Potential Role of Logistic Regression Analysis to Identify Significant Risk Factors Associated with Stroke

Objectives: This research paper aims to clarify and analyse the various risk factors contributing to the occurrence of stroke in a specific population. Material and methods: This study employed a cross-sectional analysis of the 2015 Behavioral Risk Factor Surveillance System (BRFSS) dataset. The BRFSS is an annual telephone-based survey system designed to gather information about behavioural risk factors among adults across the United States. The dataset used in this study consisted of 70,692 observations obtained from the 2015 BRFSS. It included information on 21 potential risk factors and a binary outcome variable indicating the presence or absence of a stroke. The data analysis was conducted using Google Colab, a cloud-based platform that supports the programming language Python and its libraries. Results: The logistic regression analysis revealed that the strongest associations with stroke were observed for heart disease or heart attack (p < 0.001), high blood pressure (p < 0.001), high cholesterol (p < 0.001) and difficulties in walking (p < 0.001). Other risk factors that showed significant associations with stroke were diabetes, smoking, fruit consumption, vegetable consumption, general health perception, mental health, physical health, age, education and income. It is important to note that some risk factors, including cholesterol check, physical activity, access to healthcare and absence of doctor visits, did not exhibit statistically significant associations with stroke. Conclusion: The findings revealed that heart disease or heart attack, high blood pressure, high cholesterol and difficulties in walking exhibited the strongest associations with stroke.