Nutritional intake and health status of populations and the relationship between diet and oral ulcers: A cross-sectional study based on NHANES data and machine learning predictions.

Oral ulcers are a common oral disease. This study aims to use data from National Health and Nutrition Examination Survey to analyze the related risk factors of oral ulcers and explore health status differences among various populations. The data were derived from National Health and Nutrition Examination Survey, covering various aspects of the health, lifestyle, nutritional status of the U.S. civilian population. Three thousand one hundred twenty-six participants were included in the final analysis, divided into 2 groups: with and without oral ulcers. A standardized questionnaire was used to collect information on age, gender, race, family income, body mass index, diabetes, hyperlipidemia, heart disease, smoking, alcohol consumption. The diagnosis of oral ulcers was based on participants' self-reported questionnaire results. Statistical analysis was conducted using Statistical Package for the Social Sciences software, including descriptive statistics, Spearman correlation analysis, multiple linear regression, confusion matrix analysis, forest plot analysis, restricted cubic spline regression. Significant differences were found between participants with and without oral ulcers in terms of age, gender, family income, hyperlipidemia, depression, smoking, and alcohol consumption. Age, gender, family income, magnesium, and sodium were important factors related to the incidence of oral ulcers. The model's accuracy was approximately 72.48%, precision was about 58.26%, recall was about 61.15%, and the F1 score was about 59.57%. The area under the receiver operating characteristic curve was 0.77, indicating that the classifier has a good discriminative ability. There was a significant association between age and the increased risk of oral ulcers, with the risk significantly decreasing with age. Smoking and hypertension had a significant impact on the prediction of oral ulcers, with the model tending to predict the occurrence of oral ulcers in cases with higher levels of smoking and hypertension. Age, gender, family income, hyperlipidemia, depression, smoking, and alcohol consumption are important risk factors for oral ulcers. The model has good predictive ability overall but still has room for improvement in predicting the presence of oral ulcers. There is a significant association between age and the increased risk of oral ulcers, with the risk significantly decreasing with age. Smoking and hypertension have a significant impact on the prediction of oral ulcers.