Background & Objectives: longitudinal studies play an important role in medical research. In these studies, every individual are exposed to the repeated measures over time. Specific statistical methods that consider the correlation are used to analyze for such data. But these methods differ in providing reliable and efficient results that are effective in covariates significantly. Therefore, in this study, in addition to checking factors affecting on recurrence Bipolar I Disorder using Generalized Estimating Equations and Quadratic Inference Function, we compare the efficiency of two methods.
Material & Methods: In this longitudinal study, 237 patients with Bipolar I Disorder and history of hospitalization in Zare hospital in Sari were studied . Each of patients was followed in 2007-2011. The applied methods were the Generalized Estimating Equations and Quadratic Inference Function.
Results: parameter estimation efficiency by Quadratic Inference Function was more than Generalized Estimating Equations method. Using the Quadratic Inference Function method, the effect of variables- age at onset, first-degree relatives and location were significant, but in the Generalized Estimating Equations method, location variable was not significant
Conclusion: The data used in this study showed that the estimates of the Quadratic Inference Function method is more efficient than Generalized Estimating Equations method.
Keywords: Longitudinal Data, Generalized Estimating Equations, Quadratic Inference Function, Bipolar I Disorder
Rights and permissions | |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |