Lecture

Harben Lecture
Climate change and human health: impacts, vulnerability,
and mitigation
A Haines, R S Kovats, D Campbell-Lendrum, C Corvalan

It is now widely accepted that climate change is occurring as a result of the accumulation of greenhouse gases in the Lancet 2006; 367: 2101–09
atmosphere arising from the combustion of fossil fuels. Climate change may affect health through a range of pathways— See Comment page 2039
eg, as a result of increased frequency and intensity of heat waves, reduction in cold-related deaths, increased floods and London School of Hygiene and
droughts, changes in the distribution of vector-borne diseases, and effects on the risk of disasters and malnutrition. The Tropical Medicine, London, UK
(Prof A Haines MD,
overall balance of effects on health is likely to be negative and populations in low-income countries are likely to be
R S Kovats MSc); and WHO,
particularly vulnerable to the adverse effects. The experience of the 2003 heat wave in Europe shows that high-income Geneva, Switzerland
countries might also be adversely affected. Adaptation to climate change requires public-health strategies and improved (D Campbell-Lendrum DPhil,
surveillance. Mitigation of climate change by reducing the use of fossil fuels and increasing the use of a number of C Corvalan PhD)
renewable energy technologies should improve health in the near term by reducing exposure to air pollution. Correspondence to:
Prof A Haines, London School of
Hygiene and Tropical Medicine,
Introduction of air pollutants—eg, tropospheric ozone pollution may Keppel Street, London
It has been known for thousands of years, at least since be higher in some areas of Europe and lower in others— WC1E 7HT, UK
the time of Hippocrates, that climate has wide ranging but the relations are still imperfectly understood.4 Sea [email protected]
impacts on health. Increasing recognition of the process level rise is likely to threaten low lying coastal populations, Published simultaneously with
of climate change has led to a growing interest by health particularly in countries where economic conditions do Public Health; DOI:10.1016/
j.puhe.2006.01.002
researchers in assessing the potential mechanisms by not allow construction of sea defences and other counter
which changes in climate could influence health measures. There are also concerns that flooding, drought,
(figure 1). Such health effects will be modulated by factors and environmental degradation associated with climate
such as socioeconomic development and by the degree to change may lead to population displacement and more
which effective adaptation measures are implemented. environmental refugees.
Although most studies have assessed the potential Research on the health impacts of climate change
impacts of climate change in isolation from other addresses three main topics: current associations
environmental changes, in reality climate change will be between climate and disease, the effect of recent changes
experienced against a background of other global in climate, and the evidence base for projecting the future
changes—eg, population growth, urbanisation, land use impacts of climate change on health (figure 2).
changes, and depletion of fresh water resources—that Temperatures have been increasing globally for the past
themselves have implications for health and that could, two to three decades. The detection and attribution of
in some instances, interact with climate change to health effects to these changes has become a key research
magnify the impacts. challenge.5 This climate warming is projected to continue
This article was the subject of the 2005 Harben Lecture and accelerate, so that by the end of this century global
of the Royal Institute of Public Health given by one of us mean temperature will have increased by 1·4–5·8°C.6
(AH). It covers some of the ground of previous overviews1,2 Effects at the upper end of the range are more difficult to
but adds discussion of adaptation options and the predict and likely to be more seriously adverse.7
potential use of “mitigation strategies”—eg, energy
efficiency and renewable technologies—to contribute to Has observed climate change already been
near-term reductions in mortality. affecting human health?
There are several mechanisms by which climate can A growing number of studies present evidence for the
affect health. Extremes of temperature and rainfall—eg, effects of observed climate change on vector-borne and
heat waves, floods, and drought—have direct immediate other infectious diseases. Although the literature to date
effects on mortality as well as longer-term effects. For does not constitute strong evidence of an impact of
example, populations that have experienced flooding may climate change on human vector-borne diseases (eg,
suffer from sustained increases in common mental malaria), there is now evidence of vector species
disorders.3 Climate change is also likely to affect responding to recent climate change in Europe.8 There
biodiversity and the ecosystem goods and services that have been latitudinal shifts in ticks that carry tick-borne
we rely on for human health. Changes in temperature encephalitis in northern Europe,9,10 although alternative
and rainfall may also affect the distribution of disease explanations—eg, changes in confounding factors like
vectors—eg, those of malaria and dengue—and the land use or in socioeconomic, demographic, and other
incidence of diarrhoeal diseases. Climate can affect levels environmental factors—remain plausible.

www.thelancet.com Vol 367 June 24, 2006 2101

 Lecture

Adverse health effects

Heat-related illnesses
and deaths

Extreme weather-related
health effects
Changes in intermediate factors
Air pollution concentration Air pollution-related
and distribution health effects
Natural and Climate Regional and
human variability and local weather
influences on change change Pollen production Allergic diseases
climate
Extreme weather Infectious diseases
Temperature Microbial contamination
and transmission Water-borne and
Precipitation food-borne diseases
Vector-borne and rodent-
borne diseases

Crop yield Malnutrition

Mitigation
policies
Change in sea level Coastal flooding Storm surge-related
Coastal aquifer salinity drowning and injuries
Health problems of
displaced populations

Modulating
influences
and adaptation
measures

Mitigation policies for reduction of Modulating influences Adaptation measures

greenhouse gas emissions Population density and growth Vaccination programmes
Energy efficiency Level of technological development Disease surveillance
Use of renewable energy sources Standard of living and local environmental conditions Protective technologies
Forest preservation and replanting Pre-existing health status Weather forecasting and warning systems
Quality of and access to health care Emergency management and disaster preparedness
Public-health infrastructure Public-health education and prevention
Legislation and administration

Figure 1: Potential health effects of climate variability and change

Reproduced from reference 1, with permission from the American Medical Association.

