Pulmonary Function Trajectories Preceding Death Among Older Adults: A Long-Term Community-Based Cohort Study

BACKGROUND: Poor pulmonary function (PF) has been linked to mortality, but the timing of PF changes before death remains unclear. We aimed to examine the association between PF and mortality and identify different PF trajectories precedes death. METHODS: Within the Rush Memory and Aging Project, 1 438 participants without chronic obstructive pulmonary disease were followed for up to 22 years. PF was assessed annually using a composite score (tertiled as low, medium, and high) based on forced vital capacity (FVC), forced expiratory volume in 1s (FEV1), and peak expiratory flow (PEF). Survival status was observed during the follow-up period. Data were analyzed using Cox regression, Laplace regression, and mixed-effect models. RESULTS: During the follow-up, 737 (51.25%) participants died. Compared to high PF, the hazard ratio (95% confidence interval [CI]) of mortality was 1.35 (1.05, 1.72)/1.63 (1.25, 2.12) for medium/low PF. The median survival time (95% CI) was shortened by 0.80 (0.01–1.61)/1.72 (0.43–3.01) years for participants with medium/low PF, compared to high PF. In multiadjusted trajectory analysis, the significant differences between decedents and survivors occurred at 7 years before death for composite PF (mean difference [95% CI]: 0.14 [0.02–0.25]), 6 years for FEV1 (0.21 [0.08–0.33]) and FVC (0.21 [0.08–0.34]), and 8 years for PEF (0.21 [0.06–0.37]), and became greater thereafter. CONCLUSION: Poor PF is associated with elevated mortality and shortens survival for nearly 2 years. An acceleration in PF decline tends to occur 7 years before death. Poor PF, together with its decline, might be a predictor of mortality among community-dwelling older adults.