Background: Gastrointestinal (GI) cancers and their associated treatments may adversely affect nutritional status, increasing the risk of malnutrition and physical deterioration. In this context, body composition (BC) assessment using bioelectrical impedance analysis (BIA) represents a rapid and cost-effective tool for obtaining clinically relevant information on nutritional and metabolic status, with the potential to enhance multidisciplinary patient management. However, the clinical implications of different body composition profiles in patients with GI cancers remain insufficiently explored.
Objectives: This study aims to evaluate the clinical relevance of BIA-derived body composition profiles for the early identification of nutritional impairment and sarcopenia risk in patients with newly diagnosed gastrointestinal cancers, thereby facilitating timely and targeted interventions and strengthening multidisciplinary care pathways.
Methods: Data were obtained from the NUTRISCREEN project, launched by the Istituto Nazionale Tumori IRCCS Fondazione G. Pascale of Naples. The study included patients diagnosed with gastric, pancreatic, rectal, and liver cancers. Risks of malnutrition and sarcopenia were assessed using the Nutritional Risk Screening 2002 (NRS-2002) and the SARC-F questionnaire, respectively. Nutritional assessment included body composition parameters derived from BIA. Descriptive statistics summarized categorical and continuous variables. Principal component analysis (PCA) and cluster analysis were performed to identify distinct body composition profiles combining BIA-derived parameters. Multivariable logistic regression models adjusted for age, sex, cancer site, and education level assessed associations between body composition profiles and nutritional risk.
Results: Data from 189 patients (mean age 66.0 ± 11.0 years) were analysed. Overall, 69 patients (36.5%) were at risk of malnutrition (NRS-2002 score ≥3), and 61 (32.3%) at risk of sarcopenia (SARC-F score ≥4). PCA identified two main components, and k-means clustering classified patients into two body composition profiles: a High-Quality Muscle Profile (HMP; n = 99) and a Low-Quality Muscle Profile (LMP; n = 90). Compared with HMP, LMP included a higher proportion of female patients (63.3% vs 9.1%, p < 0.001) and a greater prevalence of malnutrition (47.8% vs 27.3%, p = 0.005) and sarcopenia risk (44.4% vs 21.2%, p = 0.002). The LMP profile showed significantly lower values of phase angle (PhA: 5.2 ± 0.9 vs 6.3 ± 1.2), body cellular mass index (BCMI: 8.8 ± 1.2 vs 11.7 ± 1.5), and a higher extracellular-to-intracellular water ratio (ECW:ICW: 1.0 ± 0.2 vs 0.8 ± 0.2). In multivariable analyses, LMP was independently associated with increased risk of malnutrition (OR = 3.52, p = 0.008) and sarcopenia (OR = 3.32, p = 0.011).
Conclusions/Implications for practice or policy: The Low-Quality Muscle Profile (LMP) cluster was associated with poorer nutritional status and unfavorable body composition parameters compared with the High-Quality Muscle Profile (HMP). Integrating body composition profiling into clinical practice may improve characterization of nutritional and muscle status, providing a non-invasive, reproducible, and clinically interpretable approach to support risk stratification, individualized nutritional interventions, and longitudinal monitoring in oncology. These findings underscore the value of early, targeted nutritional strategies integrated with anticancer therapies to preserve muscle health and improve health-related quality of life in patients with gastrointestinal cancers.