A statistical analysis and generalized linear models of cerebral stroke

Cerebral stroke, a life-threatening condition that has a high mortality and morbidity rate, is the second leading cause of death worldwide. A stroke occurs when the blood supply to the brain is interrupted or reduced, resulting in potential neurological damage. Unlike many previous studies that focused on a single personal attribute related to stroke or estimated the probability of stroke, we conducted statistical analyses to investigate whether and how stroke and other variables are influenced together and amongst each other. Next, we modeled stroke, hypertension, and heart disease based on the data from 43,400 patients. We hypothesized that stroke, hypertension, and heart disease are statistically correlated to age, body mass index (BMI), glucose level, work type, marriage status, and gender, but are not related to residence type. Descriptive statistics of the data were computed and examined with statistical tests. To model the data, we performed logistic and linear regression. Our results showed that all of the variables in the dataset were related to each other except for residence type. Our data also suggested that heart disease had the strongest association with whether an individual had stroke, while smoking status had the strongest association with hypertension or heart disease. The analysis and models provide a context of risk factors with respect to cerebral stroke.