Incidence of fractures in women in the post-menopause: a cohort study in primary care in southern Brazil.

Abstract

Few studies are evaluating the incidence of fractures in Latin America and Brazil. This study aimed to estimate the incidence of bone fractures in postmenopausal women seen in primary care and evaluate the FRAX algorithm's performance in these women. A cohort study was carried out in the municipality of Santa Maria, Southern Brazil. Postmenopausal women aged 55years and over who attended primary health care were included. The recruitment period was from March 1 to August 31, 2013, and the participants were followed for 5years. The fracture risk was calculated using the FRAX algorithm. The reported incident fractures were confirmed by imaging studies or surgical reports. Of the 1057 women recruited for the study, 854 were followed. They contributed to 2732 person/year. The mean follow-up time was 3.2years (SD 1.05). The incidences of total fractures, major fractures, and hip fractures were 22.3, 15.0, and 3.3 per 1000 person/year. The most frequent fracture sites were the wrist, shoulder, and ribs. The fracture predictors were rheumatoid arthritis, previous fracture, and the use of glucocorticoids. The discriminatory capacity of incident fractures calculated by FRAX without the inclusion of BMD was AUC 0.730 (95% CI 0.570, 0.890) for hip fracture and AUC 0.691 (95% CI 0.598, 0.784) for major fractures. The FRAX algorithm showed an adequate discriminatory capacity to identify incident fractures in primary care in our study. The incidence of fractures found in our study appears to be lower than that reported in North America and Europe.