Factors Relating to the Bone Mineral Density in the Elderly Living in the Urban Community

A population-based cross-sectional study of the association between bone mineral density (BMD) and several physical factors was conducted on the Japanese community elderly in 1991. The study comprised 387 participants living in the urban community of Tokyo. BMD was measured on the third lumbar vertebra with a lateral projection by dual energy X-ray absorptiometry fixed in a van (mobile DXA). The mean BMD declined with aging in both males and females except at age-group 80-84 in males. Males had significantly higher BMD in average than females in each age-group. In univariate analysis, such many physical factors as height, weight, body mass index (BMI), grip strength, walking speed in maximum walking, serum albumin, serum intact parathyroid hormone (intact-PTH), masticatory ability and age showed significant association with BMD. Multiple regression analysis revealed that age (inversely) and BMI (directly) had significant association with BMD in both sexes. In addition, serum albumin in males and height, intact-PTH and masticatory ability in females remained as significant contributors to BMD. J Epidemiol, 1994; 4 : 83-89.

It is well known that the bone mineral contents gradually decline as a general tendency with advancing age after the peak bone mass appearing in early third decade of one' s life.Particularly, the loss of bone mineral in the elderly is a major risk of subsequent osteoporotic fractures which impair their activities of daily living and may keep earlier them bed-ridden condition.In accordance with the increasing proportion of the elderly in the population, basic information of bone mineral among the healthy elderly is crucial to establish the ways of preventing osteoporotic fractures.
Various noninvasive techniques have been developed for the quantitative assessment of bone mineral density (BMD) during the last decade.It is now well accepted that dualenergy X-ray absorptiometry (DXA) is the most reliable method for measurement of BMD, especially focusing on such critical sites of osteoporotic fractures as the lumbar

1) Overview of the research project
The Tokyo Metropolitan Institute of Gerontology launched a new project, the "Longitudinal Interdisciplinary Study on Aging (TMIG-LISA)".The project aims at verifying the etiology of geriatric diseases and chronic conditions, at identifying factors relating to the aging process, and at clarifying factors modulating the effects of life events.The project consists of 3 major disciplines ; medical, psychological, and social sciences').
The baseline survey of medical science in TMIG-LISA has been carried out on the elderly living in the Koganei City.The city lies about 25 km west of the center of the metropolis (Fig. 1) and has a population of 103,672 in an area of 16 km2 in 1991.Over the past 30 years, it has been developed as a residential area for commuters into the center of Tokyo and for retirees.Koganei City is an area which has long life expectancies at birth and at 65 years, compared with other areas in Tokyo.Economic status of the residents is also high because they are mainly engaged in professional and managerial jobs.The proportion of population aged 65 years and over has gradually increased and reached 9.74% in 1991.Koganei City has been chosen by the institute (TMIG) as a study area of the urban community for search into health aging for more than fifteen years.The outline of these survey and studies has already described elsewhere4,5).

