Commentary: Climate Change: How Can Epidemiology Best Inform Policy?

Armstrong, Ben; Hajat, Shakoor; Kovats, Sari; Lloyd, Simon; Scovronick, Noah; Wilkinson, Paul

doi: 10.1097/EDE.0b013e31826d0d13
Climate Change

From the Department of Social and Environmental Health Research, London School of Hygiene and Tropical Medicine, London, United Kingdom.

The authors report no conflicts of interest or specific sources of funding.

Correspondence: Ben Armstrong, Department of Social and Environmental Health Research, London School of Hygiene and Tropical Medicine, 15–17 Tavistock Place, London WC1H 9SH. E-mail:

Article Outline

Evidence that human activity affects global climate1 has stimulated a broad, important, and urgent research agenda. This agenda includes questions about human health that require the tools of epidemiology. Several publications have reviewed the health dimensions of the climate change research agenda.2–6 Our objective here is to map out the specific contributions within that context that epidemiology can make. Beyond proposing a general scheme of such contributions, we do not attempt to be comprehensive, preferring to highlight issues and offer examples that have received less attention.

Back to Top | Article Outline

A General Scheme for the Contribution of Epidemiology

Epidemiology can play a part in informing policy questions by clarifying the health consequences of policies:

What resources are justified to reduce carbon emissions and other human activities affecting climate?

What are the best ways to adapt to climate change?

What are the best ways to reduce carbon emissions (mitigate climate change)?

Epidemiology can contribute answers to these questions by using already published epidemiologic findings that inform health risk assessments7 and by undertaking new primary epidemiologic research to improve the evidence base for such assessments.

The first task in performing health risk assessments is to identify pathways by which a change in climate (or a measure to adapt to or mitigate it) might drive changes in health. The links in these pathways are then usually characterized separately using knowledge from the literature, including but not exclusively epidemiologic research.

The relevant pathways between climate change or policies to mitigate climate change and health may be more or less direct (as illustrated in the Figure). Examples of direct impacts are acute effects of heat and cold on mortality, and the reduction of deaths caused by air pollution if fewer fossil fuels are burned. Examples of indirect impacts are health effects from climate-driven changes in mosquito populations, population migration after desertification or coastal flooding, or electricity price increases if more expensive renewable sources replace coal for electricity production under a climate mitigation policy.

Back to Top | Article Outline

What Resources (Financial and Other) Are Justified to Reduce Carbon Emissions?

Health will clearly be damaged if climate-changing emissions are not reduced. The relevant health risk assessments estimate future health under two scenarios: one with climate changes (as estimated from climate models) and one without. Broad-brush assessments have left little doubt that effects on health will be adverse overall, and substantial in some regions, but more detailed assessments are incomplete and uncertain.8 Better information is needed on the likely magnitude of carbon emission effects and their distribution over regions and across subgroups.

Back to Top | Article Outline

What is Expected from Climate Change?

Climate science predicts higher global average temperatures and changes to precipitation patterns, but with important regional differences. In addition to affecting overall means of distributions of weather variables, these changes will affect frequency and intensity of weather events, including heat waves and floods that are extreme for the baseline period (1970–2000).1

Back to Top | Article Outline

Direct Pathways Through Temperature

An increased mean temperature would be expected in most places to increase acute heat– (and heat-wave)-related deaths and decrease acute cold–related deaths, putting aside for now the issue of adaptation. These temperature effects can be estimated using health risk assessments based on time-series (or case crossover) studies of weather and health.9–11 There is some uncertainty over the extent to which acute temperature-related mortality excesses reflect deaths that would occur otherwise a few weeks or months later (“harvesting”). This is because of dependence on time-series analyses for temperature studies, in contrast with air pollution studies, for which cohort studies now dominate health risk assessment.12 There have been no analogous studies in which long-term health has been associated with long-term weather patterns (climate), controlling convincingly for other risk factors. Such studies present challenges, even compared with equivalent air pollution studies, because climate rarely varies much over small distances; nonetheless, such studies would be helpful if feasible.

