Structural equation modeling of the relationship of bone mineral density and its risk factors in Japanese women

Several factors have been reported as risk factors for the development of osteoporosis. In this study, we aimed to examine the relationship among lifestyle factors, biologic factors, and bone mineral density (BMD) using structural equation modeling (SEM). The subjects in the present study consisted of 866 postmenopausal Japanese women aged between 40 and 80 years old. In the analysis by the SEM, we employed a multiple basic model. As the structural variables, lifestyle factors and biologic factors were selected. The goodness of fit index (GFI) of the final model was 0.991 and the Akaike’s information criteria (AIC) showed the lowest value in the peripheral models. The degree of association between biologic factors and BMD was −0.576 (direct association), 0.012 (indirect association), and −0.564 (total association). With regard to the correlation between lifestyle factors and BMD, the degrees of association were 0.085, −0.084, and 0.001, respectively. This study defined a pilot model for factors influencing BMD. Although is remains necessary to conduct further analyses with more valid measurements and constructs, this model indicated that the correlation between BMD and lifestyle factors was lower than that between BMD and biologic factors.