Prevalent vertebral deformities predict mortality and hospitalization in older women with low bone mass

OBJECTIVES: To determine the relationship between prevalent vertebral defor mities and the risk of mortality and hospitalization in older women with lo w bone mass. DESIGN: A prospective cohort study. SETTTNG: Eleven clinical centers in the United States. PARTICIPANTS: A total of 6459 community-dwelling women with low bone mass a ged 55 to 81 participated in the Fracture Intervention Trial (FIT), a multi center clinical trial of alendronate that enrolled women into one of two st udy arms based solely on the presence or absence of existing radiographic v ertebral deformities. There were 2027 women with at least one vertebral def ormity enrolled in the vertebral fracture arm of FIT and followed prospecti vely for an average of 2.9 years, whereas 4432 women with no vertebral defo rmity were enrolled in the clinical fracture arm of FIT and followed prospe ctively for an average of 4.2 years. MEASUREMENTS: Determination of prevalent vertebral deformities on baseline lateral thoracic and lumbar spine radiographs was made at the coordinating center using a combination of radiographic morphometry by digitization and semiquantitative radiologic interpretation. Deaths were confirmed by obtain ing copies of original death certificates of all participants who died. Epi sodes of hospitalization were captured through adverse event reporting; hos pitalizations resulting solely from adverse events containing the words " f racture " or " trauma " were excluded from the analyses. RESULTS: During the follow-up period, 122 women died, and 1676 women were h ospitalized on at least one occasion for reasons not related solely to frac ture. Compared with women without prevalent vertebral deformities, those wo men with prevalent deformities had higher risks of mortality (age- and trea tment assignment-adjusted relative risk 1.60, 95% confidence interval (CI), 1.10-2.32) and hospitalization (age- and treatment assignment-adjusted rel ative risk 1.18, 95% CI, 1.06-1.31); In addition, further adjustment for ot her factors, including smoking status, physical activity, hypertension, cor onary heart disease, obstructive lung disease, any fracture since the age o f 50, health status, total hip BMD, and body mass index did not alter the a ssociation between prevalent vertebral deformities and risk of mortality su bstantially (multivariate relative risk 1.49, 95% CI, 1.05-2.21). Adjustmen t for all these factors and diabetes also did not change the relationship b etween prevalent vertebral deformities and hospitalization (multivariate re lative risk 1.14, 95% CI, 1.02-1.27). Rates of mortality and hospitalizatio n increased with increasing number of prevalent vertebral deformities (test s for trend P < .01). CONCLUSIONS: Prevalent vertebral deformities in older women with low bone m ass are associated with increased risks of mortality and hospitalization. O nly a portion of this increased risk was explained by other known predictor s of these outcomes.