Volume 19, Issue 97 (7-2012)                   RJMS 2012, 19(97): 1-9 | Back to browse issues page

XML Persian Abstract Print

Tehran University of Medical Sciences
Abstract:   (7587 Views)

  Background : One of the most important complications of diabetes, is diabetic retinopathy that causes the blindness of 10,000 people every year. Different researches have been done on retinopathy risk factors in diabetic patients. This study was carried out to check the type of relationship between retinopathy risk factors and the condition of temptation it with generalized additive models. The study attempts to increase the quality of predicting the response variable and to reveal the non-linear and non-monotonic relationships between the response and the set of explanatory variables with generalized additive models.

  Methods: This cross-sectional study has been done on 367 diabetic patients who take part in assessment recall of retinopathy in Tehran. Entrants have been checked to determine their particulars, medical conditions and medicines. Finally has been used a data complex for fitting generalized additive models and binary logistic regression, including six continues explanatory variables: age, duration of diabetes, Body Mass index (BMI ), hemoglobin A1C, cholesterol, systolic blood pressure and response variable, to the presence of retinopathy Fitting model has been done with mgcvR software.

  Results: In this study 120 cases (33%) were retinopathy patient and 247 cases (67%) were not. Results of the generalized additive model were denoting that following factors have affected on retinopathy: duration of diabetes, hemoglobin and systolic blood pressure. Moreover, it has been mentioned that duration of diabetes with linear function, hemoglobin with function of degree four and systolic blood pressure with quadratic function was related to retinopathy.

  Conclusion: In addition to determination of retinopathy risk factors, it has been shown that generalized additive model can identify nonlinear relationship between variables. Therefore this model increases the quality of predicting response variable, with more information of the data relationships.

Full-Text [PDF 351 kb]   (2104 Downloads)    
Type of Study: Research | Subject: Biostatistics

Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.