Climate and human health: synthesizing environmental complexity ...

2 downloads 88 Views 221KB Size Report
Apr 17, 2007 - We summarize the principal climate variables and climate-dependent processes that are believed to impact human health across a ...
Stoch Environ Res Risk Assess (2007) 21:601–613 DOI 10.1007/s00477-007-0142-1

ORIGINAL PAPER

Climate and human health: synthesizing environmental complexity and uncertainty James D. Tamerius Æ Erika K. Wise Æ Christoper K. Uejio Æ Amy L. McCoy Æ Andrew C. Comrie

Published online: 17 April 2007 Ó Springer-Verlag 2007

Abstract Broad relationships between weather and human health have long been recognized, and there is currently a large body of research examining the impacts of climate change on human health. Much of the literature in this area examines climate–health relationships at global or regional levels, incorporating mostly generalized responses of pathogens and vectors to broad changes in climate. Far less research has been done to understand the direct and indirect climate-mediated processes involved at finer scales. Thus, some studies simplify the role of climate and may over- or under-estimate the potential response, while others have begun to highlight the subtle and complex role for climate that is contingent on other relevant processes occurring in natural and social environments. These fundamental processes need to be understood to determine the effects of past, current and future climate variation and change on human health. We summarize the principal climate variables and climate-dependent processes that are believed to impact human health across a representative set of diseases, along with key uncertainties in these relationships. Keywords Health

Climate  Disease  Ecology  Environment 

J. D. Tamerius  E. K. Wise  C. K. Uejio  A. C. Comrie (&) Department of Geography and Regional Development, University of Arizona, Harvill Building, Box #2, Tucson, AZ 85721, USA e-mail: [email protected] A. L. McCoy Arid Lands Resource Sciences, University of Arizona, 1955 East Sixth Street, PO Box 210184, Tucson, AZ 85719, USA

1 Introduction Empirically-based ecological and public health studies demonstrate that climatic variables are engaged in numerous processes that affect human health risks. However, these climate-mediated processes are of variable importance in understanding and predicting human health. Health risks are modulated by physical, geographical, biological, genetic, social, cultural, and political environments and are continuously shaping and being shaped by each other. The complex interaction between hosts, reservoirs, vectors and agents of disease results in the dispersion, emergence, reproduction, and persistence of health risks. Much of the prior literature examines climate–health links at global and national scales. Relatively less attention has been paid to outlining direct and indirect climate–health processes at finer scales. Depending on their scale, some studies may therefore over- or under-estimate the potential climate–health response, while a more limited number highlight the complex and contingent role of climate in tandem with other natural and social processes. Previous reviews have examined the social inequities of climate and health impacts, regional assessments of environmentally sensitive health outcomes, and synergistic climate and health impacts on the economy (e.g. Patz et al. 2005; Kolivras and Comrie 2004; Epstein and Mills 2005). Multiple disciplinary perspectives of health, society, and the environment are required to understand human and disease ecology and manage health risk. Adoption of Geographic Information Science perspectives and methods is widespread and growing in public health, veterinary medicine, and environmental modeling (e.g., Elliott 2000; Cromley and McLafferty 2002). Other geographical ways of knowing such as political ecology, critical geography, and physical geography challenge and re-conceptualize social,

123

602

environmental, and place-based relationships (e.g. Mayer 2000; Del Casino 2004; Smoyer 1998; Haggett 1994). This review focuses on the individual mechanisms that link climate to disease, revealing the high degree of physical and social complexity and ultimately demonstrating the uncertainty involved in developing future disease scenarios based on coarse spatio-temporal predictions of climate. Sources for this review were selected based on keyword searches from multiple electronic databases and follow-on references. Rather than strict systematic inclusion, the papers selected were the most instructive to achieve our goal of reviewing the chief climate variables and climate-dependent processes that are currently understood to affect human health across a representative set of diseases, along with key uncertainties in these relationships. We highlight established links between climate and human health risks using general mechanisms associated with transmission of the disease: airborne (non-infectious), airborne (infectious), rodentborne, insectborne, waterborne, and weather. These groupings encompass the majority of climate-mediated health risks. For each group an individual representative risk was chosen to exemplify the mechanisms linking climate with human health risks of the given group. Oftentimes, risks within each group are affected by climate through similar processes and it is therefore redundant to go through each climate-mediated health risk individually. Thus, we do not aim to provide an exhaustive review of the climate and health literature, climate change and disease, or a review of the most significant climate-mediated public health risks. Rather, our goal is to illuminate these areas by illustrating the critical and most common types of climate-mediated relationships between agents, hosts, reservoirs, vectors, and human health. The review is structured around the six representative health risks mentioned above. For each, the effects of climate were classified into two groups (1) effects of climate on the prevalence or generation of the agent in the environment and (2) effects of climate on human exposure to the agent. Frequently, climate modulates an individual health risk through several mechanisms, and the mechanisms often naturally break down along these lines. The magnitude of the effect of climate is generally greater when it affects human exposure to the agent and less so when it affects the prevalence or generation of the agent. Ultimately, it is the sum of all the effects of the climate-mediated mechanisms that dictates how sensitive the associated disease is to climate variability. The temporal and spatial variability associated with climatic mechanisms for each disease is also discussed.

