Machine learning elucidates associations between oral microbiota and the decline of sweet taste perception during aging.

Aging-induced deterioration in taste perception can result in loss of appetite and malnutrition in the elderly, posing a substantial challenge to healthy aging. In oral cavity, the oral microbiota, food particles, and taste receptors interact extensively under the flow of saliva. Although it has been hypothesized that oral microbiota may influence taste perception, evidence remains limited. Here we justified this hypothesis and further proposed that specific oral bacterial genera exhibited significant associations with age-related alterations in sweet taste perception. Notable age-related changes in taste perception were observed: the elderly presented significantly higher detection and recognition thresholds for sweet taste acuity compared to the youth. Linking back to the oral microbiota, we identified key bacteria genera Haemophilus, Lachnoanaerobaculum, Fusobacterium, Aggregatibacter and Oribacterium associated with sweet taste perception via machine learning. Correspondingly, we found several volatile compounds in the oral exhaled breath, especially the endogenous compound isoprene, that significantly correlated with oral bacteria genera and sweet taste sensitivity. Our findings in sweet taste perception-associated bacteria and metabolites can be potential biomarkers of early aging, which provides timely fresh clues for the well-being of the aging population.