Abstract
BACKGROUND: Sodium intake is undoubtedly essential for normal bodily function, but it is an important public health concern when intake exceeds dietary requirements. With salt intake exceeding recommended levels in almost all countries, high salt intake has become an important global health topic. This study aimed to be the first to combine epidemiological big data with animal behavioral experiments to systematically investigate the association between dietary sodium intake and cognitive function. METHODS: Based on NHANES data from 2011 to 2014, the association between dietary sodium intake and cognitive impairment was assessed using multivariate linear regression and five types of machine learning, including random forest and XGBoost, in 2367 participants aged 60-80 years. Cognitive function was further assessed by a 6-month high-salt dietary intervention in a C57BL/6 mouse model combined with behavioral tests such as the Morris water maze and Open Field Test (OFT). RESULTS: The results of the study revealed that the average daily sodium intake of the 2,367 middle-aged and older adults included reached 4,502 mg, with 91% exceeding the recommended standard. Dietary sodium intake was positively correlated with cognitive function scores (p < 0.001). The random forest model had the best predictive efficacy (AUC = 0.918), which was significantly better than that of the SVM and neural networks. Animal behavioral experiments revealed that high-salt dietary exposure significantly impaired neurobehavioral function in a dose-dependent manner, leading to spatial memory deficits and anxious behavior in mice. CONCLUSION: There was a threshold-dependent association between dietary sodium intake and cognitive function, and the dose-response trend was consistent between human observational studies and animal studies. This study provides translational medical evidence for the public health strategy of "salt reduction for dementia prevention.