[Skip to Navigation]
July 22, 1996

Mortality Following Fractures in Older Women: The Study of Osteoporotic Fractures

Author Affiliations

From the Department of Epidemiology and Biostatistics (Drs Browner, Nevitt, and Cummings and Ms Pressman), the General Internal Medicine Section, Department of Veterans Affairs Medical Center (Dr Browner), and the Division of General Internal Medicine (Dr Cummings), University of California—San Francisco.

Arch Intern Med. 1996;156(14):1521-1525. doi:10.1001/archinte.1996.00440130053006

Background:  Most investigators have hypothesized that the increase in mortality following osteoporotic fractures reflects poor underlying health status in addition to the acute effects of the fracture.

Methods:  We observed 9704 ambulatory women aged 65 years or older enrolled in the Study of Osteoporotic Fractures. We obtained reports of fractures and deaths every 4 months and reviewed death certificates and hospital discharge summaries. Multivariable proportional hazards models were used to determine the association between fractures and age-adjusted mortality.

Results:  During a mean follow-up of 5.9 years, 1737 women had nonspine fractures, with a postfracture mortality rate of 3 per 100 woman-years, compared with 1.8 per 100 woman-years in those who did not have fractures (P<.001). After adjusting for other factors associated with mortality, women (n=361) with fractures of the hip or pelvis had a 2.4-fold (95% confidence interval, 1.7-3.3) increase in mortality. However, only 9 (14%) of the 64 deaths that occurred after hip or pelvic fractures were caused or hastened by the fracture. By contrast, 11 (17%) of these deaths seemed to have been a result of chronic conditions that had contributed to the hip or pelvic fracture, and 44 (69%) of the deaths were not clearly related to the fracture.

Conclusions:  Mortality is increased following several types of fractures in older women. Most of the increase following hip and pelvic fractures is due to underlying conditions and probably would not be affected by reductions in the incidence of these fractures.Arch Intern Med. 1996;156:1521-1525