Multivariate Analysis of the Impact of Demographic and Clinical Factors on Cardiovascular Health

Youssef Benfallah,Wanru Guo,Shoaib H Sarfraz,Yanyu Yang
DOI: https://doi.org/10.1101/2024.05.19.24307595
2024-05-20
Abstract:Background: This study investigates the influence of demographic and clinical factors on cardiovascular health using a multivariate analysis approach. We analyzed data from one of the largest heart-related disease datasets, merging variables such as age, sex, chest pain type, to determine their association with resting blood pressure, cholesterol, and maximum heart rate. Methods: Two-means comparison is used to investigate the effect of heart disease with response variables, One-way and two-way MANOVA are used to study the relationship between chest pain type, as well as sex and heart disease on the response variables, respectively. Profile analysis and MANCOVA were also conducted to determine a final model. Results: Our findings indicate significant effects of heart disease on resting blood pressure, cholesterol and maximum heart rate. Chest pain type also has a significant effect on maximum heart rate achieved, but not blood pressure and cholesterol. There are also notable interactions between sex and heart disease status on all response variables. Finally, we formulized a multivariate regression model with age, sex, chest pain type, heart disease, and the interaction between heart disease and sex as predictors. Conclusion: These findings can inform physicians on possible treatment strategies in cardiovascular diseases.
What problem does this paper attempt to address?
The main objective of this paper is to explore the impact of demographic and clinical factors on cardiovascular health and to assess the association between these factors and cardiovascular health indicators through multivariate analysis methods. Specifically, the study analyzed how factors such as age, gender, type of chest pain, and the presence of heart disease affect cardiovascular health indicators such as resting blood pressure, cholesterol levels, and maximum heart rate. The study used datasets from five large heart disease databases (Cleveland, Hungary, Switzerland, Long Beach, and Statlog), totaling 918 unique patient records. The data were comprehensively analyzed using various statistical testing methods (such as two-sample mean comparison, univariate and bivariate multivariate analysis of variance (MANOVA), profile analysis, and multivariate analysis of covariance (MANCOVA)). Key findings include: - Heart disease significantly affects resting blood pressure, cholesterol levels, and maximum heart rate. - The type of chest pain has a significant impact on maximum heart rate but not on blood pressure and cholesterol. - There is a significant interaction between gender and heart disease status, which affects all response variables (i.e., resting blood pressure, cholesterol, and maximum heart rate). - Age as a predictor was found to significantly improve the model's performance. These results can provide information to doctors to help them develop treatment strategies for cardiovascular diseases and may guide future research in cardiovascular health management.