Back to Top | Article Outline

Indirect Pathways

Most pathways from climate change to health are likely to be more complex,8 including, for example, health effects of climate change through agriculture. Epidemiologic knowledge would certainly play a part in assessment of impact on health through these more complex pathways but might be secondary to other considerations, such as agronomy, economics and other social sciences. The usual role of epidemiology in informing health risk assessments of such indirect pathways is to furnish exposure-response relationships for links relatively downstream in a causal chain otherwise informed by other disciplines. For example, the effect of dietary changes on health may require economic models to follow changes in food prices consequent to changes in agricultural production, based on information from agronomists and climate scientists.

The possible indirect pathways from climate change to health are so varied that a list of associations that new epidemiologic research could better inform would be long and open-ended. Epidemiology can provide information on the downstream links in a disaggregated multichain pathway. Epidemiology may also sometimes escape from the “prison of the proximate”13 to directly study associations across more distal nodes in a causal network. For example, new studies of the effects of food prices on diet-sensitive health outcomes (such as stunting) might inform the climate–agriculture–health association hypothesized above, without the need to examine price-nutrient and nutrient-health associations separately. In general, however, when pathways are indirect, epidemiologists should be part of interdisciplinary teams for health risk assessments, so that the health-related links in the causal path(s) can be appropriately studied and modeled.

Vector-borne and other infectious diseases are likely to contribute a substantial proportion of the disease burden because of climate change.14 A few pathways to health are relatively direct, such as cholera outbreaks after extreme weather events, which are predicted to increase in severity and frequency with climate change.4,15,16 However, most pathways are indirect. Much work has been done on broad-scale modeling of vector-borne diseases such as malaria, but this has been largely through disease/vector models rather than epidemiology.17 One area of need, with a potentially larger role for epidemiology, is to better describe climate effects at the local scale and thus inform local strategies in sensitive areas (eg, African highlands). Diarrheal diseases and malnutrition may contribute a high proportion of the burden of disease caused by climate change, again through indirect mechanisms that have so far been imperfectly estimated.18 Epidemiologic studies of past natural climate cycles (eg, El Niňo Southern Oscillation) have been particularly informative about climate effects on infectious diseases.19

Back to Top | Article Outline

What Are the Best Ways to Adapt to Climate Change?

A second critical need is for more evidence on the capacity of populations to adapt to a changing climate. Adaptation includes physiologic, behavioral, or physical mechanisms such as building modification. Some adaptation might occur in the absence of explicit adaptation policies, and assumptions about this present a difficult issue for health risk assessments.8 A more important question, given that some amount of climate change is considered inevitable, is what policies optimize adaptation. Epidemiology can inform both of these questions, although again usually by contributing to health risk assessments that draw from many disciplines.

It is clear from temperature time-series studies that populations are able to adapt to a range of climates, with people in hotter climates apparently tolerating much higher temperatures than those in colder climates,20–22 and the reverse for toleration of cold. There is also evidence in some places that heat and cold sensitivity of populations has changed over time, but such adaptation has not as yet been well characterized.23,24

Epidemiology, including time-series studies, can play an important role in clarifying adaptation, in particular by identifying modifiers of temperature effects. Use may be made of variation in putative modifiers within cities (if the information is available for individual cases) or between cities (in multicity studies), although this ecologic approach lacks power and is more subject to bias. Several studies have found evidence that the proportion of houses with air conditioning substantially reduces heat effects at the city level,20,25 although this adaptation has obvious limitations unless electricity production is decarbonized. The study of other aspects of housing in modifying temperature-mortality functions has been limited by lack of appropriate housing data, in particular linked individually to health outcomes.

