There is growing evidence that the build-up of greenhouse gases may lead to significant increases in temperature during the next century. Climate change of the magnitude that is predicted would be likely to have a number of important impacts on human health. One of the most obvious consequences of climate change would be an increase in the frequency and severity of summer episodes of extreme heat. There is a substantial body of research evidence, much of it from the USA, showing that such heatwaves provoke substantial increases in mortality rates. This phenomenon has received less attention in Britain where typically cooler summer conditions make it a less pressing problem. This generally benign picture could change if predicted increases in summer temperatures lead to more heatwave-related deaths.
London is located in the region of the UK with the highest summer temperatures and it experiences a marked urban heat-island effect. Within Britain its population of approximately 7 million people is the one facing the greatest risk of increased heatwave-associated mortality as a result of climate change. Using data on weekly deaths and daily temperatures for London for the period 1974 to 1995 statistical models are developed of the dependence of mortality rates on temperature. These show that episodes of unusually high summer temperatures are associated with marked increases in mortality. Under present conditions this happens relatively infrequently but application of scenarios for climate change in the UK shows that such events may become much more common by the middle of the next century. Calculations based on the models of the dependence of mortality rates on temperature show that a substantial increase in the number of heatwave-related deaths can be anticipated. However, an interesting feature of the analysis is that the models show a quadratic relationship between death rates and temperature. This means that, even for the main summer months of June, July and August, temperatures are often below the climatic optimum at which mortality rates are minimised. Climate change would reduce the frequency of such cool summer conditions and the excess deaths associated with them. The paper goes on to estimate the net effects of these two opposing influences on summer mortality rates for different scenarios for climate change. The problems of predicting climate change impacts on health from analogue models based on natural climate variability are also considered