Computed tomography‐derived muscle quality and density and the risk of postoperative decline in exercise capacity in adults who received cardiac surgery: A retrospective cohort study

BACKGROUND: Cardiovascular surgery significantly improves outcomes, yet many, particularly older or frail individuals, face substantial postoperative declines in exercise capacity, prolonging recovery and worsening long-term prognosis. Early, precise risk identification is critical. Routinely performed preoperative computed tomography (CT) offers a pragmatic opportunity to assess muscle quality like psoas muscle density (PMD), reflecting physiological resilience. This study aimed to examine the predictive capabilities of CT-derived muscle mass and quality for postoperative exercise capacity decline following elective cardiovascular surgery, focusing on identifying high-risk patients for timely prehabilitation. METHODS: This retrospective study enrolled 229 consecutive patients undergoing elective thoracic cardiovascular surgery (May 2020 to March 2025). Skeletal muscle index (SMI) and PMD were calculated from preoperative CT images. The primary outcome was a postoperative exercise capacity decline, defined as a ≥50 m reduction in 6-min walk distance on postoperative Day 14. RESULTS: Eighty-seven patients (38%) experienced postoperative exercise capacity decline. PMD was significantly lower in the "6-min walk distance decline ≥50 m" (p < 0.001). Receiver operating characteristic analysis for PMD revealed an area under the curve of 0.693, with an optimal cutoff of 45.0 Hounsfield units. Lower PMD independently associated with higher odds of postoperative decline (odds ratio, 1.08 per 1-Hounsfield unit decrease, 95% CI: 1.03-1.13). SMI was not a significant predictor. CONCLUSION: Reduced preoperative PMD is a robust, independent predictor of early postoperative exercise capacity decline. The 45.0 Hounsfield units cutoff provides a clinically relevant threshold. Utilizing routinely available CT for muscle quality assessment offers a practical, cost-effective method to identify high-risk patients.