Other epidemiologic designs can also contribute to clarifying adaptation. Daily time-series studies that span several decades can investigate changes in temperature-health relationships over time (time-varying coefficient models), as noted above.23,24 If sufficiently varied cities are included, such studies might also suggest specific determinants of such changes. There are also possibilities for studying interventions to reduce the impact of heat, such as heat-watch warning systems.26 Such intervention studies can also improve the robustness with which weather effects can be estimated. Cohort studies could also study either within- or between-city modifiers. Studies of occupational groups subject to high temperature are useful in particular for effects of high temperature (50°C and higher)—to which populations have not been much exposed but may be in the future.

The role of epidemiology in regard to adaptation to less direct effects of climate change is most likely to be in the context of complex multidisciplinary health risk assessment.

Back to Top | Article Outline

What Are the Best Ways to Reduce Carbon Emissions?

There is now international consensus supporting actions to mitigate climate change by reducing the emission of greenhouse gases. However, various mitigation policies themselves may have health effects that must be clarified. Health risk assessments can inform the choices among mitigation options and, if the health effects are positive, can increase political acceptability of a mitigation policy. Many mitigation policies can have “cobenefits.”27 In particular, shifting away from burning fossil fuels generally improves health through reducing air pollution. There have also been adverse impacts identified (at least without adaptation), such as the effects of increased indoor radon concentrations with more airtight housing.28 The Table shows areas that have been identified as particularly important for climate change mitigation and possible pathways to health.

As with climate change itself, the indirect pathways by which mitigation can affect health are likely to be more important than direct pathways. For example, one alternative for motorized transport is to replace gasoline in internal combustion engines with biofuels such as ethanol or biodiesel. This can change combustion-related air pollution emissions, but the direct health impact seems likely to be relatively modest.29 A potentially larger effect could be through pressures on food prices as a result of increased competition with the energy market for land and feedstock crops.30,31 Making an estimate of impact through this pathway, however, requires information on associations of biofuels with food prices, and food prices with nutrient intakes (which also depends on income), and how the expected dietary changes would influence health (the epidemiologist’s input).

There is another example that has received scant attention from epidemiologists. Although not considered a mitigation measure, emissions of SO2 transform by atmospheric chemistry to sulfate particles (usually ammonium sulfate) and mitigate global warming in the short term by reflecting the sun’s radiation back into space.1 Pumping SO2 into the stratosphere remains, with other geoengineering options, on the outer edges of the current discussions. However, whether the climate warming caused by scrubbing SO2 out of emissions of current and planned coal-fired power stations is outweighed by health and other benefits is a question that epidemiologic evidence on particulate air pollution might inform. Toxicologists generally find pure ammonium sulfate biologically inactive.32 Epidemiologic studies (cohort and time series) have found sulfate particles associated with excess mortality,33 but it is unclear whether these associations are due to confounding by other particles. Clarifying this would not decide the question (there are other considerations around SO2 emissions and costs of scrubbing), but it could importantly inform it.

Back to Top | Article Outline


This brief overview has omitted many issues, some of which we mention briefly here. We have not mentioned direct epidemiologic study of the association of long-term changes in health with climate change. Unfortunately, the chances of disentangling the long-term effects of climate change from other changing risk factors seems too challenging to accomplish for other than a handful of climate-sensitive diseases not strongly influenced by other risk factors. This challenge will diminish, but not disappear, as climate change gathers pace.

We have suggested that the larger health effects of climate change and mitigation measures will be through complex indirect pathways. Within these pathways, there may be feedback loops (air conditioning causing global warming) and nonlinearities, including discontinuities, that require dynamic systems models. These are certainly issues for health risk assessment and, in some cases, may be so for epidemiology itself. Related to this, it is important for epidemiologists involved with “integrated assessment” of climate change effects to take part in the planning stages of such assessment, to ensure that teams modeling processes upstream to health know what data are needed and to ensure that feedback loops are considered.