There is some evidence for changes in the frequency Where health surveillance data are available for several
of weather extremes over recent decades.6 Many health decades up to the present day, it may be possible to
outcomes are sensitive to isolated extreme events (eg, determine whether any observed changes in disease might
heavy rainfall, high temperatures). Analyses of the 2003 be related to changes in climate. Interpretation is
heat wave in Europe have concluded that it was a truly complicated by potential competing explanations due to
extreme event and the summer of 2003 was probably the changes in important health determinants over time, as
hottest in Europe since 1500.11 Climatologists now well as changes in the way in which diagnoses may be
consider it “very likely” that human influence on the recorded. Empirical observation of the health consequences
global climate has at least doubled the risk of a heat wave of recent climate change, followed by formulation, testing,
such as that experienced in 2003.12 Recent evidence has and then modification of hypotheses would require long
also emerged about a possible causal role of climate time-series (probably several decades) of careful
change (and specifically the warming of sea surface monitoring. Although this process may accord with the
temperatures) in increasing the intensity of tropical principles of empirical science, it would not provide the
cyclones,13,14 although a single event such as Hurricane timely information needed to inform current policy
Katrina, which caused major or catastrophic damage decisions on greenhouse gas emission abatement, so as
along the coastlines of Louisiana, Mississippi, and to offset possible health consequences in the future. Nor
Alabama in 2005, cannot be definitely attributed to climate would it allow early implementation of policies for
change. adaptation to some level of climate change, which is now

2102 www.thelancet.com Vol 367 June 24, 2006

 Lecture

inevitable owing to past greenhouse gas emissions.

Therefore, the best estimation of the future health effects
Empirical studies Predictive modelling
of climate change will necessarily come from risk
assessment based on our current understanding of the
effects of climate variation on health from observations
made in the present and recent past, acknowledging the Learn Detect Anticipate
influence of a large range of modulating factors.
Observations of short-term variations in climate or
weather show that even small temperature increases and
Past Present Future
precipitation changes can result in measurable impacts
on malaria, diarrhoeal episodes, injuries related to floods, • Heat waves • Enteric infections (seasonal Scenario-based modelling for, eg,
and malnutrition. Knowledge of these relations allows • Weather disasters pattern of food poisoning) • Cereal grain yields
• Malaria • Vector-borne disease patterns • Malaria
approximate estimates of the health effects of past and • Dengue fever • Impacts of extreme events (heat • Dengue fever
future climate change to be made. • Diarrhoeal diseases waves, floods, cyclones)

Heat waves Figure 2: Three important research paths with examples of relevant topics
Mortality rises in hot weather, especially in elderly people. Original figure by A J McMichael (National Centre for Epidemiology and Population Health, Australian National
University, Canberra, Australia). Reproduced with permission.
It is very likely that climate change will be associated with
increases in the frequency of heat waves.15 More than
2000 excess deaths were reported in England and Wales16 occur over a few days, but complete acclimatisation may
during the major heat wave that affected most of western take several years. The rate at which changes will take
Europe in 2003 (table 1).16–23 The greatest impact on place in infrastructure is likely to be much slower.
mortality occurred in France, where it was estimated that
14 800 excess deaths occurred during the first 3 weeks of Floods, droughts, and storms
August 2003 than would be expected for that time of Natural disasters have a variety of health impacts,3,26
year.18 Deaths in Paris increased by 140%.24 The sustained ranging from immediate effects of physical injury and
period of extreme high temperatures (including the morbidity and mortality through to potentially long-
minimum temperature), unique in the recorded history lasting effects on mental health. Most flood-related
of Paris, together with housing designed for cooler deaths can be attributed to rapid rise floods,27 due to the
summers, caused a major public-health crisis. increased risk of drowning. In October 1988, a flash flood
Much of the excess mortality from heat waves is related occurred in the Nimes region of France.28 Although the
to cardiovascular, cerebrovascular, and respiratory causes homes of 45 000 people were damaged and more than
and is concentrated in elderly people. A proportion of 1100 vehicles destroyed, only nine deaths by drowning
these deaths occur in susceptible people who would (including two people who tried to rescue others) and
probably have died in the near future, but there are likely three severe injuries were reported. In 1996, 86 people
to be substantial numbers of potentially preventable died from a flood in the town of Biescas in Spain as a
deaths. In the August 2003 event, the mortality patterns consequence of the stream of water and mud that
indicate that the contribution of short-term mortality suddenly covered a campsite located near a channelised
displacement was relatively small.25 river.29
Urban centres are often particularly affected because of Many slow-rise river flood events have also been
the urban heat island effect, which results in the temp- associated with fatalities. In central Europe, the Meuse,
eratures being somewhat higher than the surrounding Rhine, Elbe, and Danube rivers have flooded in recent
suburban and rural areas. Air pollution concentrations years. In 1997, river floods in central Europe left over
may also rise during heat waves and may contribute to 200 000 people homeless, and more than 100 people were
the raised death rates. The recent experience of the heat
wave in Europe demonstrated that, even in high-income
Location (date) Excess mortality (% increase) Reference
countries, such events can cause large numbers of deaths
in the absence of a coordinated response to ensure that England and Wales (Aug 4–13) 2091 deaths (17%) Johnson et al16
elderly people are kept cool and well hydrated. Italy (Jun 1–Aug 15) 3134 (15%) in all Italian capitals Conti et al17
The impact of extreme summer heat on human health France (Aug 1–20) 14 802 (60%) Anon18
may be exacerbated by increases in humidity. A central Portugal (Aug) 1854 (40%) Botelho et al19
question in estimating future heat-related and cold- Spain (Jul–Aug) 4151 deaths (11%) Simon et al20
related mortality is the rate at which populations will Switzerland (Jun–Sept) 975 deaths (6·9%) Grize et al21
adapt to a warmer climate. Populations are likely to Netherlands (Jun–Sept) 1400–2200 deaths (not reported) Garssen et al22
acclimatise to warmer climates via a range of behavioural, Germany (Aug 1–24) 1410 deaths (not reported) Sozialministerium Baden-Wuerttemberg23
physiological, and technological adaptations. The initial Table 1: Excess mortality attributed to the 2003 heat wave in Europe
physiological acclimatisation to hot environments can

