Association of Back Pain with All-Cause and Cause-Specific Mortality Among Older Women: a Cohort Study.

Abstract

The impact of back pain on disability in older women is well-understood, but the influence of back pain on mortality is unclear. To examine whether back pain was associated with all-cause and cause-specific mortality in older women and mediation of this association by disability. Prospective cohort study. The Study of Osteoporotic Fractures. Women aged 65 or older. Our primary outcome, time to death, was assessed using all-cause and cause-specific adjusted Cox models. We used a four-category back pain exposure (no back pain, non-persistent, infrequent persistent, or frequent persistent back pain) that combined back pain frequency and persistence across baseline (1986-1988) and first follow-up (1989-1990) interviews. Disability measures (limitations of instrumental activities of daily living [IADL], slow chair stand time, and slow walking speed) from 1991 were considered a priori potential mediators. Of 8321 women (mean age 71.5, SD = 5.1), 4975 (56%) died over a median follow-up of 14.1years. A higher proportion of women with frequent persistent back pain died (65.8%) than those with no back pain (53.5%). In the fully adjusted model, women with frequent persistent back pain had higher hazard of all-cause (hazard ratio [HR] = 1.24 [95% CI, 1.11-1.39]), cardiovascular (HR = 1.34 [CI, 1.12-1.62]), and cancer (HR = 1.33, [CI 1.03-1.71]) mortality. No association with mortality was observed for other back pain categories. In mediation analyses, IADL limitations explained 47% of the effect of persistent frequent back pain on all-cause mortality, slow chair stand time, and walking speed, explained 27% and 24% (all significant, p < 0.001), respectively. Only white women were included. Frequent persistent back pain was associated with increased mortality in older women. Much of this association was mediated by disability.