1

1

0.404

0.866 0.674

1

1

1

-0.239 1

-0.028 0.502 1

-0.214 0.52 1 0.369 1

1

-0.217 0.525 0.356 0.896 0.692 1

aAll p values < O.OOO1 except ultradistal radius and age for both men and women

those of older participants at those same younger age groups. Another limitation concerns the imprecision of our ultradistal radius measurement, which may have affected our determination of whether there was age-related bone loss at this site. This site is often difficult to measure precisely since a slight twisting of the hand can change the ultradistal radius site results. Despite the imprecision in the ultradistal site measurement, our study had sufficient power to detect a quite small drop in density in the under 75 versus 75 years and older individuals (i.e., as 0.25 standard deviation, SD, for each gender separately, 80% power, one-sided test at 0.05 level), if such a difference was present. The drops in density observed for the other sites were of this order. Other studies have reported little age-related ultradistal radius bone loss in the elderly,(31 3 ) and it is likely that there is a lack of age-related bone loss in the ultradistal site. Finally, we studied a population-based group of ambulatory elderly. Nonambulatory, institutionalized elderly persons are not included in our figures, however, and convalescence and inactivity among the elderly have been shown to affect bone mass. Nursing home patients, often a frail group, may have a lower mean bone mineral density than same-aged ambulatory individuals. In addition, Browner et al. reported an association between mortality and low bone mass of the proximal radius in elderly women. (3s) They remark that an association between bone density and mortality is probably not causal and suggest that low bone mineral density is a marker of fragility or overall health status. If those with low bone mineral density have higher mortality or are more likely to be nonambulatory, then our cross-sectional data may underestimate population bone loss in older age groups. Finally, there are very few nonwhite subjects in the Framingham sample, and our results are not generalizable to this group. Since our results are similar to those of earlier cross-sectional studies, it is likely that the possible bias introduced in those studies by using volunteers was small. However, some reports of bone density stabilization in older elders

-0.135 0.504 0.378 0.727

J(J

=

0.23 and 0.52, respectively).

may be an artifact since these cross-sectional studies included volunteer elderly subjects who would tend to be healthier than other elderly subjects. In summary, our study provides information on bone mineral density in the elderly for two anatomic regions from a population-based sample of ambulatory persons. Data are presented for each sex by 5 year age groups, thus providing information on men as well as the very elderly. Men and women showed similar bone density curves with age, although women had consistently lower bone mineral density. In conclusion, our results show a decrease in bone mineral density with increasing age for femur and proximal radius sites. This decrease with age was similar in men and women. We found no significant association of bone mass with age in the ultradistal radius, but imprecise measurement makes interpretation difficult.

ACKNOWLEDGMENTS We are grateful to the Framingham cohort participants and staff and also thank the densitometer technicians, Eileen DelVecchio, Mimi Brodsky, and Cherlyn Mercier, for their invaluable assistance and support. This work was presented in part as a poster at the 1 lth Annual Scientific Meeting of the ASBMR in Atlanta; J Bone Miner Res 5 (Suppl. 1):5182, 1990. This work was supported by Arthritis Center Grant No. AR 20613 from the National Institutes of Health.

REFERENCES 1. Hagiwara S , Miki T, Nishizawa Y , Ochi H , Onoyama Y ,

Morii H 1989 Quantification of bone mineral content using dual-photon absorptiometry in a normal Japanese population. J Bone Miner Res 4:217-222. 2. Hansson T, Roos B 1986 Age changes in the bone mineral of the lumbar spine in normal women. Calcif Tissue Int 38:24925 1.

