Optimizing Nutritional Care with Machine Learning: Identifying Sarcopenia Risk through Body Composition Parameters in Cancer Patients – Insights from the NUTritional and sarcopenia RIsk SCREENing project (NUTRISCREEN)

Background/objectives: Cancer and related treatments can impair body composition (BC), increasing the risk of malnutrition and sarcopenia, which can worsen prognosis and Health-related Quality of Life (HrQoL). Bioelectrical Impedance Analysis (BIA) is a non-invasive method used to assess BC. To enhance BC parameter interpretation, this study aimed to develop a predictive model based on a Machine Learning (ML) approach for sarcopenia risk in cancer patients at the Istituto Nazionale Tumori IRCCS "Fondazione G. Pascale" (Naples).Methods: In this study, sarcopenia and malnutrition risks were assessed using validate questionnaires, anthropometric measurements and BIA. HrQoL was evaluated with the EORTC QLQ-C30 questionnaire. A PCA clustering analysis was performed to classify patients according to BC parameters and identify different BC profiles.Results Data from 879 cancer patients were collected. 117 patients (13%) and were at risk of malnutrition, while 128 (15%) were at risk of sarcopenia. PCA analysis identified three main components. Patients in C3 were older and with a higher prevalence of comorbidities. Moreover, C3 had a higher risk of malnutrition and sarcopenia. At multivariable analysis, age, lung cancer site, diabetes, malnutrition risk and C3 were significantly associated with an increased risk of sarcopenia.  Conclusion: The NUTRISCREEN project provides a personalized nutritional pathway for early screening of malnutrition andsarcopenia. Using a ML approach, we provide distinct BC profiles and valuable insights into the factors associated with sarcopenia risk. This approach in clinical practice could help define risk categories, ensure the most appropriate nutritional and therapeutic strategies and improve patient outcomes by providing data-driven care.