Biometeorological investigations of asthma morbidity in Tasmania using co-spectral analysis of time series

Abstract

The inter-relationships between asthma morbidity series in different parts of the State of Tasmania and the local, daily, meteorological factors were examined using spectral techniques. In addition to hospital morbidity series of 6 years length, volunteer asthmatics in 4 areas provided comparative series for 19 months. Biometeorological indices of wind chill, discomfort etc. were calculated for each day and used for analysis in addition to single weather elements.

Each morbidity index by geographic region and sex was analysed with each meteorological element/construct. For each pair first the individual spectra were calculated to demonstrate the dominant frequencies in their respective time series and then the co-spectrum between each pair of spectra was computed to assess the degree of coherence, or correlation between their frequency components. An examination of the coherence between the series provided evidence against the existence of cyclical meteorotropic behaviour in any of the study areas. Periodic fluctuations were seen to exist in the morbidity spectra but these were not related to any of the weather parameters that were investigated.