TY  - JOUR
AU  - Hosseini-Esfahani, Firoozeh
AU  - Alafchi, Behnaz
AU  - Cheraghi, Zahra
AU  - Doosti-Irani, Amin
AU  - Mirmiran, Parvin
AU  - Khalili, Davood
AU  - Azizi, Fereidoun
PY  - 2021
DA  - 2021/9/2
TI  - Using Machine Learning Techniques to Predict Factors Contributing to the Incidence of Metabolic Syndrome in Tehran: Cohort Study
JO  - JMIR Public Health Surveill
SP  - e27304
VL  - 7
IS  - 9
KW  - metabolic syndrome
KW  - Tehran Lipid and Glucose Study
KW  - data mining
AB  - Background: Metabolic syndrome (MetS), a major contributor to cardiovascular disease and diabetes, is considered to be among the most common public health problems worldwide. Objective: We aimed to identify and rank the most important nutritional and nonnutritional factors contributing to the development of MetS using a data-mining method. Methods: This prospective study was performed on 3048 adults (aged ≥20 years) who participated in the fifth follow-up examination of the Tehran Lipid and Glucose Study, who were followed for 3 years. MetS was defined according to the modified definition of the National Cholesterol Education Program/Adult Treatment Panel III. The importance of variables was obtained by the training set using the random forest model for determining factors with the greatest contribution to developing MetS. Results: Among the 3048 participants, 701 (22.9%) developed MetS during the study period. The mean age of the participants was 44.3 years (SD 11.8). The total incidence rate of MetS was 229.9 (95% CI 278.6-322.9) per 1000 person-years and the mean follow-up time was 40.5 months (SD 7.3). The incidence of MetS was significantly (P<.001) higher in men than in women (27% vs 20%). Those affected by MetS were older, married, had diabetes, with lower levels of education, and had a higher BMI (P<.001). The percentage of hospitalized patients was higher among those with MetS than among healthy people, although this difference was only statistically significant in women (P=.02). Based on the variable importance and multiple logistic regression analyses, the most important determinants of MetS were identified as history of diabetes (odds ratio [OR] 6.3, 95% CI 3.9-10.2, P<.001), BMI (OR 1.2, 95% CI 1.0-1.2, P<.001), age (OR 1.0, 95% CI 1.0-1.03, P<.001), female gender (OR 0.5, 95% CI 0.38-0.63, P<.001), and dietary monounsaturated fatty acid (OR 0.97, 95% CI 0.94-0.99, P=.04). Conclusions: Based on our findings, the incidence rate of MetS was significantly higher in men than in women in Tehran. The most important determinants of MetS were history of diabetes, high BMI, older age, male gender, and low dietary monounsaturated fatty acid intake. 
SN  - 2369-2960
UR  - https://publichealth.jmir.org/2021/9/e27304
UR  - https://doi.org/10.2196/27304
UR  - http://www.ncbi.nlm.nih.gov/pubmed/34473070
DO  - 10.2196/27304
ID  - info:doi/10.2196/27304
ER  -