Depressive symptom trajectories around incident hip fracture in older adults: an event-centered matched analysis from the Health and Retirement Study

OBJECTIVE: To examine depressive symptom trajectories around incident hip fracture and determine whether excess burden reflects pre-fracture worsening, peri-fracture change, or post-fracture trajectory differences. METHODS: We used waves 3-13 (1996-2016) of the Health and Retirement Study. Eligible participants were aged 65 years or more, free of baseline hip fracture, and had repeated depressive symptom assessments. Incident hip fracture was defined as the first positive report after a prior negative report. Depressive symptoms were measured using the 8-item Center for Epidemiologic Studies Depression Scale. Cases were matched to controls using coarsened exact matching. Piecewise linear mixed-effects models estimated group differences in pre-event slope, interval-level change around the fracture-reporting interval, and post-event slope. RESULTS: The matched sample included 5098 participants (mean age 75.0 years; 60.3% recorded as female), including 552 incident hip fracture cases and 4546 controls. The main fracture-related difference was an excess increase in depressive symptoms around the fracture-reporting interval (coefficient = 0.249, 95% confidence interval 0.076 to 0.423). This increase was not preceded by a differential pre-event slope (coefficient = 0.001, 95% confidence interval - 0.015 to 0.017), and was followed by a flatter post-event slope (coefficient = -0.041, 95% confidence interval - 0.080 to -0.002). CONCLUSION: Incident hip fracture was associated mainly with an acute worsening in depressive symptoms around the fracture-reporting interval rather than excess pre-fracture deterioration or sustained divergence over longer follow-up. The peri-fracture period may be an important window for psychological assessment and support.