Nonlinear Association Between Subcutaneous Fat Area and All-Cause Mortality in Postoperative Patients with Stage I-III Colorectal Cancer: A Multicenter Retrospective Study.

backgroundColorectal cancer (CRC) represents a significant global health burden as one of the most prevalent malignancies of the gastrointestinal tract. Recent research has increasingly focused on the prognostic implications of host-related factors, particularly body composition. Among these, subcutaneous adipose tissue (SAT) area, a critical biomarker reflecting nutritional status and metabolic homeostasis, has shown inconsistent associations with CRC outcomes, suggesting a potential non-linear relationship. This investigation aims to elucidate the non-linear correlation and identify potential threshold effects between SAT area and overall survival (OS) in postoperative stage I-III CRC patients.

methodsWe conducted a multicenter retrospective cohort study encompassing 3,765 patients diagnosed with stage I-III CRC who underwent radical resection between 2010 and 2020 across three tertiary medical centers. SAT area was quantified at the third lumbar vertebra (L3) level using preoperative computed tomography (CT) imaging. Restricted cubic spline (RCS) regression analysis was employed to evaluate the non-linear association between SAT area and OS, with subsequent threshold determination using piecewise regression modeling. Based on the identified threshold value (45.32 cm2), patients were stratified into low SAT (< 45.32 cm2) and high SAT (≥ 45.32 cm2) cohorts. To address potential confounding factors, propensity score matching (PSM) was implemented for baseline characteristic adjustment, followed by Cox proportional hazards regression analysis to assess the independent prognostic value of SAT area on OS.

resultsIn this multicenter cohort of 3,764 patients with stage I-III CRC (median age 59 years; 53.6% female) undergoing radical resection, RCS analysis revealed a significant nonlinear "L-shaped" association between SAT and OS (nonlinear test P < 0.001). Segmented regression identified an inflection point at 45.32 cm2 (95% CI: 42.15–48.67): below this threshold, each 10 cm2 decrease on SAT was associated with an 18.2% increase in mortality risk (HR = 1.182, 95% CI: 1.104–1.264); above the threshold, there was no significant association between SAT and OS (HR = 1.012, 95% CI: 0.983–1.045). Multivariate Cox regression confirmed that low SAT was an independent risk factor for OS (HR = 1.59, 95% CI: 1.26–1.99, P < 0.001). After PSM, the 5-year OS rate in the low SAT group was significantly lower than that in the high SAT group (77.4% vs. 81.8%, Log-rank P = 0.041), and subgroup analysis showed that this association was stable across different clinical characteristics (heterogeneity test I2 = 0%, P = 0.58).

conclusionThere is a nonlinear "L-shaped" relationship between SAT and postoperative survival in patients with stage I-III CRC, with a threshold of 45.32 cm2. SAT below this threshold is an independent poor prognostic factor, and it is recommended to incorporate it into the preoperative risk assessment system to provide a basis for individualized nutritional intervention and prognosis management.