Related Experiment Videos
Identifying risk factors in a mostly overweight patient population with coronary artery disease
Saim Yoloğlu1, Alpay Turan Sezgin, Ramazan Ozdemir
1Department of Biostatistics, Faculty of Medicine, Inonu University, Malatya, Turkey. syologlu@inonu.edu.tr
Insights
Obesity is a modifiable risk factor for coronary artery disease (CAD). Multivariate analysis effectively identified CAD predictors in an overweight population, offering higher accuracy than univariate methods.
Area of Science:
- Cardiology
- Public Health
- Epidemiology
Background:
- Overweight and obesity are recognized as complex, multifactorial chronic disorders.
- The American Heart Association (AHA) identifies obesity as a modifiable risk factor for coronary heart disease (CAD).
- Understanding CAD risk factors in overweight populations is crucial for targeted interventions.
Purpose of the Study:
- To evaluate the association between CAD and conventional/novel risk factors in a predominantly overweight patient population (MOP).
- To compare the predictive performance of univariate and multivariate logistic regression models for CAD.
- To determine the optimal logistic regression model using the Hosmer-Lemeshow goodness-of-fit test.
Main Methods:
- Univariate and multivariate logistic regression analyses were employed.
- Systematic application of regression algorithms to assess risk factors.
- The Hosmer-Lemeshow statistic was used to evaluate model fit.
Main Results:
- Univariate analysis revealed significant associations between CAD in MOP and various risk factors, but with low sensitivity, specificity, and accuracy.
- Multivariate analysis demonstrated a good model fit with high sensitivity, specificity, and accuracy, despite some factors not being significant predictors.
- The Hosmer-Lemeshow test confirmed the superior performance of the multivariate model.
Conclusions:
- Multivariate logistic regression provides a more accurate and reliable model for identifying CAD risk factors in overweight individuals compared to univariate analysis.
- The findings underscore the importance of comprehensive risk factor assessment using advanced statistical methods in managing CAD in obese populations.
- Effective identification of CAD predictors can inform public health strategies and clinical practice for disease prevention.
Abstract:
Overweight/obesity is a complex multifactorial chronic disorder, and the American Heart Association (AHA) has recently classified as a modifiable risk factor for coronary heart disease (CAD). This study (1) evaluates the association between CAD in a patient population mostly overweight (MOP) and conventional and novel coronary risk factors by using univariate and multivariate logistic regression analysis and (2) seeks to find the best model by comparing univariate and multivariate logistic regression analysis algorithms, which were systematically applied to risk factors by using Hosmer-Lemeshow statistic test. In univariate analysis, there were significant associations between CAD in MOP and conventional and novel risk factors. However, the model's sensitivity, specificity, and accuracy levels were weak. In multivariate analysis, although some risk factors were not found as predictors of coronary artery disease, the model showed good fit to data and had high sensitivity, specificity, and accuracy levels. This was also confirmed by using the Hosmer-Lemeshow goodness of fit test, more specifically.