The leading causes of Ischemic Heart disease Essay

Abstract Hypertension is one of the leading causes of Ischemic Heart disease. The study aims to find out the risk of getting Ischemic heart disease due to stage 2 hypertension. Other factors which contribute to the risk are Cholesterol levels, Age of the person, smoking status ,adrenal gland tumor, sleep apnea,diabetes and high stress levels.For this study, with the outcome variable as Ischemic heart disease,the variable of of interest is hypertension while controlling for variables like cholesterol levels, smoking status, age and adrenal gland tumor. The dataset is from Evans county in Georgia which is a 9 year follow-up study of 609 males at risk of getting a Ischemic Heart disease.Data collected had 71 males who had ischemic heart disease and 255 males who had stage 2 hypertension. The race of the observations in the dataset is white and gender being males leaves the scope for further study to include different races and females which may lead to different outcomes and interpretations.
Methods The goal of the study is a three way comparison between a Linear Probability Model (LPM), a Probit regression model and a Logit regression model. Logistic regression model will help to find out the odds of developing the Ischemic heart disease with hypertension as the exposure variable. The limitations and advantages of each model over the other will be studied and the model with the better fit will be used to predict the probability of developing a Ischemic heart disease with various variables.
Results
(1) Significant t value statistic were obtained for Hypertension, Cholesterol levels, Smoking and Adrenal gland tumor while running regression for Linear probability model and Logit Model. Insignificant t value for cholesterol was found while running the profit regression model. Age was not significant variable in all the three models.
(2) Age divided into 5 groups(i.e less than 40 years, 40-45 years,45-50 years,50-55 years,55-60 years, more than 60 years) had a significant t value statistic for persons above 60 years of age while running the Linear probability model.
(3) 44 observations gave the probability of less than 0/more than 1 while running the LPM. Although LPM is easy to interpret,Logit model was preferred over LPM due to this limitation(More than 5% values gave the probability of less than 0/more than 1).
(4) Running Logit and Logistic regression gave the following results
(a)Hypertension increases the probability for a person having Ischemic heart disease by 0.081. (b)1mg/dl increase in Cholesterol increase the probability of Ischemic Heart disease by 0.000689.
(c)Smoking increases the probability of Ischemic Heart disease by 0.05862.
(d)Adrenal gland tumor increases the probability of Ischemic Heart disease by 0.2499.
Conclusion By using the Logit model, we reject the null hypothesis as Stage 2 Hypertension increases the risk for Ischemic heart disease.
Omitted variable bias Sleep apnea may exaggerate the effect of Hypertension on developing Ischemic Heart Disease.
Limitation Diet, Socioeconomic condition, diabetes and stress levels could have been used as control variables to fit the model better. Inclusion of variables like race and gender can give interesting results.