JOURNAL OF BONE AND MINERAL RESEARCH Volume 6, Number 10, 1991 Mary Ann Liebert, Inc., Publishers

Bone Mineral Density and Mass by Total-Body Dual-Photon Absorptiometry in Normal White and Asian Men MARY RUSSELL-AULET, JACK WANG, JOHN THORNTON, EDWARD W.D. COLT, and RICHARD N. PIERSON, JR.

ABSTRACT Total body bone mineral density, total body bone mineral mass (TBBM), and bone mineral densities (BMD) in seven different regions of the body were measured in 238 normal men (154 Whites and 84 Asians), age 2294, using dual-photon absorptiometry. Although Asian men had lower TBBM and BMD in all regions (p c 0.05) except the arms, when multiple regression was done with body weight, height, and age, no significant differences were found between Asians and Whites for bone measurements. Thus lower bone mineral densities and bone mineral mass in Asian males compared to White males appear to be due to differences in height and weight rather than to ethnic differences. The two groups were combined to derive multiple regression equations for TBBM, total-body BMD, and regional BMD. Age, weight, and height were significant in the multiple regression equations for TBBM, ribs BMD, and legs BMD. Age and weight, but not height, were significant for total-body BMD, trunk BMD, spine BMD, arm BMD, and pelvis BMD. Weight, but neither height nor age, was significant for head BMD. These reference normal bone mineral density and bone mineral mass standards are appropriate for both Asian and White males when adjusted for weight, height, and age.

INTRODUCTION

Both Asian and White males were included in this study to determine whether normal reference bone mineral density (BMD) and bone mineral mass ranges for males are different between Asians and Whites. The purpose of this paper is twofold. One is to establish normal standards for males for total-body bone mineral density and mass. The other is to compare Asian and White males to determine whether different normal reference ranges by ethnic group are required.

STEOPOROSIS has been considered primarily a disease that affects females. However, Gallagher et al. indicate that 21% of all nontrauma, nonmetastatic hip fractures were in males.(11As the average age of the general population increases, the incidence of osteoporotic fractures in men may also be expected to increase. To evaluate bone mineral measurement results for a specific male patient, normal reference ranges are needed for males. NorMETHODS AND MATERIALS mal reference ranges have been published for males for specific sites of the body, such as the lumbar spine, femur, and radius.(z-81Reference normal standards for total-body A total of 238 adult males (154 White and 84 Asians) bone mineral values in males are needed. were recruited from the hospital staff, by poster advertiseAsians living in the United States have fewer hip frac- ment, and through various civic and social groups in the t u r e ~ ‘ ~despite ) having lower bone density values.‘lO,ll’ area. Asians were defined as having both parents and all

0

Department of Medicine, Body Composition Unit, St. LukeWRoosevelt Hospital Center, Columbia University College of Physicians and Surgeons, New York, New York.

1109

1110

RUSSELL-AULET ET AL. grandparents from an Asian country. There were 77 Chinese, 3 Japanese, 3 Korean, 1 Philippine, and 1 Indonesian subjects. Only I 1 of these Asians were born in the United States: 5 Chinese, 2 Japanese, and 2 Koreans. Whites were of European background. All subjects were at the time of study living in the United States. All subjects were ambulatory, and none used maintenance steroids or high-dose diuretic medication. Total-body bone mineral mass (TBBM) and total-body BMD were measured by dual-photon absorptiometry (Lunar DP4, software version 5E) as described by Peppler et aI.l1*)Software algorithms divide the body into head, trunk, spine, arms, ribs, legs, and pelvis (see Fig. I). These regions are graphically displayed, and the final cut lines dividing the regions are adjusted by the operator to account for individual variation. Each study required approximately 50 minutes lying supine. The reproducibility of the measurements was established by scanning four subjects five times in a 3 week period. Although reproducibility data were obtained in only 4 of the 238 subjects, they included low and high extremes of totalbody bone density range (0.955-1.425 g/cm’). All subjects gave informed consent to the testing as approved by the St. LukeWRoosevelt Hospital Center Internal Review Board.

Statistics Intraclass correlations and 95% confidence intervals were calculated.(13’To test for differences in height, weight, age, body mass index (BMI), TBBM, total-body BMD, and regional BMD of Asian versus White groups, ttests were done. Multiple regression analysis was done to determine if ethnic group, age, weight, and height added FIG. 1. Graphic image from DPA scan with regional di- significantly to the prediction of TBBM, total-body BMD, and BMD of the different regions of the body. visions.

