Time-Varying Social Integration and Long-Term All-Cause and Cause-Specific Mortality After Traumatic Spinal Cord...

To determine whether lower time-updated social integration and within-person decline are associated with subsequent mortality after traumatic spinal cord injury (SCI), and to examine cause-specific and educational gradients.

Longitudinal secondary cohort analysis using annually lagged exposure, a 5-year validity window, and discrete-time survival models with participant-clustered standard errors.

Spinal Cord Injury Model Systems 2021 public-use database.

Adults with traumatic SCI and at least one eligible Craig Handicap Assessment and Reporting Technique (CHART) Social Integration assessment. The primary cohort included 16,783 participants; the repeated-assessment model included 6,610 participants, entering follow-up after their second assessment.

Not applicable.

All-cause mortality was primary. Natural-cause, external-cause, and suicide mortality were secondary; suicide was a subset of external causes. CHART and education were exposures or covariates.

During 122,626 exposure-covered person-years, 2,965 deaths occurred. Crude mortality increased from 17.3 per 1,000 person-years at CHART 100 to 46.8 at CHART 0-49. Each 10-point lower current score was associated with higher mortality (hazard ratio [HR], 1.09; 95% confidence interval [CI], 1.07-1.10). A lower first score (HR, 1.078; 95% CI, 1.049-1.108) and within-person decline (HR, 1.077; 95% CI, 1.051-1.104) were associated with mortality per 10 points. Associations were observed for natural (HR, 1.08) and external mortality (HR, 1.11); the suicide estimate was imprecise (HR, 1.05; 95% CI, 0.96-1.15). Lower education corresponded to higher standardized 10-year mortality risk. Estimates were similar when the first assessment occurred within two years of injury. Primary cause was unavailable for 23.2% of deaths, and early postinjury mortality was incompletely represented in follow-up.

Lower and declining social integration were associated with higher subsequent mortality after traumatic SCI. Repeated assessment may provide prognostic information, but survivor selection, missing causes, and residual confounding limit interpretation. These findings do not establish that modifying social integration or education changes survival.