What determines spatiotemporal variations in cold-weather-related mortality in England?

Abstract

Mortality rates in England are higher during the winter period (December-March) compared to other seasons. Most excess winter deaths are caused by respiratory and circulatory conditions, which are exacerbated by cold temperatures. Excess winter mortality rates vary between areas and years. This research investigated spatial and spatiotemporal variations in excess mortality across small areas of England in relation to winter season, cold weather and other explanatory factors. A systematic review was undertaken, which identified factors associated with modified risk of specific types of winter- and cold-related adverse health and social outcome, and assessed the effectiveness of interventions on reducing these adverse impacts. Evidence-based pathway models were developed of associations between winter season or cold exposure and circulatory and respiratory health outcomes, in relation to explanatory factors. Secondary data were identified to represent variables from the pathway models in analyses. Poisson regression models were implemented using a Bayesian approach to evaluate spatial and spatiotemporal variations in observed-to-expected mortality ratios from circulatory and respiratory conditions in relation to covariates, across English Local Authorities and between: winter seasons, periods of cold and warmer weather and months of the year. Climatic factors were associated with spatiotemporal variations in mortality ratios between winter periods, but were less important determinants of excess mortality across the year. This supports the use of weather forecasting services to alert health and social care providers to predicted adverse weather conditions, in order to support individuals with circulatory and respiratory conditions during the winter period. The effects of social factors were similar for circulatory and respiratory conditions and were predominantly non-seasonal. Thus, interventions could be developed to reduce spatial variations in excess mortality from both condition groups on a year-round basis. Further research using morbidity outcome data could provide information to reduce spatial variations in excess medical consultation rates across England.