Dose-response curve of annoyance with adjustment for agebased on the standard population

摘要

Some papers detected significant relationship between age and annoyance of noise. However, most of the reported dose-response curves were not adjusted for age. Comparisons of crudernresponse rates may be misleading if the age composition differs in the populations being compared.rnThis paper proposes a method to obtain a dose-response curve with adjustment for age based on thernWHO world standard population. A logistic function was applied as a dose-response curve since it hasrnsome advantages e.g. (1) the response curve ranges from 0 to 1 and (2) the accuracy of the curve canrnbe evaluated. Multiple logistic regression yields dose-response curves of all age groups at the samerntime. The age-adjusted curve was approximately derived by standardizing the logits of these curvesrnbased on the standard population. To simplify this method, a special coding of dummy variables was introduced into the logistic regression. The age-adjusted logistic function and its confidence interval can be obtained easily by using this coding. The proposed method was applied to a field study conducted along trunk roads. The dose-response curve showed a remarkable change after the adjustment for age. This result indicates that the reported dose-response curves could be confounded if they are not adjusted for age.