Dietary Soybeans Intake and Bone Mineral Density among 995 Middle-Aged Women in Yokohama

To investigate relationship of dietary factors, especially source of calcium intake, to bone mineral density (BMD) among Japanese middle-aged women, a total of 995 healthy women age of 40 to 49 (mean•}SD, 45•}3), who lives in Yokohama-city, were recruited through convenience sampling by the municipal information paper and health announcement at each 18 public health center in 18 wards for the three-day course on prevention of osteoporosis from October 1996 to March 1998. The BMD of the 2nd metacarpal bone was measured using Computed X-ray Densitometry (CXD) method, by a trained radiologist. Dietary intake of calcium was assessed by self-reporting food frequency questionnaire on calcium dietary sources such as milk, dairy

The bone mass loss since menopause start is a major determinant of the bone mass later in life and is thus considered to be an important relevant factor to the risk of osteoporotic frac-tures1).Therefore, the risk of osteoporosis might be greatly depends on BMD loss in menopause, which is likely to be influenced by multiple environmental factors including dietary factors, such as calcium intake2), physical activity, and smoking, etc.There are many reports on bone mineral and dietary factors, however a few studies have been conducted on relationship between beneficial effects of vitamin-K, soyprotein, and isoflavonoid and bone mass among middle aged women in Japan, which might differ from those among Caucasians.We carried out the cross-sectional study on dietary factors and BMD among 995 Japanese middle-aged women, aged 40-49, residing in Yokohama, urban area of Japan.Subjects were not following any special dietary regimens or not taking any calcium supplement.

BMD measurements
The BMD of the 2nd metacarpal bone was measured using Computed X-ray Densitometry (CXD) method (Bonalizer, Teizin Co.Tokyo), by a trained radiologist at each public health center.

Dietary assessment
Each dietary calcium intake was assessed by self-reporting food frequency questionnaire on major calcium dietary sources such as milk, dairy products, small fish, vegetables, and soybeans and carefully checked by trained dieticians at each public health center.
On this questionnaire, portion size for each major calcium dietary sources were listed for participants and calcium intake from each sources were calculated by the same dieticians to estimate weekly calcium intake.Concretely, portion size of several foods for 100mg of calcium, for example 200m1 of milk, were listed, and participants were requested to fill in the amount of portion taken for one week for each sources of calcium (Table 1).This procedure was developed through several committees meetings on dietary assessment in the study, in which all dieticians and investigators were involved.

Statistical analysis
Correlation coefficients between BMD and age, height, weight, and weekly calcium intake were calculated using Peason correlation for continuous variables.The relationships between non-adjusted and adjusted mean BMD controlled for age, height, body weight, weekly calcium intake and dietary intake frequency were assessed by analysis of variance (ANOVA) and analysis of covariance (ANCOVA), respectively.All analyses were conducted by using SPSS for Windows ver.7.5.

RESULTS
The average age, height, body weight, weekly calcium intake, and BMD are summarized in Table 2.
Correlation between BMD and age, height, body weight, and weekly calcium intake are indicated in Table 3.In 40 to 49 years, BMD was positively related with height (r=0.125,p=0.000008)and negatively related with age (r=-0.136,p==0.00004)significantly.Weight and weekly calcium intake had no apparent relation with BMD.
Adjusted mean of BMD controlled for age, height, weight, and weekly calcium intake by frequency of consumption of dietary sources of calcium are summarized in Table 4.A significant independent gradient of BMD through only for soybeans (p=0.03) was noted.BMD of those who have soybeans more than 2 times per week were significantly higher than those who have 0 or I times.
Table 5 indicates the relationship between BMD and menstrual cycle of participants.A significant independent negative Table 1.Frequency questionnaire for calcium intake to selected foods.Q.How often do you eat following foods?Please fill in following columns.