OF TOTAL-BODY SCAN MEASUREMENTS~ TABLE1. REPRODUCIBILITY

Intraclass correlation 95Vo Confidence intervals

SD

cv

Measurement

Estimate

Lower limit

Upper limit

(%)

TBBM, g BMD, g/cm’ Total body Head Trunk Spine Arms Ribs Legs Pelvis

0.997

0.989

0.999

1.4

39

0.997 0.986 0.997 0.976 0.939 0.991 0.997 0.987

0.987 0.945 0.986 0.909 0.784 0.966 0.988 0.950

0.999 0.999 0.999 0.998 0.996 0.999 0.999 0.999

0.9 2.0 1.1 2.8 4.8 1.3 1.1 2.9

0.0109 0.0437 0.0108 0.0325 0.0422 0.0094 0.0134 0.0338

~

In trasubject

Bet ween subjects 760 0.1908 0.3670 0.1836 0.2091 0.1658 0.1010 0.2489 0.2990

~

aFour subjects had five DPA scans within a 3 week period. TBBM, total-body bone mineral mass; BMD, bone mineral density; CV, coefficient of variation = 100 x SDlntrasublect/mean.

1111

DPA NORMALS FOR ASIAN AND WHITE MALES TABLE2. CHARACTERISTICS OF SAMPLE STUDIEDa Age (years)

Height (cm)

Weight (kg)

BMI (kg/m2)

50 f 16

175 f 7b (1 58-194) 170 f 6b (157-185) 173 f 7 (157-194)

76 f l l b (55-1 16) 67 f 8 b (48-91) 73 f 11 (48-1 16)

25 f 3 b (18-34) 23 f 3 b (17-30) 24 f 3 (1 7-34)

Whites, N = 154 Asians, N = 84 Total, N = 238

(22-94) 51 f 17 (20-81) 51 * 17 (20-94)

aMeans. standard deviations, and ranges. bp < o.oO01.

RESULTS Reproducibility data are shown in Table 1. Intraclass correlations ranged from 0.939 to 0.997. Coefficients of variation ranged from 0.9 to 4.8%. Physical characteristics of the sample studied are shown in Table 2. Asians were shorter, lighter, and had a lower body mass index (BMI; p < 0.OOOl) than Whites. Age was not significantly different between Asians and Whites. Asians had lower TBBM, total-body BMD, and regional BMD in all areas except the arms (see Table 3). The results of multiple regression of bone measurements against ethnic group, age, weight, and height are shown in Table 4. Weight was a significant predictor for all bone measurements by multiple regression prediction equations. Age was significant in all equations, except for the prediction of head bone density. Height was significant only in the predictive equations for TBBM, rib BMD, and leg BMD. Ethnic group was not a significant predictor of TBBM, total-body BMD, or any regional bone density measures once weight, height, and age were taken into consideration.

DISCUSSION General impressions in the literature suggest that Asians have lower bone mineral densities and bone mineral mass than Whites.(lO.") Indeed, our results show that Asian males had lower TBBM, total-body BMD, and regional BMD (except for the arms) than White males. Asian males were also smaller (i.e., lighter and shorter) than White males. When bone measurements were adjusted for age, weight, and height, however, White-Asian bone mineral differences no longer existed. Thus the lower bone density values in age-matched Asian males appears to be due to differences in weight and height between the two ethnic groups. These results are similar to a study reported by Yano et al. that found age and weight to be significant predictors (by multiple linear regression) of forearm and 0s

calcis bone mineral content in 1368 Japanese-American males; height was also significant in forearm bone mineral content. ( l o ) Bone mineral density of the head is anomalous, showing no consistent relationship to any parameter except weight, where r' is only 0.12,a finding similar with the Gallagher group's study in premenopausal women. 1 4 ) Otherwise, weight and age are dominant predictors of all bone mineral mass and BMD measurements. Although height by itself is correlated with all measurements, when weight and age are included in multiple regressions, height no longer improves the prediction except in TBBM, rib BMD, and leg BMD. The relationship of height with leg BMD is similar to Gallagher et al.' finding(l4)in premenopausal women and suggests that taller people may stress their longer legs more, leading to increased leg BMD. Ribs, being neither weight bearing nor obviously related to height, showed a slightly negative function with height. Spine BMD decreased less with age than any other region of the body. This finding is in agreement with other studies that show that in normal males, the lumbar BMD decrease less rapidly with age than other parts of the body, such as the f e m ~ r . ' ~ ~ ~ ~ ' ' The hypothesis that a different relationship (slope) exists between the bone loss and age before and after age 40 was explored. Adjusting for weight and height, the hypothesis that age coefficient and intercept are the same at age 540 versus age >40 was tested and not rejected (p = 0.6 for TBD and p = 0.5 for TBBM). The same test was done for age breaks at age 50 and 60 with similar results (p values range from 0.4 to 0.9). Also, the residual analysis did not give any suggestion that the models were not valid representations for these data. Of course, adjusting for weight and height is critical. If there is in fact a nonlinear (age break) relationship between bone loss and age in the height- and weight-adjusted data, it is not significant, that is, not strong enough to be detected with this data set. Indeed, our data are consistent, with no age break. It is possible that a cohort effect is seen - that men from earlier generations have lower relative height, weight, or

Total White Asian

P