Association of controlling nutritional status with fecal incontinence and fecal incontinence severity index in individuals with spinal cord injury: mediating roles of interleukin-6 and uric acid.

BACKGROUND

Fecal incontinence (FI) is common after spinal cord injury (SCI), but modifiable risk factors are unclear. The association between malnutrition and FI in patients with SCI remains unclear.

METHODS

This cross-sectional study included 2,256 SCI patients. Nutritional status was assessed using the CONUT score (albumin, cholesterol, lymphocytes). FI was self-reported, and the fecal incontinence severity index (FISI) was used to quantify severity. Logistic and linear regressions tested CONUT associations with FI and FISI. Restricted cubic splines examined the dose-response. Exploratory mediation analyses assessed IL-6 and uric acid (UA). The predictive performance of CONUT was evaluated using AUC, NRI, and IDI.

RESULTS

Among 2,256 patients, 267 (11.8%) reported FI. After fully adjusting for covariates, higher CONUT was linked to increased FI odds (OR per 1-unit increase, 1.27; 95%CI, 1.09-1.67) and higher FISI (β = 0.51; 95%CI, 0.32-0.74) in a near-linear dose-response manner. The findings were consistent across subgroups and sensitivity analyses. CONUT showed moderate predictive ability for FI (AUC = 0.788; 95%CI, 0.759-0.818), outperforming the geriatric nutritional risk index (AUC = 0.632) and prognostic nutritional index (AUC = 0.626). Adding CONUT improved reclassification (NRI = 0.19; 95%CI, 0.11-0.30; IDI = 0.12; 95%CI, 0.06-0.20). In exploratory mediation analyses, IL-6 accounted for 16.37 and 18.21% of the total association with FI and FISI, respectively; UA accounted for 14.89 and 10.11%, respectively.

CONCLUSION

CONUT was positively and nearly linear-dose-responsively associated with FI odds and severity, with IL-6 and UA partially accounting for these associations. Among three nutritional indices examined, CONUT demonstrated the strongest association with FI severity. However, these findings reflect relative performance among nutritional markers rather than established clinical predictive utility.