Paraspinous muscle sarcopenic indices for predicting mortality in patients undergoing hemodialysis

We examined the association between computed tomography (CT)-derived paraspinous muscle density (PSMD) and index (PSMI) and all-cause mortality in patients undergoing hemodialysis.

We included 234 hemodialysis patients and used CT images to measure PSMD and PSMI at the third lumbar vertebra level. PSMD and PSMI were determined as the average CT-based value of the paraspinous muscle in Hounsfield units (HUs) and as the paraspinous muscle area adjusted by height (cm²/m²), respectively. We used univariate and multivariate regression analysis to investigate the associations of PSMD and PSMI with baseline variables. Moreover, we used the Kaplan-Meier methods, Cox regression analysis, and prediction model discrimination to examine the relationships of these paraspinous sarcopenic indices with mortality risk.

The PSMD and PSMI were independently associated with the C-reactive protein level and geriatric nutritional risk index, respectively. The receiver operating curve-derived sex-specific cutoffs of PSMD and PSMI for predicting mortality were 32.1 HU and 13.29 cm²/m² in women, and 38.4 HU and 14.97 cm²/m² in men, respectively. Eighty-one patients died during a median follow-up period of 4.0 years. Lower PSMD alone was independently associated with an increased risk of all-cause mortality (adjusted hazard ratio [aHR]: 3.14, 95% confidence interval [CI]: 1.81-5.46). In addition, only when PSMD was added to the baseline risk model, the net reclassification improvement significantly improved (0.516, p = 0.0001).

PSMD may be better than PSMI for predicting all-cause mortality risk in patients undergoing hemodialysis.