It is natural that epidemiology will initially be focused on the more easily estimable effects—typically, the more direct pathways, such as the acute impact of heat or particulate pollution on mortality. But this practice can become a problem if more complex pathways are ignored entirely, or health assessment is presented without acknowledgment of its incompleteness—bad information may be worse than no information. Uncertainties other than those arising from epidemiologic associations may in any case dominate.34 For many policy questions, local assessments, which are not as uncertain as broad-brush global ones, may be the most relevant.

Epidemiologists have contributed, of course, to a substantial body of work relevant to climate change beyond the scope of this commentary—in particular, many of the high-profile synthesis reports of organizations concerned with climate change (ie, the Intergovernmental Panel on Climate Change [IPCC],8 United Nations Environment Program,35 and the World Health Organization.)18 As these issues continue to take prominence, more epidemiologists will be needed to work in these organizations and to augment the epidemiologic evidence base on which synthesis reports rely.

In summary, epidemiologists can play an important role in the policy choices relating to climate change by contributing in multidisciplinary teams to health risk assessments of climate change itself, of adaptation to climate change, and of policies to mitigate climate change. New epidemiologic studies can reduce uncertainty in such assessments and will require an assortment of study designs, depending on the research question. Because most of the effects on health are likely to be through complex indirect pathways, epidemiologists should be prepared to tackle such pathways to remain relevant.

