Research Article: Evaluating objective nutritional biomarkers for the prediction of sarcopenia in Crohn’s disease: development and internal validation of a clinical nomogram
Abstract:
Sarcopenia is prevalent in Crohn’s disease (CD) and correlates with adverse clinical outcomes; however, routine computed tomography (CT) screening is limited in daily practice. This study aimed to comparatively analyze three objective nutritional indices—the Geriatric Nutritional Risk Index (GNRI), Advanced Lung Cancer Inflammation Index (ALI), and Prognostic Nutritional Index (PNI)—and to develop a predictive nomogram for sarcopenia in CD.
A retrospective cohort of 333 patients with CD was randomly allocated into a training ( n =?234) and an internal validation ( n =?99) cohort. Sarcopenia was defined using gender-specific cut-offs of the CT-derived skeletal muscle index (SMI) at the third lumbar vertebra (L3) level. Unadjusted restricted cubic splines (RCS) evaluated continuous associations. Multivariable logistic regression was utilized to identify associated risk factors and construct the nomogram. Model performance was assessed via the area under the receiver operating characteristic curve (AUC), calibration plots, and decision curve analysis (DCA). Interaction tests evaluated prognostic stability.
The overall prevalence of sarcopenia was 48.3%. All three indices demonstrated linear inverse continuous associations with sarcopenia risk and maintained robust statistical associations after multivariable adjustment. The GNRI demonstrated comparable discriminative accuracy to the ALI, significantly outperformed the PNI, and remained highly stable across all clinical subgroups (all P for interaction > 0.05). Conversely, the predictive capacities of the ALI and PNI were significantly confounded by disease activity. The formulated nomogram, integrating the GNRI with sex, Montreal location, Montreal behavior, and disease activity, exhibited excellent discrimination (training AUC?=?0.847; validation AUC?=?0.750). Calibration plots indicated optimal agreement, and DCA confirmed maximum net clinical benefit for the GNRI-based model.
The GNRI is a robust surrogate biomarker for identifying sarcopenia in CD, demonstrating excellent predictive performance comparable to the ALI, alongside high clinical stability. The developed GNRI-based nomogram provides an accurate, non-invasive screening tool to facilitate early identification and guide targeted nutritional interventions.
Introduction:
Sarcopenia is prevalent in Crohn’s disease (CD) and correlates with adverse clinical outcomes; however, routine computed tomography (CT) screening is limited in daily practice. This study aimed to comparatively analyze three objective nutritional indices—the Geriatric Nutritional Risk Index (GNRI), Advanced Lung Cancer Inflammation Index (ALI), and Prognostic Nutritional Index (PNI)—and to develop a predictive nomogram for sarcopenia in CD.
Read more