Association between ambient temperature and dementia‐related hospital admissions in England

Abstract

AbstractBackgroundThe impacts of a changing climate on dementia burdens have not been widely explored, despite the increasing pressures associated with dementia on the public health and social care systems. This study explores the association between ambient temperature and the dementia‐related emergency hospital admission, using the national hospital inpatient dataset for England. We also examine important modifiers of the heat‐risk, including separate results for each Government region of the country.MethodsDaily counts of emergency admissions with a primary diagnosis of dementia (ICD codes F00‐F03) and daily ambient temperature data from Met Office land surface stations were obtained for the period 2000‐2009 in England. Poisson regression was used to detect short‐term associations between daily mean temperature and daily counts of dementia admissions, adjusting for overdispersion and day‐of‐week effects, and including spline functions to model any underlying trends and seasonal patterns in the series. Distributed lag non‐linear models employing cross‐basis functions were used to characterise the exposure‐response relationship, at both national and regional levels at different lags of temperature; a 7‐day lag structure was used to explore potential delayed effects of temperature on the week leading up to admission. The modifying effects of age‐group, urban/rural residence, and quintiles of area‐level deprivation were also examined.ResultThere was some suggestion of seasonality in dementia admissions, with numbers peaking in the summer and winter months each year. After seasonal adjustment, there was an increase in the risk of dementia‐related hospital admission associated with high temperature, with adverse effects becoming apparent once daily mean temperatures rose above 15.0°C. At the national level, we estimated an approximately 10% increase in the risk of dementia admission for every 5°C increase in temperature above 15.0°C. When disaggregated by regions, the risk was highest in London. Separate results by age‐group, deprivation and urban/rural residence will be presented.ConclusionThe findings are likely to have important public health implications for priority settings, such as for hospital emergency department planning during hot seasons. The dangers of heat‐induced symptoms among dementia sufferers can be anticipated by caregivers whenever hot weather is forecast.