AUTHOR=Gutiérrez-Esparza Guadalupe , Pulido Tomas , Martínez-García Mireya , Ramírez-delReal Tania , Groves-Miralrio Lucero E. , Márquez-Murillo Manlio F. , Amezcua-Guerra Luis M. , Vargas-Alarcón Gilberto , Hernández-Lemus Enrique 

TITLE=A machine learning approach to personalized predictors of dyslipidemia: a cohort study

JOURNAL=Frontiers in Public Health

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2023.1213926

DOI=10.3389/fpubh.2023.1213926

ISSN=2296-2565

ABSTRACT=Mexico ranks second in the global prevalence of obesity in the adult population, which increases the probability of developing dyslipidemia. Dyslipidemia is closely related to cardiovascular diseases, which are the leading cause of death in the country. Therefore, developing tools that facilitate the prediction of dyslipidemias is essential for prevention and early treatment. In this study, we used a dataset from a Mexico City cohort of 2621 participants, men and women between 20 and 50 years old, with and without some type of dyslipidemia. The main goal of this study is to identify the potential factors associated with dyslipidemia types in both men and women. This will be achieved by applying machine learning algorithms. To this end, we used VIM of RF, XGBoost, and GBM for feature selection. Similarly, SMOTE was used as a resampling method to balance the dataset, which contains anthropometric measurements, biochemical tests, dietary intake, family health history, and other health parameters related to smoking habits, alcohol consumption, quality of sleep, and physical activity. The results showed that the VIM algorithm of RF got the best subset of attributes followed by GBM, reaching up to 80% of balanced accuracy. The best subset of attributes is selected by implementing the machine learning technique comparing the better performance (as measured with balanced accuracy, sensitivity, and specificity) between the classifiers. Body mass index, high uric acid levels, age, sleep disorders, and anxiety are the top five features that increase the risk of several types of dyslipidemia.