www.thelancet.com Vol 367 June 24, 2006 2103

 Lecture

killed. Floods in Dresden, Germany, in 2002 left large Changes in climate that can affect the transmission of
parts of the city without power and freshwater for several vector-borne infectious diseases include temperature,
days.30 Four of the six main hospitals in Dresden were humidity, altered rainfall, soil moisture, and sea level rise.
sited near the river Elbe and affected by the flooding. Over It is a complex task to determine how these factors may
1300 deaths, approximately 2000 injuries, and more than affect the risk of vector-borne diseases. In addition to
1 million displaced people resulted from Hurricane climatic factors, the incidence and geographic distribution
Katrina, which is now the most expensive “natural” of vector-borne diseases are influenced by many
disaster in US history. demographic and societal factors. Transmission requires
In some cases flooding may lead to mobilisation of that the reservoir host, a competent vector, and the
dangerous chemicals from storage or remobilisation of pathogen be present in an area at the same time, and in
chemicals already in the environment—eg, pesticides. A adequate numbers to maintain transmission.
case study of heavy metal soil contamination after the Global climate change could cause increases or
flooding of the river Meuse during the winter of 1993–94 decreases in the overall incidence of vector-borne diseases,
concluded there was a potential health risk for river-bank and the duration of the transmission season, in particular
inhabitants as a consequence of lead and cadmium sites. Small changes in seasonality may be important,
contaminations of the flood plain soils.31 Hazards may be since transmission rates tend to increase non-linearly in
greater when industrial or agricultural land adjoining relation to the transmission season. Furthermore,
residential land is affected. However, there is insufficient increases or decreases in the geographic distribution of
research on flooding that causes chemical contamination disease transmission may occur, since climate-driven
to detect any causal effect on the pattern of morbidity and changes in vectorial capacity cause transmission to
mortality in the affected populations.32 become unsustainable in previously endemic areas, or
Following floods, increases in diarrhoeal and respiratory sustainable in previously non-endemic areas. Even small
diseases are reported in both high-income and low- increases in disease distributions may mean that new
income countries;3,33 transmission is increased where populations are exposed. New populations often lack
there is crowding of displaced populations. In acquired immunity, which can result in more serious
industrialised countries, although infections are much clinical disease.
less of a problem, the impact on the local economy may There is now a substantial body of literature on the
still be severe and increases in common mental disorders association between the El Niño cycle, a major determinant
such as anxiety and depression are common. These of global weather patterns, and some infectious diseases.
increases are probably related to damage to the home For example, there is reasonably strong evidence for an
environment and economic losses and may persist for association with El Niño and malaria epidemics in parts
more than a year after flooding. of south Asia and South America and with cholera in
Increased vulnerability of populations in low-income coastal areas of Bangladesh.36
countries may be related to the habitation of high-risk The relations between climate and disease distribution
areas such as flood plains and coastal zones, the presence and transmission have been investigated for many vector-
of a limited public-health infrastructure, and the borne diseases (table 2), including the development of
substantial damage to local and national economies, predictive models. Predictive models can be broadly
which is proportionally much greater than in industrialised classified as “biological” (based on aggregating the effect
countries.34 Inevitably, low-income populations are also of climate on the individual components of the disease
less likely to be covered by insurance. transmission cycle) or “statistical” (derived from direct
Droughts may have wide ranging effects on health correlations between observed geographic or temporal
including on nutrition, infectious diseases, and on forest variations in climate, and associated variations in disease
fires causing air pollution, particularly in low-income incidence or distribution). Most modelling of the effects
countries. The number of people worldwide affected by of climate change has focused on malaria,37–39 but the
drought is influenced strongly by the El Niño cycle.35
Vector Major diseases
Infectious diseases
Mosquitoes Malaria, filariasis, dengue fever, yellow fever,
Transmission of many infectious disease agents is West Nile fever
sensitive to weather conditions, particularly those Sandflies Leishmaniasis
spending part of their life cycle outside the human body. Triatomines Chagas disease
Pathogens that are carried by insects are exposed to Ixodes ticks Lyme disease, tick-borne encephalitis
ambient weather. Vector-borne diseases typically exhibit Tsetse flies African trypanosomiasis
seasonal patterns in which the role of temperature and Blackflies Onchocerciasis
rainfall is well documented. Some vector-borne diseases, Snails (intermediate host) Schistosomiasis
such as malaria, also display considerable year-to-year
Table 2: Examples of vector-borne diseases likely to be sensitive to
variation in some regions that can also be partly explained
climate change
by climatic factors.36

