The risk of fracture among women starting selective serotonin reuptake inhibitors

Abstract Increased fracture risk has been reported in patients using selective serotonin reuptake inhibitors (SSRIs). However, prior studies have had limited information regarding BMD and symptoms of depression, both potentially important confounders. We examined a longitudinal cohort of women who initiated SSRIs, other antidepressant (AD) medications, or no AD to estimate the risk of fracture associated with start of SSRIs. The Study of Women’s Health Across the Nation (SWAN) is a longitudinal cohort of diverse women from across the US transitioning across the menopause. Study visits are approximately yearly, with reporting of medication use, fracture incidence (any and non-traumatic), mental health scales (CES-D), and BMD (the latter occurring in selected SWAN sites). We estimated fracture incidence and relative risk among women starting SSRIs or other AD, and compared them with women not starting SSRIs or other AD. Multivariable Cox regression models with increasing adjustment were constructed. As well, secondary analyses focused on non-traumatic fractures and women with BMD measurements. The Study of Women’s Health Across the Nation includes 3302 total women, of which 286 were excluded because of prevalent AD use and 1167 because they did not have adequate follow-up time, had a fracture prior to start of an AD, or could not be matched; this left 1849 women for analysis. The incidence rates for any fracture (per 100 person-years) for SSRI users was 2.64 (95% CI: 1.82-3.71), other AD users 0.80 (95% CI: 0.22-2.04), and non-users 1.21 (95% CI: 1.07-1.36). Fully adjusted regression models found an increased hazard ratio for any fracture among women starting SSRIs compared with no AD (HR 1.77, 95% CI: 1.15-2.74). These results were consistent for non-traumatic fractures and in subgroups with BMD included as a covariate. Initiation of SSRI among women in mid-life was associated with an increased risk of fracture.