Environmental noise pollution is an important social problem. Noise is known to have an adverse effect on human emotions and bodies. However, the methodology of previous studies did not consider selection bias in eliminating participants during the screening process. Therefore, for this study, we propose a framework that combines propensity score matching with a generalized additive model to reduce that sampling problem. Within this framework, we use health data from the National Health Insurance Service and noise data from the National Noise Information System in Korea. Using the proposed framework and data set, we analyze the effects of noise on cardiocerebrovascular disease. Our results show that, when daytime noise increases by 1 A-weighted decibel (dB(A)), cerebrovascular disease increases by 0.66%, hypertension increases by 0.17%, and heart disease increases by 0.38%. Moreover, we conducted a scenario analysis to investigate the effects of noise reduction policies. When noise levels are reduced to meet regulatory targets, cerebrovascular diseases decrease by 2077 per million people, high blood pressure decreases by 5705 per million people, and heart disease decreases by 1151 per million people. Our results thus provide information about noise exposure-response functions in Korea that could be used to establish noise reduction policies.
Keywords: Cardiocerebrovascular; Cohort; Generalized additive model; Health effect; Noise; Propensity score matching.
Copyright © 2018 Elsevier B.V. All rights reserved.