2104 www.thelancet.com Vol 367 June 24, 2006

 Lecture

potential impact on the global distribution of dengue has decreases in precipitation are projected, particularly
also been estimated.40 The calibration and validation of around the Amazon and in Central America. The
global dynamic models is difficult because the underlying overwhelming majority of the burden of malaria currently
systems are never closed. The requisite historical data are occurs in sub-Saharan Africa. Study of the effects of
not often available with sufficient spatial coverage. The climate changes over much of the past century suggests
use of assumptions and simplifications potentially that areas showing a statistically significant trend towards
decrease the quantitative accuracy of the assessment. increasing suitability for malaria are broadly countered
Hence, research is now focused on the development of by areas showing a decrease.41 The most detailed study,
regional models that can allow for validation and adequate verified against a large database of historic malaria
prioritisation and estimation of risk. surveillance data, suggests that climate change will cause
It is likely that additional populations put at risk by a small (5–7%) increase in the population at risk in Africa,
climate change will be in low-income countries, since it is mainly through expansion into higher altitudes. The
generally assumed that more developed countries, which study indicates that climate change will also lengthen the
currently control malaria, will remain able to do so. transmission season in many areas, causing a 16–28%
Malaria in poorer countries is currently only restricted by increase in the total number of person-months of
climate factors in specific arid and highland regions. The exposure.42
ability of these countries to manage any climate-induced
increase in malaria will depend on their capacity to Estimating the global burden of disease due to
develop and sustain malaria control programmes. climate change
The effect of climate change will vary geographically. WHO has recently undertaken an exercise to estimate
Malaria transmission may decrease in many areas where the global burden of disease that could be due to climate

2100 2200
Greenhouse gas
emissions
scenarios

Time

Global climate
modelling
Generates series of maps of
predicted future climate

1·8
1·6
1·4
Health impact model
1·2
Estimates the change in relative 1·0
risk of specific diseases 0·8
0·6
0·4
0·2
0
2020s 2050s 2080s

Total DALYs/million
Subregion Malnutrition Diarrhoea Malaria Floods All causes population
Conversion to a AFR-D 293 154 178 1 626 2185·78
single health AFR-E 323 260 682 3 1267 3839·58
AMR-A 0 0 0 4 4 11·85
measure AMR-B 0 0 3 67 71 166·62
DALY (disability AMR-D 0 17 0 5 23 324·15
adjusted life year) EMR-B 0 14 0 6 20 147·57
EMR-D 313 277 112 46 748 2145·91
EUR-A 0 0 0 3 3 6·66
EUR-B 0 6 0 4 10 48·13
EUR-C 0 3 0 1 4 14·93
SEAR-B 0 28 0 6 34 117·19
SEAR-D 1918 612 0 8 2538 2080·84
WPR-A 0 0 0 1 1 8·69
WPR-B 0 89 43 37 169 111·36
World 2846 1459 1018 193 5517 925·35

Figure 3: Overview of the comparative risk assessment process for climate change and health

www.thelancet.com Vol 367 June 24, 2006 2105

 Lecture

change in terms of disability adjusted life years (DALYs) The analyses suggested that climate change will bring
lost. This measure makes it possible to take into account some health benefits, such as lower cold-related mortality
impacts that do not necessarily lead to death but cause and greater crop yields in temperate zones, but these
disability. Climate scenarios are derived from the output benefits will be greatly outweighed by increased rates of
of global climate models that are, in turn, driven by other diseases, particularly infectious diseases and
scenarios of future greenhouse gas emissions (figure 3). malnutrition in developing regions. A small proportional
The attributable burden of climate change was estimated increase in cardiovascular disease mortality attributable
in relation to three (future) climate scenarios relative to to climate extremes is likely in tropical regions, and a
the baseline climate (ie, the average climate from 1960 to small benefit in temperate regions, caused by warmer
1991) representing little or no anthropogenic climate winter temperatures. Since there is evidence that some
change. Epidemiological models were used to estimate temperature-attributable mortality represents small
the degree to which these climatic changes are likely to displacements of deaths that would occur soon in any
affect a limited series of health outcomes (malaria, case, no assessment was made of the associated increase
diarrhoeal disease, malnutrition, flood deaths, direct or decrease in disease burden. Climate change is
effects of heat and cold). These measures of proportional estimated to increase the burden of diarrhoea in regions
change can be applied to projections of the burden of each made up mainly of developing countries by approximately
of these diseases in the future to calculate the possible 2–5% in 2020. Richer countries (GDP>US$6000/year),
impacts of climate change on the overall disease burden. either now or in the future, were assumed to suffer little
The methods have been described more fully elsewhere.43 or no additional risk of diarrhoea. Much larger
To generate consistent estimates, the analysis attempted proportional changes are likely in the numbers of people
to account for current geographic variation in vulnerability killed in coastal floods (approximately a doubling in the
to climate, where not already incorporated into the former socialist economies), and inland floods (up to five
predictive models. It also attempted to account for future times greater risk in developed regions). Although the
changes in disease rates due to other factors (eg, decreasing proportional change is much larger than for other health
rates of infectious diseases due to technological advances outcomes, the baseline disease burden is much lower, so
or improving socioeconomic status), and for changes in that the aggregate effect is comparatively small.
population size and age structure (eg, potentially greater Substantial proportional changes were estimated in the
proportion of older people at higher risk of mortality risk of falciparum malaria in countries at the edge of the
related to cardiovascular disease in response to thermal current distribution. However, most of the estimated
extremes). These potential future changes can be attributable disease burden is associated with small
addressed by applying the estimates of relative risks under proportional increases in regions that already suffer
alternative climate change scenarios to the global burden heavily from malaria, principally through extensions in
of disease projections of disease rates, population size, the altitudinal and latitudinal range in Africa.
and age structure. These alternative scenarios attempt to On aggregate, it was estimated that climate change
take into account the effects of changing gross domestic may already (by 2000) be causing in the region of
product (GDP), “human capital” (as measured by average 150 000 deaths (0·3% of global deaths per year) and
years of female education), and time (to account for trends 5·5 million lost DALYs/year (0·4% of global DALYs lost
such as technological development)44 on the overall per year).45 Even taking into account increasing wealth
“envelope” of cause-specific mortality and morbidity for and some level of behavioural and socioeconomic
diseases affected by climate change. The assumptions adaptation, the disease burden caused by climate change
made about future adaptation and vulnerability are is likely to increase substantially over time. Overall, the
outlined in table 3. effects are predicted to be heavily concentrated in poorer

