Risk factors for osteoporosis related to their outcome: fractures.

The aim of the study was to determine to what extent easy obtainable bone mineral density (BMD)-related risk factors are associated with the occurrence of fractures and to what extent changes in these determinants during a patient's lifetime are relevant. A cross-sectional population-based study was carried out on 4725 postmenopausal women, 50-80 years of age, registered with 23 general practitioners (GPs). The women were questioned and examined. BMD of the lumbar spine was measured using dual-energy X-ray absorptiometry (QDR-1000, Hologic). We analyzed the total population as well as a random sample of 1155 women for whom additional data were collected on recalled weight at age 20-30 years and on self-reported height. Body mass index (BMI) was estimated in two ways: (2) objective BMI [= measured weight/(measured height)2]; (2) recalled BMI [= recalled body weight at age 20-30/(self-reported height)2]. Fractures (dependent variable) were categorized as: (1) fractures sustained during the patient's lifetime; (2) fractures after the age of 50 years; (3) fractures that had occurred during the 5 years before BMD measurement took place. Multivariate stepwise backward and forward logistic regression analyses, using fractures as the dependent variable, were performed with all discrete and non-discrete variables (divided into quartiles). The relationship between the presence of osteoporosis and the presence of fractures was related to the changes in BMI (recalled BMI versus objective BMI). More advanced age, positive family history of fractures and BMD had a positive association with the presence of fractures. Low recalled BMI was a statistically significant predictor of 'fractures during the patient's lifetime' and of 'fractures after the age of 50'. Hysterectomy was associated with a higher prevalence of 'fractures during the patient's lifetime'. Perimenopausal complaints in the history seemed to be associated with a lower prevalence of 'fractures after the age of 50'. Moderate (and heavy) occupational exercise in the past were associated with the presence of fractures 'after the age of 50' and 'fractures during the past 5 years'. Sporting activities in the past showed a slightly positive relationship with the presence of 'fractures during the patient's lifetime' and 'fractures after the age of 50'. Bivariate analysis revealed that current smokers had not sustained significantly more fractures than current nonsmokers, but within the subgroup of current smokers, the prevalence of fractures was significantly higher among those women who had smoked for more than 35 years. Smoking was statistically significantly associated with early menopause. Early menopause was not statistically significantly related to the presence of osteoporosis but appeared to be statistically significantly associated with the prevalence of fractures in the age categories over 65 years. The absolute risks of sustaining one or more fractures ranged from 3% to 44%. Women in the lowest quartile of recalled and objective BMI were often osteoporotic (40%). In this category, women with normal BMD had a statistically significant lower fracture risk than osteoporotic women. Women with a possibly decreased BMI were most often osteoporotic and had sustained more 'fractures during the past 5 years' than expected. Women who had (probably) always been obese were less often osteoporotic and had a much lower fracture risk. It is concluded that decreased BMI is associated with a higher risk of developing fractures at an older age. Prevention of fractures should include fall prevention. In addition, in lean women treatment of low BMD is important.