2) Subjects
Subjects were 996 persons (438 males and 558 females) ranging in age from 65 to 84 who were chosen by one-tenth of random sample from all eligible citizens living in Koganei City at June 1991.A comprehensive study on aging and the aged has been carried out from August through October in 1991.Table 1. Number of participants for DXA examina-tion.
Among those eligible persons, 814 (81.7%) responded to the interview survey.However, in spite of enthusiastic recommendation to undergo an medical examination and every endeavor including free transport service, 405 (40.7%) persons could participate in the successive medical examinations.Regarding the demographic differences between the non-participants and the participants to the medical examinations including BMD measurements, sex-ratio was almost the same between two groups (82.7 percent vs 82.3 percent).Furthermore, such items as age-distribution, self-rated health, education, past histories of stroke, heart disease, hypertension and diabetes, and hospitalization for the last one year were investigated from the interview.No significant difference was found in males except the frequency of the past history of heart disease.In females, the participants were significantly (p<0.05)younger than the non-participants (71.1*5.Oyears vs 72.7*5.4years).
The difference was mainly due to lower percentage of old persons aged 80 and more in the participants than in the non-participants (7.7% vs 14.2%, respectively).It thus appeared that the participants would not be completely representative of the elderly population in Koganei City because of existence of a few demographic differences between the participants and the non-participant.
Bone mineral measurements were undertaken for 387 persons (387/405 ; 95.5%) with other medical examinations after informed consent.The age-group and the number of the participants were shown in Table 1.These subjects were ambulatory and healthy without any serious diseases affecting bone mineral content.
3) Methods Medical, psychological and sociological variables were investigated at the baseline survey for this longitudinal and interdisciplinary study.
Concerning bone mineral contents in the body, BMD was measured by DXA (Hologic QDR-1000W by Hologic Inc., MA, UAS) fixed in the van running from one community to another in Koganei City.The inside of the van was divided into three sections, i.e., operation-section with computer and display system, examination-section for measurements of BMD, and waiting-section for the participants to change their clothes.The subjects were laid on the instrument (table mattress) in left side-down position for lateral projection.BMD in the third lumbar vertebra (L3) with lateral projection was taken instead of the usual measurements of BMD from the second (L2) to the fourth (L4) lumber vertebrae with antero-posterior (AP) projection which may often overestimate BMD because of contamination by the calcification of abdominal aorta and spondylosis deformans of the vertebral body in the elderly.In the lateral projection, BMD in L3 is more reliable than L2 or L, which may often show overlapping with either ribs (L2) or iliac crest (L4).In order to eliminate secondary hyperostotic changes or senile ectopic calcification in and around vertebrae, the analysis of BMD was carried out by setting the range of interest in the center of vertebral body where the trabecular bones are predominant.The reproducibility represented by coefficient of variation of this DXA instrument and analytical method for computing BMD have already reported in detail by the authors6).
The other following variables were chosen as relevant factor for changes of bone mineral contents in the present study : Body weight (kg) was measured to the nearest 100 g and body height (cm) was to the nearest mm.The Quetelet's index (kg/cm2xl04) was adopted as a measure of body mass index (BMI).
Grip strength (kg) was to measured on the stronger hand using a hand dynamometer.
Walking speed at the fastest walking (m/sec) was measured to the average time in complete two trials on a middle 5 m length during flat 15 m walking course.
As blood components relating to BMD, the serum albumin and intact parathyroid hormone were measured in non-fasting blood samples obtained by venopuncture.Serum albumin (Alb : g/dl) was measured by standard kit with BCG method.Serum intact parathyroid hormone (intact-PTH : pg/ml) was measured using the Allegro intact PTH kit by RIA method.
Masticatory ability was evaluated by geriatric dentists using jelly (G-l Jelly(R) during oral and dental examination.The texture of jelly was divided into three grades from grade-1 (soft) to 3 (hard) according to the hardness of daily foods7).
