Abstract

Author(s): Kaoru SAKATANI*, Seika KAMOHARA, Kenji KARAKO and Katsunori OYAMA

Recent studies suggest that systemic metabolic disorders may lead to cognitive impairment and potentially increase the risk of dementia onset. Specifically, lifestyle diseases an contribute to vascular cognitive impairment resulting from atherosclerosis. Moreover, other systemic metabolic factors, such as malnutrition, can also influence the risk of cognitive impairment. Based on the relationship between the brain and systemic metabolism, we utilized AI technology to assess the risk of dementia using health check-up data including basic blood data that did not include dementia-specific biomarkers. The AI model was trained to estimate cognitive function, represented by the Mini-Mental State Examination (MMSE) from basic blood test data and age. The trained AI model was able to estimate MMSE scores with high accuracy using blood test data and subject’s ages as inputs. Although the estimation accuracy slightly decreased when age was excluded from the input features, a significant correlation with the actual measurements was still observed. The most important variables in the estimation were subject’s ages, followed by blood parameters indicating malnutrition, anemia, renal function, and liver function. These variables showed significant correlations with MMSE. Such findings suggest that the pathology of dementia should not be limited to the brain alone, but rather should be considered as a systemic metabolic disorder. By considering dementia in this way, it becomes possible to evaluate the risk of dementia using health check-ups. Furthermore, it becomes possible to identify metabolic disorders underlying an individual’s dementia risk based on abnormalities in blood data and provide personalized dietary interventions accordingly. Currently, dietary therapies aimed at preventing cognitive impairment are mainly uniform diets intended for the prevention of lifestyle diseases, such as the Mediterranean diet. However, combining individualized dietary therapies tailored to metabolic disorders contributing to an individual’s dementia risk may offer a more effective approach to dementia prevention.