Biological* adaptation affecting relative risks Socioeconomic adaptation affecting relative risks

Direct effects of heat and cold Yes. Temperature associated with lowest mortality was None
assumed to change directly with temperature increases
driven by climate change
Diarrhoea None Assumed RR=1 if GDP per capita rises above US$6000/year
Malnutrition None Food-trade model assumed future increases in crop yields from
technological advances, increased liberalisation of trade, and increased
GDP
Disasters: coastal floods None Model assumed the relative risk of deaths in floods decreases with GDP
Disasters: inland floods and landslides None Model assumed the RR of deaths in floods decreases with GDP
Vector-borne diseases: malaria None None (for RR)

GDP=gross domestic product; RR=relative risk. *Physiological, immunological, and behavioural.

Table 3: Assumptions on adaptation and vulnerability for each health outcome, as applied in the WHO global burden of disease exercise

2106 www.thelancet.com Vol 367 June 24, 2006

 Lecture

Health outcome Public health Surveillance

Mortality and morbidity due to heat waves Public-health education Enhance health surveillance of routine data for early detection of heat wave effects
Heat health warning systems (eg, monitoring from funeral homes, calls to NHS Direct)
Emergency preparedness
Floods Public-health education—eg, boil water notices Surveillance for flood effects, with long-term follow-up
Emergency preparedness Coordinated national surveillance for flood deaths, injuries, and illnesses
Check list for post-flood activities
Air quality Warnings for high pollution days Daily air pollution measurements
Vector-borne diseases Public education, especially to avoid contact with ticks Monitoring of vectors and reservoir host
Integrated surveillance for human and animal diseases
Food-borne disease Maintenance and strengthening of food hygiene measures Integrated surveillance for human and animal diseases
Water-borne diseases Risk assessment for extreme rainfall events Increased microbiological monitoring of public water supplies and private wells, and
Risk assessment of health effects of algal blooms enhanced surveillance during and following heavy rainfall events

Table 4: Summary of public-health adaptation measures in relation to the health impacts of climate change (applicable to European populations)46,47

populations at low latitudes, where the most important demographic groups; (2) how to exploit opportunities, as
climate-sensitive health outcomes (malnutrition, well as to reduce risks; (3) the cost of adaptation and the
diarrhoea, and malaria) are already common, and where potential to exacerbate climate change (eg, extensive use
vulnerability to climate effects is greatest. These diseases of air conditioning); (4) the need for adaptation to multiple
mainly affect younger age groups, so that the total burden factors, including climate change; (5) the systemic nature
of disease due to climate change appears to be borne of climate impacts that means that many sectors will need
mainly by children in developing countries. to be involved to reduce public-health impacts; and (6)
Considerable uncertainties surround these estimates maladaptation—ie, policies that increase vulnerability to
and the range of impacts included is not comprehensive. climate change can result in serious negative effects.
There is, for example, uncertainty over future climate Some examples of public-health adaptation strategies to
change (particularly future greenhouse gas emissions), climate variability and change are given in table 4.
uncertainty about climate/health relations, and most
importantly, uncertainties around the degree to which Mitigation strategies and health
current climate/health relations will be modified by Climate change poses a major threat to sustainable
socioeconomic adaptation in the future. These development because adverse effects are likely to be
uncertainties could be reduced in subsequent studies by directed particularly at poor populations that currently
(1) applying projections from a range of climate models also suffer disproportionately from a lack of reliable
and/or their probability distributions, (2) relating climate energy at the level of the household and the community.
and disease data from a wider range of climatic and The easy availability of cheap energy from fossil fuels has
socioeconomic environments, (3) more careful validation underpinned the economic development of industrialised
against patterns in the present or recent past, and countries and has therefore contributed substantially to
(4) more detailed longitudinal studies of the interaction the dramatic advances in health observed over the past
of climatic and non-climatic influences on health. century or so. With current energy sources, recent and
continuing patterns of economic development contribute
Climate change and public health more to climate change than population growth.48 During
The current state of knowledge about climate change is the 20th century the world population grew almost
such that some specific measures for health protection fourfold, at the same time emissions of carbon dioxide
can now be recommended. The summer of 2003 grew around 12-fold. Population growth in low-income
illustrated a lack of public-health capacity in Europe to and middle-income countries poses major challenges for
deal with heat waves. A recent WHO publication greenhouse gas emissions in the future if economic
encourages public-health decision makers to act now to growth is based on fossil fuel use.
address climate hazards, as well as address adaptation Although the Kyoto protocol is an important political
strategies in the longer term.46 Although there is initiative to engage countries in developing policies to
uncertainty about future climate change, failure to invest reduce greenhouse gas emissions, the modest targets in
in adaptations may leave a nation poorly prepared to cope the protocol would not have much impact on some of the
with adverse changes and increase the probability of major adverse impacts.49 For example, to keep the
severe consequences.47 concentration of carbon dioxide from exceeding the
A number of important factors related to the design and doubling of the pre-industrial concentration of 275 parts
implementation of strategies must be considered in the per million, reductions of more than two-thirds in
assessment of health policies, measures, and strategies, emissions would be needed, assuming a population of
including: (1) the variation of appropriateness and 9 billion by 2050.50 The industrialised nations, which have
effectiveness of adaptation options by region and across benefited so much from fossil fuels, should take the lead

