predicting metabolic syndrome based on nutrient intakes in iranian women using a decision tree data-mining approach

Background and aims: the increasing incidence of metabolic syndrome (mets) has become a major public health concern globally. nutrients and dietary patterns are influential factors associated with the incidence of mets. the main purpose of this study was to apply machine learning approaches to predict mets based on micronutrients and macronutrients intakes in adult females from mashhad, northeast of iran. method: this cross-sectional study was carried out on 2975 women, 35-65 years old, who participated in the mashad cohort study. mets was defined according to the international diabetes federation (idf). dietary intakes were measured using a 65-items food frequency questionnaire. logistic regression (lr) and decision tree (dt) algorithms examined the associations between micro/macronutrients intakes and the risk of mets. results: according to the lr model, calcium, phosphate, potassium, vitamin b12, thiamine, selenium, magnesium, and sodium were significantly related micronutrients associated with an increased prevalence of mets. fiber was the only macronutrients associated with mets. according to the dt model, in micronutrients, magnesium was the most related factor related to the risk of mets, followed by phosphate, potassium, sodium, and selenium. fiber was the most important macronutrient associated with mets.