
In older people with heart failure and atrial fibrillation, undernutrition is one of the key factors leading to inflammation, loss of function, disability and, ultimately, death. Since the inflammation is closely related to the intestinal microbiome, shaping the gut microbiota composition with a probiotic-based food could be an efficacious and safe approach to improve the cognitive functioning and skeletal muscle mass. AMBROSIA (Microbiota-Inflammation-Brain axis in heart failure: new food, biomarkerS and AI Approach for the prevention of undeRnutrition in Older) aims to develop an innovative food product to prevent undernutrition in HF and AF patients: a new chocolate bar containing a specific mix of probiotic strains and a cocktail of micro/macronutrients. The efficacy of the AMBROSIA bar on undernutrition prevention and its impact on cognitive functioning and skeletal muscle mass of older HF and AF patients will be evaluated through a prospective monocentric interventional clinical study. Several experimental high-throughput datasets such as biomarker, lipidomics and metagenomics profiles from urine, blood and saliva samples will be generated and analyzed during the AMBROSIA study. Using these data, statistical and machine learning methods will be developed for the identification of features from the “Microbiota-Inflammation-Brain axis” that are predictive for undernutrition (biomarkers) and/or related to AMBROSIA bar treatment outcome. The AMBROSIA network consists of an international consortium with academic and industrial partners from five European countries (Italy, Spain, UK, Ireland, Germany). As an AMBROSIA partner, Genevention will develop a semantic data integration platform and knowledgebase for FAIR management and rich, harmonized annotation of clinical and experimental high-throughput data. We will develop, evaluate and integrate machine learning methods for identification of features from the “Microbiota-Inflammation-Brain axis” that are predictive for undernutrition and could be used as biomarkers. Statistical and machine learning tools will be developed and integrated in the platform for analysis of patient parameters related to treatment with the AMBROSIA bar. This work is financially supported by the German Federal Ministry of Food and Agriculture (BMEL) through the Federal Office for Agriculture and Food (BLE), grant number 2820ERA20E, (Cofund ERA-NET “ERA-HDHL”).