Application of penalized logistic regression for detecting risk factors interactions of children with Cerebral Palsy

Identification of risk factors and their interactions is important in medical studies. The aim of this study was to identify the interaction of risk factors of cerebral palsy in children with 1-6 years of age. In this cross-sectional study the data of 225 children, 1-6 years of age and from 2008 to 2009, were corporate in penalized logistic regression analysis to identify the interactions between the risk factors for cerebral palsy in affected children. Data analysis was performed using the software R version 2.15.2. Selective regression model analyzed the data from 89 children with cerebral palsy and 109 healthy children by using forward step-wise procedure showed the significant difference between the main effects of asphyxia (P<0.001), consanguinity (P=0.003), preterm delivery (P=0.002) and current pregnancy (P=0.001) with cerebral palsy. While using a forward stage-wise procedure of any main effects of asphyxia (P<0.001), consanguinity (P<0.001) and the effects of preterm delivery: consanguinity (P=0.040), current pregnancy: consanguinity (P=0.002) were identified. Penalized logistic regression is suggested to identify interactions between risk factors of disease such as cerebral palsy