Quasi-Fourier Modeling Individual and Count Outcomes

Abstract

This article develops a model based on time-series analysis for investigating health outcome, such as disease or death, in relation to levels of air pollution. The model is built under assumption that any health outcome belongs to multivariate hierarchical system and depends on meteorology, pollution, geophysical and socio-cultural variables in it. The possible connections in this system are considered and mathematically formalized by means of trigonometric and polynomial functions. The model is tested by two simulations: (1) with count outcomes taken from Philadelphia daily mortality records for the period from 1974 to 1980; (2) with the individual outcomes taken from the peak flow measurements in 48 adults with vulnerable respiratory system in Leipzig, Germany, October 1990 – April 1991.