www.thelancet.com Vol 367 June 24, 2006 2107

 Lecture

and their proportional reductions will need to be much Sources of renewable energy such as photovoltaic, solar
greater than the less developed nations to converge on a thermal, wave, and wind power do not appear to have any
much lower level and more equal distribution of important adverse effects on health and their overall
emissions. impacts are likely to be overwhelmingly beneficial.54 The
Approximately 2 billion people lack access to electricity barriers to their uptake relate particularly to the cost of
and suffer substantial ill health as a result. Around half electricity generated in these ways. However, there is
the global population cook daily with traditional biomass substantial evidence that dams for the generation of
fuels (eg, dung, crop residues, wood, and charcoal), hydropower may have adverse effects, for example, by
resulting in exposure to very high concentrations of air affecting the distribution of vector-borne diseases and
pollutants indoors and extensive time spent in collection displacing populations.55,56 The health impact assessment
of wood or other fuel and the attendant opportunity costs, of dams is therefore an important aspect of the planning
particularly for women. Improved energy efficiency cook process. The assessment of the impacts of the expansion
stoves are becoming increasingly available in a number of of nuclear power is complex and beyond the scope of this
countries and can substantially cut the use of biomass article. Increasing costs and concerns about the security
fuels with subsequent health, environmental, and of fossil fuels provide added impetus to seek alternatives.
economic benefits.51 These populations would also
obviously benefit from access to affordable electricity. Conclusions
A WHO publication has demonstrated the potential The effects of climate change on health are likely to be
near-term benefits to health of strategies to reduce predominately negative and impact most heavily on low-
greenhouse gas emissions applied in China.52 The authors income countries where capacity to adapt is weakest, but
concluded that the benefits to human health arising from also on the most vulnerable groups in developed
changes in energy use in the housing sector are many countries. Adaptation strategies should blunt some of
times larger than those in the electric power sector. the adverse impacts but will pose difficulties of
Economic benefits of reducing exposure to indoor air implementation, particularly in low-income countries.
pollution were thought to be substantially larger than the With climate change already underway, there is a need to
cost of reducing greenhouse gas emissions, particularly assess vulnerabilities and identify cost-effective inter-
when this was achieved by improving energy efficiency. vention/adaptation options in the health sector and in
Fossil fuel combustion is a cause of both local air other sectors that have direct links to human health.
pollutants (especially particulates, ozone, methane, Early planning can help reduce future adverse health
nitrogen oxides, and sulphur dioxide) and greenhouse impacts and mitigation strategies—eg, using a number
gases. Policies that aim to address global anthropogenic of renewable energy sources—can improve health by
climate change can therefore also benefit health in the reducing air pollution as well as addressing climate
near term by reducing the concentration of urban air change.
pollutants. A recent paper demonstrates the potential Conflict of interest statement
benefits of converting all US on-road vehicles to AH is a reviewer and RSK a participant in the UN Intergovernmental
hydrogen fuel-cell vehicles.53 Such vehicles powered by Panel on Climate Change for the Fourth Assessment Report. DC-L and
CC declare that they have no conflict of interest.
hydrogen from renewable energy sources (eg, wind
power) could save 3700–6400 lives annually from Acknowledgments
This article is based on a paper presented to the World Climate
reduced air pollution as well as benefiting climate Change Conference (Moscow, Russian Federation, Sept 29–Oct 3,
change. “Ancillary benefits” are the monetised secondary 2003), which has been updated for the 2005 Harben Lecture of the
(or side) benefits of mitigation policies on problems Royal Institute of Public Health. We thank Tony McMichael for
such as reductions in local air pollution associated with permission to use figure 2 and acknowledge the contributions of
many scientists to the Comparative quantification of health risks:
the reduction of fossil fuels. Multiple, wider health global and regional burden of disease due to selected major risk
ancillary benefits of mitigation are possible by improving factors report. The views expressed in this article are those of the
transport policies in both developed and developing authors and do not necessarily reflect the position of the WHO.
countries. Transport is projected to have the fastest References
proportional growth in greenhouse gas emissions of 1 Haines A, Patz J. Health effects of climate change. JAMA 2004; 291:
99–103.
any sector from 1990–2020, and there are direct 2 McMichael AJ, Woodruff RE, Hales S. Climate change and human
connections with urban air pollution (around health: present and future risks. Lancet 2006; 367: 859–69.
800 000 deaths per year globally), road traffic accidents 3 Ahern M, Kovats RS, Wilkinson P, Few R, Matthies F. Global health
(1·2 million deaths per year), and physical inactivity impacts of floods: epidemiologic evidence. Epidemiol Rev 2005; 27:
36–46.
(1·9 million deaths a year).45 There are therefore 4 Langner J, Bergstom R, Foltescu V. Impact of climate change on
potentially major synergies in terms of reduced surface ozone and deposition of sulphur and nitrogen in Europe.
greenhouse gas emissions and direct health benefits Atmos Environ 2005; 39: 1129–41.
5 Kovats RS, Campbell-Lendrum D, McMichael AJ, Woodward A,
from sustainable transport systems that make more use Cox J. Early effects of climate change: do they include changes in
of public transport, walking, and cycling, especially in vector-borne disease? Philos Trans R Soc Lond B Biol Sci 2001; 356:
rapidly developing countries such as China and India. 1057–68.

