Risk factors associated with cancer grade progression: A longitudinal study using generalized estimating equations

Cancer is the second leading cause of death worldwide and remains a major public health challenge.It results from the uncontrolled proliferation of abnormal cells due to disruptions in normal cellular processes, leading to tumor formation.Identifying factors that influence cancer progression, particularly changes in cancer grade, is essential for improving patient outcomes.This study analyzes longitudinal data to evaluate the effects of several risk factors on variations in cancer grade.The Generalized Estimating Equations (GEE) method, an extension of Generalized Linear Models (GLMs), was used to analyze correlated longitudinal data.Hypothesis testing was conducted using the Wald test, while model selection was based on the quasi-likelihood under the independence model criterion (QIC) and its corrected version (QICu).The analyses were performed using SPSS and R. The results showed that sex had no significant effect on cancer grade progression, whereas race, age, surgery, and laterality significantly influenced cancer grade progression.The exchangeable correlation structure provided the best model fit.Overall, the findings demonstrate that GEE is an effective method for identifying determinants of cancer grade progression in longitudinal studies.