Modelling Diabetes Risk: Social and Behavioural Factors in BRFSS Data
This project analyses data from the 2024 Behavioral Risk Factor Surveillance System (BRFSS), a large U.S. health survey of over 450,000 adults, to understand how demographic, socioeconomic, and behavioural factors are associated with diabetes risk.
The goal is to identify which factors show the strongest relationships with a diabetes diagnosis, and how those patterns change after accounting for differences across the population.
The project is organised into three parts:
Overview
A concise summary of the key findings, with visualisations showing how diabetes risk varies across age, body weight, and income.
Exploratory Data Analysis
A breakdown of the sample, variable distributions, and unadjusted relationships with diabetes status.
Logistic Regression
Full model specification, diagnostics, and adjusted results, including odds ratios, model performance, and assumption checks.