High temperatures and cause-specific mortality

• Epidemiology
• public health
• climate
• cancer
• hygiene/occupational hygiene

High ambient temperatures are hazardous, with a significant body of research documenting increases in morbidity and mortality during hot weather.1 And yet, every heat-related death is preventable. Improving understanding of who is most vulnerable, where they are located and early warning systems that effectively communicate the risks of hot temperatures could decrease the current health burden and increase resilience to a warmer world.

Surprisingly, understanding is limited of the characteristics that increase vulnerability.1 Gasparrini et al (this issue) provide a comprehensive analysis of specific causes of death that were increased during hot weather in regions of England and Wales. Overall mortality increased 2.1% per degree Celsius above a heat threshold identified for each region. For stroke, ischaemic heart disease, chronic ischaemic heart diseases and respiratory infections, the attributable burden increased >10%. The attributable burden is lower for other causes, but the absolute numbers of deaths is large (eg, there were >1000 deaths each over the study period for myocardial infarction, chronic obstructive pulmonary disease, nervous system disorders and external causes). Effective early warning systems and healthcare could have prevented these deaths. These and similar studies can inform physiological studies to better understand the mechanisms by which hot temperatures exacerbate existing health conditions. It is important to distinguish the degree to which a disease process increases risk versus drug treatment (drugs can affect thermoregulatory mechanisms) versus general fitness versus socioeconomic and other factors. Such information will help healthcare providers improve care for those who may be susceptible to hot weather and is critical for developing effective early warning systems that recommend tailored actions for at risk groups, including older adults, children, those with certain chronic diseases and physical labourers.

An issue not addressed in Gasparrini et al is whether vulnerability to hot temperatures is affected by mortality in the previous winter season. Studies in Sweden and Italy report that lower than average winter mortality increases the health impacts of hot weather the following summer.2 3 If these results are replicated elsewhere, then this suggests that individuals with certain diseases may have a general susceptibility to extreme temperatures such that deviations from average temperatures are potentially fatal. Such information could be used to enhance health protection under current conditions. Furthermore, this suggests that current projections of the burden of heat-related mortality under climate change may underestimate impacts. Projections also are needed of health impacts during longer and more intense heatwaves and during heatwaves occurring earlier in the summer to better estimate the total health burden of changing hot weather patterns.4

Although early warning systems save lives,5 continued heat-related deaths indicates that improvements are needed in effectively motivate susceptible individuals to follow guidance on actions to reduce their core body temperature during hot weather. The focus of many heatwave early warning systems is on determining the threshold at which mortality begins to increase; these systems implicitly assume that communicating warnings is sufficient for those at risk to take action. However, a particular challenge is that vulnerable individuals often do not perceive that they are at risk.6 Research is needed on how to encourage these individuals to recognise their susceptibility and make appropriate behavioural changes.

Better understanding the factors that increase vulnerability to hot temperatures, how those factors may change in the future and how to increase health protection is of increasing importance as temperatures continue to increase with climate change. Ambient temperatures increased over the past several decades, with greater change in more pole ward regions.7 There is a greater than 90% likelihood that average Northern Hemisphere temperatures during the period 1950 to 2000 were higher than during any other 50-year period in the past 500 years, with a greater than 66% likelihood that they were the highest in at least the past 1300 years.7 In addition, there has been an increase in the frequency and duration of heatwaves. Further increases in ambient temperatures and associated increases in the frequency, intensity, duration and spatial extent of heatwaves are projected in the coming decades.6 Unless actions are taken, these increases will be associated with a greater burden of heat-related mortality.8 9 However, hot weather is not the only factor affecting vulnerability; urban infrastructure, housing characteristics, population ageing, increasing numbers of people with chronic diseases, acclimatisation and other factors determine the overall burden of hot temperatures. The next generation projections of heat-related mortality should aim to estimate the extent to which climate change and these and other factors interact to affect health burdens.

Increasing heat-related mortality with climate change underscores the importance of implementing a range of strategies, policies and measures to reduce vulnerability over the short and longer term, including improving building and community design to reduce urban heat islands, management of chronic diseases taking into account temperature susceptibility, vulnerability mapping, designing early warning and response systems that motivate appropriate behavioural change, and reducing the greenhouse gas emissions underlying climate change.