Back to Top | Article Outline


1. Solomon S, Qin D, Manning M, et al. Climate Change 2007: The Physical Science Basis: Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change.. 2007 Cambridge, UK: Cambridge University Press
2. Epstein PR. Climate change and human health. N Engl J Med. 2005;23:1433–1436
3. Hanna EG, McMichael AJ, Butler CD. Climate change and global public health. Routledge Handbook of Global Public Health Taylor & Francis. 2011
4. McMichael AJ, Lindgren E. Climate change: present and future risks to health, and necessary responses. J Intern Med. 2011;270:401–413
5. McMichael AJ, Woodruff RE, Hales S. Climate change and human health: present and future risks. The Lancet. 2006;367:859–869
6. Martens WJM, McMichael AJ Environmental Change, Climate, and Health: Issues and Research Methods. 2002 Cambridge, UK: Cambridge University Press
7. EPA.NRCCoIRAAUbtU, Studies NRCBoE, Earth NRCDo, Studies L. Science and decisions: advancing risk assessment. 2009 Natl Academy Press
8. Parry ML, OFC JP, Palutikof PJ, van der Linden CE, Hanson eds. Climate Change 2007: Impacts, Adaptation and Vulnerability: Contribution of Working Group Ii to The Fourth Assessment Report Of The Intergovernmental Panel On Climate Change. 2007 Cambridge, UK: Cambridge University Press
9. Huang C, Barnett AG, Wang X, Vaneckova P, FitzGerald G, Tong S. Projecting future heat-related mortality under climate change scenarios: a systematic review. Environ Health Perspect. 2011;119:1681–1690
10. Langford IH, Bentham G. The potential effects of climate change on winter mortality in England and Wales. Int J Biometeorol. 1995;38:141–147
11. Peng RD, Bobb JF, Tebaldi C, McDaniel L, Bell ML, Dominici F. Toward a quantitative estimate of future heat wave mortality under global climate change. Environ Health Perspect. 2011;119:701–706
12. Künzli N, Kaiser R, Medina S, et al. Public-health impact of outdoor and traffic-related air pollution: a European assessment. Lancet. 2000;356:795–801
13. McMichael AJ. Prisoners of the proximate: loosening the constraints on epidemiology in an age of change. Am J Epidemiol. 1999;149:887–897
14. Confalonieri U, Menne B, Akhtar R, et al.Parry ML, Canziani OF, Palutikof JP, Linden PJvd, Hanson CE Human health. Climate 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. 2007 In: Cambridge, UK Cambridge University Press:391–431
15. Meehl GA, Stocker TF, Collins WD, et al.Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL Global climate projections. In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. 2007 Cambridge, UK Cambridge University Press
16. Schwartz BS, Harris JB, Khan AI, et al. Diarrheal epidemics in Dhaka, Bangladesh, during three consecutive floods: 1988, 1998, and 2004. Am J Trop Med Hyg. 2006;74:1067–1073
17. Gething PW, Smith DL, Patil AP, Tatem AJ, Snow RW, Hay SI. Climate change and the global malaria recession. Nature. 2010;465:342–345
18. McMichael AJ, Campbell-Lendrum D, Kovats RS, et al.Ezzati M, Lopez AD, Rodgers A, Murray CJL Global climate change. In: Comparative Qantification of Health Risks: Global and Regional Burden of Diseases Attributable to Selected Major Risk Factors. 2005;Vol. 1 Geneva, Switzerland World Health Organization:1543–1649
19. Kovats RS, Bouma MJ, Hajat S, Worrall E, Haines A. El Niño and health. Lancet. 2003;362:1481–1489
20. Anderson BG, Bell ML. Weather-related mortality: how heat, cold, and heat waves affect mortality in the United States. Epidemiology. 2009;20:205–213
21. Baccini M, Biggeri A, Accetta G, et al. Heat effects on mortality in 15 European cities. Epidemiology. 2008;19:711–719
22. McMichael AJ, Wilkinson P, Kovats RS, et al. International study of temperature, heat and urban mortality: the ‘ISOTHURM’ project. Int J Epidemiol. 2008;37:1121–1131
23. Barnett AG. Temperature and cardiovascular deaths in the US elderly: changes over time. Epidemiology. 2007;18:369–372
24. Carson C, Hajat S, Armstrong B, Wilkinson P. Declining vulnerability to temperature-related mortality in London over the 20th century. Am J Epidemiol. 2006;164:77–84
25. Braga AL, Zanobetti A, Schwartz J. The effect of weather on respiratory and cardiovascular deaths in 12 U.S. cities. Environ Health Perspect. 2002;110:859–863
26. Michelozzi P, de’ Donato FK, Bargagli AM, et al. Surveillance of summer mortality and preparedness to reduce the health impact of heat waves in Italy. Int J Environ Res Public Health. 2010;7:2256–2273
27. Haines A, McMichael AJ, Smith KR, et al. Public health benefits of strategies to reduce greenhouse-gas emissions: overview and implications for policy makers. Lancet. 2009;374:2104–2114
28. Wilkinson P, Smith KR, Davies M, et al. Public health benefits of strategies to reduce greenhouse-gas emissions: household energy. Lancet. 2009;374:1917–1929
29. CETESB. . Tabela 7 - Fatores médios de emissão de veículos leves novos. Available at:—fatores-m?dios-de-emiss?o-de-ve?culos-leves—novos-(1). Accessed April 3, 2012
30. Naylor R, Liska A, Burke M, et al. The ripple effect: biofuels, food security, and the environment. Environment: Science and Policy for Sustainable Development. 2007;49:30–43
31. Rosegrant MW, Zhu T, Msangi S, Sulser T. Global scenarios for biofuels: impacts and implications. Review of Agricultural Economics. 2008;30:495–505
32. Schlesinger RB, Cassee F. Atmospheric secondary inorganic particulate matter: the toxicological perspective as a basis for health effects risk assessment. Inhal Toxicol. 2003;15:197–235
33. Smith KR, Jerrett M, Anderson HR, et al. Public health benefits of strategies to reduce greenhouse-gas emissions: health implications of short-lived greenhouse pollutants. Lancet. 2009;374:2091–2103
34. Dessai S. Heat stress and mortality in Lisbon Part II. An assessment of the potential impacts of climate change. Int J Biometeorol. 2003;48:37–44
35. UNEP. Integrated Assessment of Black Carbon and Tropospheric Ozone: Summary for Decision Makers.. 2011 Nairobi, Kenya United Nations Environment Programme & World Meteorological Association
© 2012 Lippincott Williams & Wilkins, Inc.