Objective: To extend healthy life expectancy, the prevention of fractures due to reduced bone density is important because they are a major cause of elderly people becoming bedridden. However, at our center, only about 2% of males and 25% of females underwent bone density examinations. In order to encourage people to have these examinations, we studied association between arteriosclerosis-related items and bone density, and determined the high-risk items.
Methods: Our subjects were 321 women who underwent bone density examinations for the first time and internal abdominal fat measurements by CT in health check-ups between April 1 2008 and January 31 2014. Using statistical methods, we studied possible association between examination date, lifestyle habit items and decreased bone density.
Results: The t test revealed significant differences in the following items, with
p<0.001 for age at examination and hypertension;
p<0.01 for HbA1c value and
p<0.05 for visceral fat, BMI and drinking. Also, the χ
2 test revealed significant differences in the following items, with
p<0.001 for menopause and hypertension and
p<0.05 for visceral fat and BMI. Further, in a logistic regression analysis model with age at examination, subcutaneous fat area, HbA1c value, total cholesterol and smoking as explanatory variables, the accuracy rate was 83.4%, and in the Hosmer-Lemeshow test, the value of χ
2 was 8.060 and that of
p was 0.428 (greater than significance level of 0.05). This indicated that results were optimal and demonstrated the usefulness of these variables in predicting reduction in bone density.
Conclusion: While an association with smoking was not immediately obvious in the χ
2 test and t test, its addition to the explanatory variables for logistic regression analysis had a major impact on predicting osteoporosis. While there is a tendency to focus on association with single item, we think it is necessary to consider association between multiple items and reflect this in lifestyle guidance.
View full abstract