BONE DENSITY IN THE ELDERLY 3. Hui SL, Wiske PS. Norton JA, Johnston Jr CC 1982 A prospective study of change in bone mass with age in postmenopausal women. J Chron Dis 35:715-725. 4. Linquist 0, Bengtsson C, Hansson T, Jonsson R 1983 Changes in bone mineral content of the axial skeleton in relation to aging and the menopause. Scand J Clin Lab Invest 4 3 ~333-338. 5 . Kleerekoper M, Peterson E, Nelson D, Tilley B, Phillips E, Schork MA, Kuder J 1989 Identification of women at risk for developing postmenopausal osteoporosis with vertebral fractures: Role of history and single photon absorptiometry. Bone Miner 7:171-186. 6. Ribot C, Tremolliers F, Pouilles JM, Louvet JP, Guiraud R 1988 Influence of the menopause and aging on spinal density in French women. Bone Miner 5:89-97. 7. Nilas L, Christiansen C 1988 Rates of bone loss in normal women: Evidence of accelerated trabecular bone loss after the menopause. Eur J Clin Invest 18529-534. 8. Riggs BL, Wahner HW, Melton J L 111, Richelson LS, Judd HL, Offord KP 1986 Rates of bone loss in the appendicular and axial skeletons of women: Evidence of substantial vertebral bone loss before menopause. J Clin Invest 77:14871491. 9. Hui SL, Slemenda CW, Johnston CC Jr, Appledorn CR 1987 Effects of age and menopause on vertebral bone density. Bone Miner 2:141-146. 10. Mazess RB 1982 On aging bone loss. Clin Orthop 165:239252. 11. DeSimone DP, Stevens J, Edwards J, Shary J, Gordon L, Bell NH 1989 Influence of body habitus and race on bone mineral density of the midradius, hip and spine in aging women. J Bone Miner Res 4:827-830. 12. Mazess RB, Barden HS, Ettinger M, Johnston C, DawsonHughes B, Baran D, Powell M, Notelovitz M 1987 Spine and femur density using dual-photon absorptiometry in US white women. Bone Miner 2:211-219. 13. Eastell R, Riggs BL, Wahner HW, O’Fallon WM, Amadio PC, Melton LJ 111 1989 Colles’fracture and bone density of the ultradistal radius. J Bone Miner Res 4:607-613. 14. Heuck AF, Block J, Glueer CC, Steiger P , Genant HK 1989 Mild versus definite osteoporosis: Comparison of bone densitometry techniques using different statistical models. J Bone Miner Res 42391-899. 15. Davis JW, Ross PD, Wasnich RD, Maclean CJ, Vogel JM I989 Comparison of cross-sectional and longitudinal measurements of age-related changes in bone mineral content. J Bone Miner Res 4:351-357. 16. Pocock NA, Eberl S, Eisman JA, Yeates MG, Sambrook PN, Freund J, Duncan A 1987 Dual-photon bone densitometry in normal Australian women: A cross-sectional study. Med J Aus 146:293-297. 17. Schaadt 0, Bohr H 1988 Different trends of age-related diminution of bone mineral content in the lumbar spine, femoral neck and femoral shaft in women. Calcif Tissue Int 42~71-76. 18. Riggs BL, Wahner HW, Dunn WL, Mazess RB, Offord KP, Melton LJ 1981 Differential changes in bone mineral density of the appendicular and axial skeleton with aging. J Clin Invest 67:328-335. 19. Mazess RB, Barden HS, Drinka PJ, Bauwens SF, Orwoll ES, Bell NH 1990 Influence of age and body weight on spine and femur bone mineral density in U.S. white men. J Bone Miner Res 5:645-652.

553 20. Orwoll ES, Oviatt SK, Mann T 1990 The impact of osteophytic and vascular calcifications on vertebral mineral density measurements in men. J Clin Endocrinol Metab 70:12021207. 21. Thomsen K, Gotfredseri A, Christiansen C 1986 Is postmenopausal bone loss an age-related phenomenon? Calcif Tissue Int 39:123-127. 22. Jackson JA, Kleerekoper M 1990 Osteoporosis in men: Diagnosis, pathophysiology, and prevention. Medicine (Baltimore) 69:137-152. 23. Kelly PJ, Twomey L, Sambrook PN, Eisman JA 1990 Sex differences in peak adult bone mineral density. J Bone Miner Res 5:1169-1175. 24. Meier D, Orwoll ES, Jones JM 1984 Marked disparity between trabecular and cortical bone loss with age in healthy men. Ann Intern Med 101:605-612. 25. Yano K, Wasnich RD, Vogel JM, Heilbrun LK 1984 Bone mineral measurements among middle-aged and elderly Japanese residents in Hawaii. Am J Epidemiol 119:751-764. 26. Massachusetts Institute for Social and Economic Research 1988 Massachusetts estimated 1990 city and town populations. University of Massachusetts Press, Amherst. 27. Riggs BL, Wahner HW, Seeman E, Offord KP, Dunn WL, Mazess RB, Johnson KA, Melton LJ Ill 1982 Changes in bone mineral density of the proximal femur and spine with aging: Differences between the postmenopausal and senile osteoporosis syndromes. J Clin Invest 70:716-723. 28. Melton LJ 111, Kan SH, Frye MA, Wahner HW, O’Fallon M, Riggs BL 1989 Epidemiology of vertebral fractures in women. Am J Epidemiol 129:1000-1011. 29. Hopkins A, Zylstra S, Hreshchyshyn M 1988 Normal and abnormal features of the lumbar spine observed in dual-photon absorptiometry scans. Clin Nucl Med 14:410-714. 30. Smith-Bindman R, Cummings SR, Steiger P, Genant HK 1991 A comparison of morphometric definitions of vertebral fracture. J Bone Miner Res 6:25-34. 31. Ross PD, Davis WJ, Vogel JM, Wasnich RD 1990 A critical review of bone mass and the risk of fractures in osteoporosis. Calcif Tissue Int 46:149-161. 32. Stutzman ME, Yester MV, Duborsky EV 1987 Technical aspects of dual photon absorptiometry of the spine. J Nucl Med Techno1 15177-181. 33. Dawson-Hughes B, Dallal GE 1990 Effect of radiographic abnormalities on rate of bone loss from the spine. Calcif Tissue Int 46:280-281. 34. Riggs EL, Melton LJ 111 1983 Evidence for two distinct syndromes of involutional osteoporosis. Am J Med 75:899-901. 35. Browner WS, Seeley DG, Black D, Cauley J, Hailey SB, Cummings SR, and Study of Osteoporotic Fractures Research Center 1990 The association between bone mineral density and mortality in elderly women. In: Christiansen C and Overgaard K (eds.) Osteoporosis 1990. Osteopress, Copenhagen, pp. 71-72.

Address reprint requests to: Marian T. Hannan Boston University School of Medicine 80 East Concord Street A203 Boston. MA 02118 Received for publication July 29, 1991; in revised form December 12, 1991; accepted December 13, 1991.