Applying Linear and Nonlinear Regression Analysis Models to Study the Possible Relationships between Cancer’s Risk and CVD/Stroke Risk over a 12-Year Period for a Type 2 Diabetes Patient Based on GH-Method: MathPhysical Medicine (No. 551)

Abstract

In the author’s previous medical research reports, he mainly applied physics theories, engineering models, mathematical equations, computer big data analytics and artificial intelligence (AI) techniques, as well as some statistical approaches to explore and interpret various biophysical phenomena. However, the majority of medical research papers he has read thus far are primarily based on statistics. As a result, in this article, he selects some basic statistical tools, such as correlation, variance, p-values, and regression analyses, to study the predicted Cancer risk probability as the output (dependent variable) by using his MI-based CVD/Stroke risk as the input (independent variable). Since 1/1/2012, the author has been collecting various data related to his health (~3 million data) which includes 4 categories of medical conditions, obesity, diabetes, hypertension, and hyperlipidemia (m1 through m4), along with 6 categories of lifestyle details, including exercise, water intake, sleep, stress, food, and daily life routines (m6 through m10). Due to his limited knowledge in earlier years, the datasets from 2010 to 2012 are incomplete; therefore, the data used in this study for the initial period of 2010-2012 are his best-guessed data. Previously, he researched and published a few articles regarding the risks of having CVD/Stroke and Cancers based on his enhanced metabolism index (MI) model. In this particular paper, adopting a regression analysis model, he is able to compare the previously calculated Cancer risks based on the enhanced MI model versus the newly regression predicted Cancer risks using his MI-based CVD risks as input. In this study, he will not repeat the detailed introduction of the regression analysis in the Method section because it is available in many statistics textbook. It should be noted that in regression analysis, the correlation coefficient R should be > 0.5 or 50% to indicate a strong inter-connectivity and the p-value should be < 0.05 to be considered as statistically significant. In summary, there are two specific conclusions worth mentioning: 1. The time-domain analysis results have revealed that the correlation between his Cancer risk and CVD risk are very high (89%). In addition, his enhanced MI-based model Cancer Risk and Regression Predicted Cancer Risk are highly correlated as well. 2. The space-domain linear regression analysis has shown that there is a variance of 1.0 existing between his Cancer risk and CVD risk. When using the nonlinear polynomial model, his variance still reaches 99.6%. This finding has further proven his Cancer risk and CVD risk are highly correlated together. Furthermore, he has conducted two more linear regression analyses of Cancer risk vs. Lifestyle and CVD risk vs. Lifestyle which reveal two extremely high linear variances at 1.0. These findings offer additional indication that Lifestyle, the common root-cause, contributes highly on both CVD risk and Cancer risk.