Temperature related effects on mortality and years of life lost in the UK forcurrent and future climates

Abstract

Background: In the context of a warming climate, understanding current and potential future temperature risks to health is vital to allow effective prioritisation of resources and protect public health. However, important gaps in knowledge remain around current temperature-related risks and premature mortality, population adaptation to temperature effects, the best use of temperature thresholds in epidemiological models and local climate health impacts under conditions of the 2015 Paris Agreement, which aims to limit increases in global average temperature to well below 2 °C (preferably below 1.5°C) compared to preindustrial levels. Methods: In this thesis, I used a number of methods to assess current and future impacts of temperature on health, including: a systematic review to assess temporal changes in temperature-related health risks; epidemiological time series regression analysis of UK mortality data from Greater London, Greater Manchester and the West Midlands to estimate effects of ambient temperature on mortality and years of life lost (YLL) and evidence for temporal changes in temperature-related risk; a case study approach to examine cold definition and threshold use within studies; and quantitative health impact assessment methods to estimate changes in heat related mortality (HRM) under the Paris agreement in the 3 largest UK conurbations. Results: In Greater London, estimated risk and attributable burdens are sensitive to cold threshold choice (below which effects are quantified). Integrating evidence from multiple disciplines allowed causality across the range of ‘cold’ temperatures used and the implications of threshold placement for policy and research to be better understood. Evaluating temperature effects on (premature) mortality, I found an increased risk of YLL and mortality for each 1°C above or below the heat and cold thresholds, e.g. heat-effects were greatest in London, where for each 1 °C above the heat-threshold the risk of mortality increased by 3.9% (95% CI 3.5%, 4.3%) and YLL increased by 3.0 % (95% CI 2.5%, 3.5%). The systematic review found evidence of decreasing temporal susceptibility to heat in certain populations but little evidence for changes in cold related risks. Analysing UK data, however, I found no evidence of decreasing vulnerability to risk of heat or cold related mortality or YLL between 1996-2013. Lastly, I found that under conditions of the Paris agreement in the 3 largest English conurbations, HRM is projected to increase by 60-68% if the climate stabilises at 1.5°C compared to an increase of 100-110% under 2°C scenarios, depending on location. Conclusions: This PhD has demonstrated an increased risk of heat and cold related mortality and YLL in the 3 largest UK conurbations. Alongside the lack of evidence for attenuation in this risk and projections of future HRM, this has direct implications for UK public health planning and adaptation needs. Quantifying avoidable deaths under 1.5 degrees compared to 2 degrees of global warming provides timely motivation for increased climate mitigation ambition.