베타회귀를 사용한 지역별 유병률과 미세먼지 연평균 농도의 관계 분석

초록

Although prevalence is a commonly used value in epidemiology, it is not modeled practically. Generally, epidemiologist modeled using logistic regression or Poisson regression, with related values of prevalence. In this paper, the prevalence was analyzed with a beta regression using the characteristics of prevalence with continuous values between 0 and 1. Using community health survey data and particulate matter data from the National Institute of Environmental Research, the prevalence of asthma, diabetes, hypertension, atopic dermatitis or dyslipidemia and the concentration of particulate matter, PM10 and PM2.5 were estimated in 2015. The analysis showed that there were some differences between PM10 and PM2.5, but there were the significant associations with the prevalence of asthma, diabetes, hypertension, atopic dermatitis or dyslipidemia. This is no different from the recent research on the particulate matter and diseases and we can see that the regional level of particulate matter needs to be controlled to prevent the disease.

키워드