Risk factors of the childhood asthma in the Colombo Municipal Council area: Bayesian approach on logistic regression model

Abstract

Asthma is one of the most common chronic diseases in the child-hood in both developing and developed countries. The objective of this study was to examine the relationship between the childhood asthma and the risk factors among children aged 5-12 years in the Colombo Municipal Council area using Bayesian inferential method. In Bayesian inferential methods Gibbs sampling algorithm was used to select samples from posterior distribution of logistic coefficient under non-informative prior using WinBUGS 1.4 software. In checking convergence “live trace plots” of sample of parameters were obtained. Also the simulation was run until the Monte Carlo error for each parameter of interest is less than about 5% of the sample standard deviation. Then mean standard deviation Monte Carlo error and some percentiles of posterior distribution were computed. According to the study, asthma was more likely to occur among boys than girls (Odds Ratio [OR] =1.48, 95% Credible Interval [Cl] =1.09 to 1.98). Children those who do not have allergies (OR=0.59, 0=0.39 to 0.90) and children those who have had no family history of asthma (OR=0.27, 0=0.19 to 0.36) have a lower risk of having asthma. If the parents’ income is high then possibility to getting exposed to diseases becomes low. Children, who used mattresses to sleep, have a possibility of getting exposed to disease asthma. If children live with more number of persons in the house, the possibility of occurring disease is low.