2108 www.thelancet.com Vol 367 June 24, 2006

 Lecture

6 Intergovernmental Panel on Climate Change. Climate change 2001: 31 Albering HJ, van Leusen SM, Moonen EJ, Hoogewerff JA,
impacts, adaptation and vulnerability. Contribution of Working Kleinjans JC. Human health risk assessment: a case study involving
Group I to the third assessment report of the Intergovernmental heavy metal soil contamination after the flooding of the river Meuse
Panel on Climate Change. Cambridge: Cambridge University Press, during the winter of 1993–1994. Environ Health Perspect 1999; 107:
2001. 37–43.
7 Kovats RS, Campbell-Lendrum D, Matthies F. Climate change and 32 Euripidou E, Murray V. Public health impacts of floods and
human health: estimating avoidable deaths. Risk Anal 2005; 25: chemical contamination. J Public Health 2004; 26: 376–83.
1409–18. 33 Siddique A, Baqui A, Eusof A, Zamon K. 1988 flood in Bangladesh:
8 Purse BV, Mellor PS, Rogers DJ, Samuel AR, Mertens PP, Baylis M. pattern of illness and cause of death. J Diarrhoeal Dis Res 1991; 9:
Climate change and the recent emergence of bluetongue in Europe. 310–14.
Nat Rev Microbiol 2005; 3: 171–81. 34 Few R, Matthies F, eds. Flood hazards and health: responding to
9 Lindgren E, Tälleklint L, Polfeldt T. Impact of climatic change on present and future risks. London: Earthscan, in press.
the northern latitude limit and population density of the disease 35 Bouma MJ, Kovats SR, Goubet SA, St H Cox J, Haines A. Global
transmitting European tick Ixodes ricinus. Environ Health Perspect assessment of El Niño’s disaster burden. Lancet 1997; 350: 1435–38.
2000; 108: 119–23. 36 Kovats RS, Bouma MJ, Hajat S, Worrell E, Haines A. El Niño and
10 Skarphedinsson S, Jensen PM, Kristiansen K. Survey of tick borne health. Lancet 2003; 361: 1481–89.
infections in Denmark. Emerg Infect Dis 2005; 117: 1055–61. 37 Martin PH, Lefebvre MG. Malaria and climate—sensitivity of
11 Luterbacher J, Dietrich D, Xoplaki E, Grosjean M, Wanner H. malaria potential transmission to climate. Ambio 1995; 24: 200–07.
European seasonal and annual temperature variability, trends and 38 Rogers DJ, Randolph SE. The global spread of malaria in a future,
extremes since 1500. Science 2004; 303: 1499–1503. warmer world. Science 2000; 289: 1763–66.
12 Stott PA, Stone DA, Allen MR. Human contribution to the 39 Thomas CJ, Davies G, Dunn CE. Mixed picture for changes in
European heatwave of 2003. Nature 2004; 432: 610–14. stable malaria distribution with future climate in Africa.
13 Emanuel K. Increasing destructiveness of tropical cyclones over the Trends Parasitol 2004; 20: 216–20.
past 30 years. Nature 2005; 436: 686–88. 40 Hales S, de Wet N, Macdonald J, Woodward A. Potential effect of
14 Webster PJ, Holland GJ, Curry JA, Chang HR. Changes in tropical population and climate changes on global distribution of dengue
cyclone number, duration, and intensity in a warming environment. fever: an empirical model. Lancet 2002; 360: 830–34.
Science 2005; 309: 1844–46. 41 Small J, Goetz SJ, Hay SI. Climatic suitability for malaria
15 Hulme M, Jenkins GJ, Lu X, et al. Climate change scenarios for the transmission in Africa, 1911–1995. Proc Natl Acad Sci USA 2003;
United Kingdom: the UKCIP02 Scientific Report. Norwich, UK: 100: 15341–45.
Tyndall Centre for Climate Change Research, School of 42 Tanser FC, Sharp B, le Sueur D. Potential effect of climate change
Environmental Sciences, University of East Anglia, 2002. on malaria transmission in Africa. Lancet 2003; 362: 1792–98.
16 Johnson H, Kovats RS, McGregor G, et al. The impact of the 2003 43 McMichael AJ, Campbell-Lendrum D, Kovats RS, et al. Global
heatwave on mortality and hospital admissions in England. climate change. In: Ezzati M, Lopez AD, Rodgers A, Murray CJ, eds.
Health Stat Q 2005; 25: 6–11. Comparative quantification of health risks: global and regional
17 Conti S, Meli P, Minelli G, et al. Epidemiologic study of mortality burden of disease due to selected major risk factors, vol 2. Geneva:
