Highlight
- Lower and declining social integration, measured repeatedly, is strongly associated with increased all-cause mortality after traumatic spinal cord injury (SCI).
- Both initial low levels and within-person decline in social connections predict higher mortality risk, independent of education level.
- Mortality associations are observed across natural and external causes, though suicide-specific effects were imprecise.
- Findings underscore the prognostic value of longitudinal social assessments but caution against inferring causality due to selection bias and confounding.
Study Background
Traumatic spinal cord injury (SCI) is a devastating condition resulting in long-term disability and increased mortality. Survivors often face complex challenges not only from physical impairments but also from psychosocial factors, including social isolation and decreased social integration. Social integration—comprising the quantity and quality of social relationships and participation—is a known determinant of health and mortality in the general population but remains underexplored longitudinally in traumatic SCI cohorts. Given the chronic nature of SCI, understanding how dynamic changes in social connectivity over time impact survival can illuminate targets for holistic rehabilitation and supportive interventions.
Study Design
This analysis leveraged the publicly available 2021 Spinal Cord Injury Model Systems (SCIMS) database, comprising adults with traumatic SCI who underwent serial assessments with the Craig Handicap Assessment and Reporting Technique (CHART) Social Integration scale. The primary cohort included 16,783 individuals with at least one eligible CHART assessment. A subset of 6,610 participants with at least two assessments comprised the repeated-measures cohort to evaluate within-person changes over time.
The study employed a longitudinal secondary cohort design with annually lagged exposure, utilizing a five-year validity window for social integration scores. Discrete-time survival models, accounting for participant clustering, assessed associations of time-updated social integration scores and within-person changes with subsequent mortality outcomes. Mortality was ascertained as all-cause death, with secondary analysis on cause-specific mortality categories: natural causes, external causes (including accidents and violence), and suicide (a subset of external causes). Education level was included as a covariate and effect modifier.
Key Findings
Over 122,626 person-years covered by social integration data, 2,965 deaths occurred. Crude mortality rates showed a clear gradient based on CHART social integration scores: from 17.3 deaths per 1,000 person-years at the highest score (CHART 100) to 46.8 deaths per 1,000 person-years at the lowest social integration levels (CHART 0-49). These findings emphasize the strong inverse relationship between social connectivity and mortality after SCI.
Hazard ratio (HR) estimates for every 10-point decrement in social integration revealed:
– Current (time-updated) social integration score: HR 1.09 (95% confidence interval [CI], 1.07–1.10), indicating a 9% increased mortality risk per 10-point decline.
– First recorded social integration score: HR 1.078 (95% CI, 1.049–1.108).
– Within-person decline in social integration over time: HR 1.077 (95% CI, 1.051–1.104).
Secondary analyses showed similar risk increases for natural-cause mortality (HR 1.08) and external-cause mortality (HR 1.11). However, estimates for suicide-related mortality were imprecise and not statistically significant (HR 1.05; 95% CI, 0.96–1.15).
Lower educational attainment was associated with higher standardized 10-year mortality risk, reinforcing socioeconomic factors’ interplay with social integration in influencing outcomes.
Additional sensitivity analyses restricted to participants whose first social integration evaluation occurred within two years post-injury produced consistent estimates, supporting the robustness of findings early after SCI.
Expert Commentary
These findings substantiate the concept that social determinants, notably social integration, exert a meaningful influence on long-term survival after traumatic SCI. The data corroborate existing evidence linking social isolation with adverse health outcomes and extend it by capturing dynamic social trajectories rather than static baseline measures.
However, several limitations temper causal interpretation. The observational nature of the study leaves residual confounding from unmeasured factors possible. Approximately 23% of deaths lacked confirmed primary cause, which may bias cause-specific analyses. Early post-injury mortality might be underrepresented due to enrollment and assessment timing, potentially skewing mortality risk estimates.
Survivor bias may affect participants who lived long enough to provide repeated assessments—the observed within-person declines in social integration might also reflect worsening health or other conditions leading to increased mortality, rather than social integration being an independent driver. While education correlates with social integration, the study design does not clarify whether enhancing education or social integration can causally improve survival.
Mechanistically, social integration could influence mortality through multiple pathways including emotional support reducing stress, promoting health behaviors, and facilitating access to care—especially critical in the complex long-term management of SCI-related medical and psychological comorbidities.
Conclusion
The longitudinal SCIMS study provides compelling epidemiologic evidence that lower and declining social integration are robustly associated with increased long-term mortality following traumatic spinal cord injury. The graded nature of risk across increments in social integration underscores the vital role of sustained social connectivity in the survivorship of this vulnerable population.
Repeated measurement of social integration may serve as an informative prognostic indicator for clinicians to identify individuals at elevated risk. However, due to methodological limitations, these results do not prove causality, nor do they establish that interventions to improve social integration or education level will necessarily extend survival. Future research should explore targeted social interventions and their potential to modify mortality risk while accounting for confounders and mediating factors.
This study reinforces the importance of holistic, multidisciplinary care approaches addressing psychosocial as well as physical health domains to optimize outcomes for people living with traumatic SCI.
Funding and ClinicalTrials.gov
The study analyzed publicly available data from the Spinal Cord Injury Model Systems, funded by the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR). No additional intervention or funding was reported. ClinicalTrials.gov registration was not applicable as this was a secondary database analysis.
References
1. Calderone A, Marra A, De Luca R, et al. Time-Varying Social Integration and Long-Term All-Cause and Cause-Specific Mortality After Traumatic Spinal Cord Injury: A Longitudinal SCIMS Study. Arch Phys Med Rehabil. 2026 Oct 11. doi: 10.1016/j.apmr.2026.09.004. PMID: 42860197.
2. House JS, Landis KR, Umberson D. Social relationships and health. Science. 1988 Jul 29;241(4865):540-5.
3. Krause JS. Social support, social integration, and health outcomes in persons with traumatic spinal cord injury. Arch Phys Med Rehabil. 1995;76(7):697-701.
4. World Health Organization. International perspectives on spinal cord injury. WHO; 2013.
5. Holt-Lunstad J, Smith TB, Layton JB. Social relationships and mortality risk: a meta-analytic review. PLoS Med. 2010;7(7):e1000316.