DISCUSSION
The bone mass of middle aged women decreases rapidly since the menopause period start.The population of the elderly has been expanding rapidly in Japan.The incidence of osteoporosis and the number of elderly patients suffering from bone fractures are increasing drastically 3).Then the prevention of bone mineral loss during menopause may be effective preventive method for osteoporosis and fracture later in life 1).
Osteoporosis is defined as a systemic skeletal disease characterized by low bone mass and microarchitectural deterioration of bone tissue, with a resulting in an increase of fracture.And many reports revealed the close relationship between BMD and osteoporosis.Therefore BMD is thought to be better indicator to predict the risk of bone fracture of later in life.
Nowadays various apparatus for measuring BMD have been developed.
In this study, Computed X-ray Densitometry (CXD) method was applied in order to measure BMD of many cases simultaneously assessing bone cortical width and its BMD of the 2nd metacarpal bone.This method has the advantages of simple operation and high reproducibility.With regard to validation, CXD method are well correlated with those by other methods including dual energy X-ray absorptiometry (DXA) 4,5,6) There was no apparent relation between weekly total calci-um intake and BMD (Table3).There are some reports that calcium is a threshold nutrient which could explain inconsistency among population based epidemiological studies 7,8).With regard to dietary assessment, the validity and repeatability of our self-reporting food frequency questionnaire in this study were not investigated.Therefore we cannot deny the possibility that the quality of this questionnaire might obscure the true relationship between calcium intake and BMD.
We investigated the relationships of dietary factors, especially dietary source of calcium intake, to BMD among healthy Japanese middle aged women in Yokohama, urban area of Japan.The results suggested that the BMD might be more essentially affected by variety of dietary source of calcium and possible dietary beneficial factors than total amount of calcium intake.
The study subjects were recruited not through random sampling but convenience sampling at each 18 public health center.This might reduce the power to infer these results to ordinary population, however the distribution of BMD among the study subjects were similar to those reported as normal reference data for women 40 to 49 years of age in the Japanese population 9).
There was negative significant correlation between age and BMD, which suggests peak bone mass may be younger than 40 years of age.In another cross sectional study about Japanese women reported peak bone mass was 40 years of age 10).
There was positive significant correlation of height to BMD .Since height and age was also significantly negatively correlat-  * adjusted for age , height, weight, and weekly calcium intake by ANCOVA method ** p-value by ANOVA *** p-value by ANCOVA ed (r=-0.093,p=0.0001),positive significant correlation of height to BMD might appear reflecting the association of age to BMD.
There was no apparent relation between weekly total calcium intake and BMD.However a significant positive gradient of BMD through intake frequency of soybeans (p=0.02) was noted.This tendency remained after an adjustment was made controlling for age, height, weight, and weekly calcium intake (p=0.03).
These findings imply the possibility that soybeans might affect BMD of middle-aged women.Soybeans are rich with vitamin K, soy protein, and isoflavonoid, which might affect beneficially on BMD.
Previously, we demonstrated the relationship between dietary calcium and BMD among premenopause women aged 20 to 40 in Yokohama 11).Small fish accompanied by higher salt intake associated with BMD adversely.On the contrary, there was no significant relationship between soybean intake and BMD that is suggested in the present study among middle aged women aged 40 to 49.
The relationship between dietary factors and BMD might be affected by age especially in terms of menopause.
Bonjour JP, et al. reported that vitamin K deficiency might contribute to bone fragility in-patients sustaining hip fracture 12).Kanai T, et al. investigated the relationship between BMD and vitamin K in postmenopausal women, and reported that women with reduced BMD showed lower levels of vitamin K1 and K2 than those with normal BMD 13) The influence of vitamin-K2 (menakinone) on bone metabolism is thought as follows.MK-4 is one of 14 vitamin-K isomers, which have almost same biological function as MK-7 rich in soybeans.In patient with established osteoporosis, Orimo et al. reported that treatment with MK-4 (45mg of menatetrenone daily) for 48weeks resulted in an increase in metacarpal BMD, increased serum osteocalcin, and reduced urinary calcium excretion 14,15) A hypothetical mechanism to explain these findings would be the presence of a renal vitamin K-dependent protein whose activity decreased if vitamin K is not sufficient, thus leading to excessive calcium loss in the urine.This would lead to secondary hyperparathyroidism, increased bone turnover, and ultimately to increased urinary hydroxyproline excretion that would be corrected by vitamin K supplementation.These preliminary human studies suggest that pharmacological doses of MK-4 may have an anti-resorptive effect on bone, but substantial further investigation in this area is required.Supplementation with MK-4 has also been reported to have a positive effect in vitro bone mineralization system 16).Because soybeans are rich of vitamin K2, there was a significant trend for higher BMD with more frequent intake of soybeans.These findings suggest that soybeans may causatively increase BMD.
The hypercalciuric effect of protein has been proposed as one factor contributing to the high rates of osteoporosis in Western countries, where protein intake greatly exceeds requirement 17).However, soy protein causes much less Ca excretion in the urine as a result of the somewhat lower sulfur amino acid, which increase Ca excretion from kidney 18).The soy protein is rich of isoflabonoid(phytoestrogen), which is known for the beneficial effect on cardiovascular and breast cancer risk factors in females 19.20).
Soy protein may also help to promote bone health 21) Yamaguchi M, et al. investigated the effect of genistein, which is one of isoflavonoids on bone resorption in vitro, and showed its direct inhibitory effect on bone resorption in tissue culture 22).
The effect of isoflabonoid to gonadotropin in postmenopausal and that in premenopausal seems to be different.In the postmenopausal women, isoflabonoid shows antiestrogenic effect.On the other hand, in the premenopausal women, isoflabonoid shows estrogenic effect 23,24).
This study was carried out by cross-sectional method, with no longitudinal data for the change of BMD and dietary factors.In order to investigate the causative relationship between BMD and dietary factors, comprehensive longitudinal study should be required.
Beyond this limitation, this study shows significant decrease of BMD with menstrual cycle, and significant positive relationship between BMD and frequency of soybeans intake.Bone mineralization effect of vitamin-K , weak hypercalciuric effect of soy protein, and phytoestrogen effect of isoflabonoid might be related with significant positive association between BMD and soybeans intake.
Considering the natural history of osteoporosis from middle age, BMD decreases by age, and decreases drastically since menopause start.It is important to prevent the BMD decreasing by healthy lifestyle including appropriate diet.
In this paper, we investigated relationship between BMD and dietary sources of calcium intake in middle-aged women .Soybeans intake significantly positively associated with BMD which suggest possible causative beneficial factors , such as vitamin-K, soyprotein, and isoflabonoid rich in soybeans could be important, apart from total calcium intake , for prevention of osteoporosis.

Sample 1 :
If you take 200cc of milk 2 times daily, please fill in 2 times per day and 7 days per week.Sample 2: If you take 400cc of milk 1 times daily, please fill in 2 times per day and 7 days per week.

Table 2 .
Study participants: anthropometric, weekly calcium intake, and bone mineral density.
*r : correlation coefficient gradient of adjusted BMD controlled for age, height, weight, and weekly calcium intake through menstrual cycle (p=0.001)was noted.

Table 4 .
Dietary sources of calcium intake frequency and bone mineral density (n=995).

Table 5 .
menstrual cycle and bone mineral density.