during the summer 2003 heat wave in Italy. Environ Res 2005; 98: WHO, 2004: 1543–649.
390–99. 44 Murray CJL, Lopez AD, eds. The global burden of disease: a
18 Anon. Impact sanitaire de la vague de chaleur en France survenue comprehensive assessment of mortality and disability from
en août 2003. Paris: Institut de Veille Sanitaire, 2003. diseases, injuries and risk factors in 1990 and projected to 2020.
19 Botelho J, Caterino J, Calado R, Nogueira PJ, Paixao E, Falcao JM. Cambridge, MA, USA: Harvard School of Public Health/WHO,
Onda de calor de Agosto de 2003. Os seus efeitos sobre a 1996.
mortalidade da populacao portuguesa. Lisbon: Instituto Nacional de 45 WHO. The world health report 2002: reducing risks and promoting
Saude Dr. Ricardo Jorge, 2005. healthy life. Geneva: WHO, 2002.
20 Simon F, Lopez-Abente G, Ballester E, Martinez F. Mortality in 46 Menne B, Ebi KL, eds. Climate change and adaptation strategies for
Spain during the heat waves of summer 2003. Euro Surveill 2005; human health Hessen, Germany, Dr Dietrich Steinkopff Verlag
10: 156–60. GmbH & Co. KG, 2005.
21 Grize L, Hussa A, Thommena O, Schindlera C, 47 Smith JB, Lenhart SS. Climate change adaptation policy options.
Braun-Fahrlandera C. Heat wave 2003 and mortality in Switzerland. Climate Research 1996; 6: 193–201.
Swiss Med Wkly 2005; 135: 200–05. 48 McMichael AJ, Powles JW. Human numbers, environment,
22 Garssen J, Harmsen C, de Beer J. The effect of the summer 2003 sustainability, and health. BMJ 1999; 319: 977–80.
heat wave on mortality in the Netherlands. Euro Surveill 2005; 10: 49 Parry ML, Arnell NW, Hulme M, Nicholls RJ, Livermore MT.
165–67. Adapting to the inevitable. Nature 1998; 395: 741.
23 Sozialministerium Baden-Wuerttemberg. Gesundheitliche 50 Wigley TML, Richels R, Edmonds JA. Economic and environmental
Auswirkungen der Hitzewelle im August 2003. Freiburg, Germany: choices in the stabilization of atmospheric CO2 concentrations.
Sozialministerium Baden-Wuerttemberg, 2004 Nature 1996; 379: 240–43.
24 Vandentorren S, Suzan F, Medina S, et al. Mortality in 13 French 51 Kammen DM. Cookstoves for the developing world. Sci Am 1995;
cities during the August 2003 heat wave. Am J Public Health 2004; 273: 72–75.
94: 1518–20.
52 Wang X, Smith KR. Near-term health benefits of greenhouse gas
25 Le Tertre A, Lefranc A, Eilstein, D, et al. Impact of 2003 heat wave on reductions: a proposed method for assessment and application in
all cause mortality in 9 French cities. Epidemiology 2006; 17: 75–79. two energy sectors of China. Geneva: WHO, 1999. WHO/SDE//
26 Noji EK. The public health consequences of disasters. New York: PHE/99.01.
Oxford University Press, 1997. 53 Jacobson MZ, Colella WG, Golden DM. Cleaning the air and
27 French J, Ing R, Von Allmen S, Wood R. Mortality from flash improving health with hydrogen fuel-cell vehicles. Science 2005;
floods: a review of national weather service reports, 1969–81. 308: 1901–05.
Public Health Rep 1983; 98: 584–88. 54 Haines A, Kammen D. Sustainable energy and health.
28 Duclos P, Vidonne O, Beuf P, Perray P, Stoebner A. Flash flood Global Change and Human Health 2000; 1: 2–11.
disaster—Nimes, France, 1988. Eur J Epidemiol 1991; 7: 365–71. 55 Diop M, Jobin WR. Senegal River Basin Health Master Plan Study.
29 Estrela T, Menéndez M, Dimas M, et al. Sustainable water use in Water and Sanitation for Health (WASH) Field Report number 453.
Europe – part 3: extreme hydrological events: floods and droughts. Washington, DC: US Agency for International Development, 1994.
Copenhagen: European Environment Agency, 2001. 56 Ghebreyesus TA, Haile M, Witten KH, et al. Incidence of malaria
30 Meusel D, Kirch W. Lessons learned from the 2002 flood in among children living near dams in northern Ethiopia: community
Dresden, German. In: Menne B, Bertollini R, Kirch W, eds. Extreme based incidence survey. BMJ 1999; 319: 663–66.
weather events and public health response. Berlin: Springer-Verlag,
2005: 175–84.

www.thelancet.com Vol 367 June 24, 2006 2109