Related Experiment Video
Updated: Jun 14, 2025
07:31
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
7.0K
Metabolic syndrome predictive modelling in Bangladesh applying machine learning approach
Md Farhad Hossain1,2, Shaheed Hossain2, Mst Nira Akter2
1Division of Computing, Analytics and Mathematics, Department of Mathematics and Statistics, School of Science and Engineering, University of Missouri, Kansas City, MO, United States of America.
Plos One
|September 5, 2024
Summary
Metabolic syndrome (MetS) affects 27.8% of the Bangladeshi population, with higher prevalence in females. Key risk factors identified include age, blood pressure, and obesity indicators, informing predictive modeling.
Area of Science:
- Cardiovascular Health
- Metabolic Disorders
- Data Science in Healthcare
Background:
- Metabolic syndrome (MetS) is a critical public health concern, characterized by a cluster of risk factors including abdominal obesity, hypertension, and hyperglycemia.
- These factors significantly elevate the risk of cardiovascular diseases, stroke, and type 2 diabetes.
- Understanding MetS prevalence and its determinants in specific populations is crucial for targeted interventions.
Purpose of the Study:
- To identify key risk factors associated with Metabolic syndrome in the Bangladeshi population.
- To develop and evaluate machine learning (ML) models for predicting MetS.
- To assess the prevalence of MetS using established diagnostic criteria.
Main Methods:
- Utilized the Adult Treatment Panel III (ATP III) criteria to diagnose MetS in a dataset of 8185 Bangladeshi participants.
- Employed Chi-Square and Random Forest techniques to identify significant risk factors.
- Trained and evaluated various ML models including Decision Trees, Random Forests, SVM, XGBoost, KNN, and Logistic Regression.
Main Results:
- Identified a MetS prevalence of 27.8% in the study population, with a higher proportion among females (58.3%) than males (41.7%).
- Key predictive factors identified include Age, Systolic Blood Pressure (SBP), Waist-to-Height Ratio (WHtR), Fasting Blood Glucose (FBG), Waist Circumference (WC), Diastolic Blood Pressure (DBP), marital status, Hip Circumference (HC), Triglycerides (TGs), and smoking.
- The Waist-to-Height Ratio (WHtR) was specifically highlighted as an anthropometric index for diagnosing abdominal obesity.
Conclusions:
- The study successfully identified significant risk factors for Metabolic syndrome in Bangladesh and demonstrated the potential of ML models for prediction.
- Further research is needed to refine the precision of these classification tools and enhance predictive accuracy for MetS.
- The findings underscore the importance of addressing modifiable risk factors like obesity, hypertension, and hyperglycemia within the Bangladeshi population.
