Logistic Regression Analysis on 250k+ patients to identify non-linear risk factors in Type 2 Diabetes
This project analyzes a real-world dataset of 253.680 patients to determine the physiological and lifestyle factors contributing to Type 2 Diabetes.
The analysis challenges the assuption of linear risk, applying Logistic Regression to quantify how the risk escalates exponentially with BMI categories.
- Non-Linear Risk
- Exponential grow of risk
- Intervention protocols should prioritize patients before they cross the obesity threshold, patients knowing this could improve adherence to treatment as they could get more benefit from taking action.
- Language: R
- Libraries: 'tidyverse', 'ggplot2'
- Technique: Binary Logistic Regression ('glm'), Odds Ratio Calculation, Data Visualization
Author: Miguel Ángel Vicente | Health Data Analyst & Physiotherapist