Statistical analysis was conducted on the SPSS-X software on VAX/VMS of the TIMG mainframe computer.Multiple linear regression analysis was done between BMD as dependent variables and the other physical variables as independent variables.RESULTS 1) Mean and standard deviation (S D.) of the variables by sex.Table 2 shows mean and S.D. of the continuous variables at the baseline examination.A significant sex difference was found in level of all variables except age ; six variables (BMD, weight, height, grip strength, walking speed and masticatory ability) showed greater average in males than those in females, and other three variables (Quetelet's index, serum albumin and serum intact-PTH) showed smaller average in males.Table 3 indicates mean and S.D. of BMD in L3 with lateral projection on each age-group.The mean BMD declined with aging in a graded fashion in both sexes except age-group 80-84 in males.A marked sex difference was found in every agegroups,i.e., males had higher level of BMD in every agegroup.The regression coefficient between BMD and age of males was -0.008, and that of females was -0.007.There were significant differences of mean BMD between age-group 65-69 and other age-groups in both sexes.In fact, our study on the precision using the spine phantom bone showed higher coefficient of variation (CV) in lateral-BMD than that in AP-BMD (1.04% in lateral vs 0.31% in AP).However, in spite of the relatively poor precision , only lateral spinal projection in L3 was adopted for BMD measurement in the present study, partly because the time and cost allocated for BMD measurment was limitted in the medical examination, and partly because the lateral spinal projection is useful in the elderly who have more or less osteophytes and sclerotic facets in the vertebrae as well as abnormal calcification in the aortic artery which could certainly elevate AP-BMD.
According to the previous studies, the following factors are suspected of having principle influences more or less on bone mineral condition i.e., race11,12) , sex 13), age 14,15), and physique or body size [16][17][18].In the present study, these principle variables could be also available for the analysis.In addition to sex, age, weight, height and BMI, the following variables were adopted in this study ; walking speed at the fastest walking which is believed to depend on the individual kinetic ability including maximum torque in the knee extension during the walking19), serum (intact) parathyroid hormone as an indicator of bone metabolism, serum albumin representative as well-nourished status and a possible predictor of survivorship of the elderly20,21), and masticatory ability which is relating to general health condition of the elderly including physical activity22).
The serum calcium, phosphate and alkaline-phosphates, in spite of the good indicators of many bone disorders, were not considered in this study because of their relative constancy in senile osteoporosis23).
The uivariate analysis revealed that the all but two (height and intact-PTH in males) had significant relations with BMD in L3.In the multivariate analysis, the relationship of age (negatively) and BMI (positively) in both sexes remained significant, which coincide completely with preceding findings from a number of cross-sectional studies concerning the risk factors for bone mineral condition 14,15,18) Serum albumin in males and height, serum intact-PTH and masticatory ability in females showed also independent contributions to BMD.Serum albumin concentration has been thought as not only a possible predictor of mortality20) but also an index of aging, since the ageassociated decline of this serum protein is considered to be mainly attributable to intrinsic changes in albumin metabolism with aging21).Presumably, senile osteoporosis with gradual decrease of bone minerals in males observed usually at about age of seventy and over may coincide with the gradual decline of serum albumin concentration.
In contrast with serum albumin influencing to BMD in male, serum intact-PTH, one of endogenous bone-metabolic hormone, showed negative association to BMD in female.It is known that PTH becomes relatively important on the bone metabolism in the postmenopausal women whose estrogen is exhausted24).According to a clinical data from healthy postmenopausal Japanese women, the same result has been obtained in relation between PTH and BMD with negative and significant correlation coefficient (-0.742)23).In generally, among the elderly aged 65 and more, these results may indicate that nutritional and exogenous factors are predominant to BMD in males and, on the other side, endogenous hormones such as estrogen and PTH have primary and definit influences to BMD in females.
Little has been known on the association between masticatory ability and bone mineral contents in the body either in clinical studies or in population-based studies.Mastication, as is well-known, consists of voluntary and complicated movements of the muscles attached to the mandible.The teeth rigidly suspended in the sockets of the jaw bones (maxilla and mandible) plays also important role in mastication.Deterioration of masticatory ability, mainly results from the loss of the natural dentition, leads to functional disturbance of the muscles, salivary glands and total alimentary system, which results necessarily in the malnutritional status.In addition, there is a significantly positive relationship between masticatory ability and physical strength such as BMI and length of time for standing on one leg with eyes open in the elderly from preceding study22).In this context, as well as physical exercise for BMD25-27), good masticatory ability may have a positive effect on BMD as shown in the present study.
The present study identified several physical factors relating to BMD.Further research is warranted to examine relationship between other variables such as lifestyle and BMD from the standpoint of preventing osteoporosis in the elderly.

Table 2 .
Means (*S.D.) of BMD and other physical variables for both sexes.