2 Airborne (infectious and allergic) diseases The pollen and spores of various flora including molds and fungi, often affect human health by causing allergic

123

Stoch Environ Res Risk Assess (2007) 21:601–613

responses and in some cases infectious disease. Coccidioidomycosis is an infectious disease caused by soil dwelling fungi whose connections with climate are still actively under study. 2.1 Coccidioidomycosis Coccidioidomycosis (also known as Valley fever) is a disease caused by the inhalation of spores from the soil fungi Coccidioides immitis and C. posadasii (Comrie 2005). The fungus is endemic in isolated regions of North, South, and Central America that typically experience arid to semiarid conditions, mild winter temperatures, and prolonged hot seasons (Pappagianis 1988). The disease is not transmissible from person to person, and life-long immunity is typically acquired after the initial infection (Pappagianis 1988). Although the vast majority of human cases of coccidioidomycosis are asymptomatic or mild, about 1% of the cases become severe and can be fatal (Kolivras and Comrie 2003). 2.2 Climate and agent prevalence Precipitation directly influences the prevalence of Coccidioides in the environment through several mechanisms. Initially, precipitation provides soil moisture that is necessary for the growth of the fungus (Kolivras and Comrie 2003). Although a positive correlation exists between precipitation and the incidence of disease, there is no immediate effect on health, since it may take weeks to years for fungal spores to be released and infect a mammal (Comrie 2005; Kolivras and Comrie 2003). However, the increase in incidence associated with increased precipitation has been demonstrated to persist for 1.5–2 years (Comrie 2005). In regions of the southwest US, precipitation during the arid fore-summer (April–June) has been demonstrated to have an amplified effect on the incidence of subsequent Valley fever incidence (Comrie 2005). The fore-summer is typically the driest and warmest part of the year in this region, which leads to seasonal soil desiccation and vegetation dormancy. Consequently, precipitation during this period has been hypothesized to give Coccidioides a competitive advantage over other biota (Comrie 2005). 2.3 Climate and human exposure Although precipitation is necessary for the growth of the fungus in soil, the absence of precipitation plays a pivotal and direct role in the exposure of hosts to fungal spores. Prolonged dry conditions cause the fungus to develop spores and release them into the soil (Kolivras and Comrie 2003; Comrie 2005). Dry soil in conjunction with wind and other physical processes increases the likelihood that the

Stoch Environ Res Risk Assess (2007) 21:601–613

spores will become airborne, resulting in a greater incidence of Valley fever (Kolivras and Comrie 2003). Consequently, it has been demonstrated that the annual incidence of coccidioidomycosis is correlated with ambient particulate matter (PM) concentrations (Comrie 2005). Precipitation can cause a rapid reduction in the incidence rate of Valley fever due to its dust-inhibiting properties (Kolivras and Comrie 2003). 2.4 Temporal and spatial variability Since climate variables, particularly precipitation, are essential components in the disease dynamics of coccidioidomycosis, incidence rates exhibit distinct seasonal and inter-annual patterns that are correlated with climate variability. Maximum incidence rates of infection occur during or following periods of high dust concentration, which typically coincide with dry weather, and lower rates occur during wetter periods with reduced dust concentrations. Consequently, a bimodal annual incidence pattern occurs in the southwest US, where two distinct dry and wet seasons are experienced (Comrie 2005). Although increased precipitation is positively correlated with disease in subsequent years, a long-term drought beginning in the mid 1990s in the southwest US may be at least partially responsible for an epidemic that has occurred over that same period (Comrie 2005). This suggests that although precipitation is important for the prevalence of the fungus, there are sufficient amounts of Coccidioides spores in the environment during periods of drought to cause epidemics. 2.5 Future research Statistics on the incidence of infection is the only proxy that has been used to understand the spatial variability of Coccidioides in the environment. However, inconsistencies in diagnosis and reporting of Valley fever, as well as the wind’s ability to disperse the spores large distances cause this to be an imperfect method for calculating the true spatial variability of the fungus in the soil (Comrie 2005). Understanding the true spatial distribution of the fungus would assist in further refining the environmental variables that contribute to the propagation of the fungus in the environment and ultimately affect human health.

3 Rodentborne diseases Rodents have long been known to be carriers of pathogens that cause disease in humans. The most famous example is the Bubonic Plague that killed an estimated 25 million people in a 5-year stretch in the middle of the fourteenth

603

century. However, many other lesser known pathogens carried by rodents exist that are known to cause human disease such as lassa fever, lyssavirus, and hantavirus. 3.1 Hantavirus—Sin Nombre virus The Sin Nombre virus (SNV) is a particularly dangerous serotype of hantavirus that was identified in the southwest US during an outbreak of Hantavirus pulmonary syndrome (HPS) in 1993. Although this particular strain of the virus had not been identified prior to the initial outbreak, it is now known that it has been endemic in regions of the United States for a substantial length of time (Yates et al. 2002). The primary reservoir of SNV is the deer mouse (Peromyscus maniculatus), which inhabits much of the western US. Humans are primarily exposed to SNV when excretions from infected deer mice become airborne and are inhaled. 3.2 Climate and agent prevalence Although the basic mechanisms that relate climate and HPS risk have been determined, they have yet to be refined. The inability of studies to pinpoint the precise effects of climate on HPS incidence may be due to the indirect nature of the relationships. Generally, above-average precipitation from fall through spring has been demonstrated to be positively correlated to HPS cases (Parmenter et al. 1993; Engelthaler et al. 1999). This observation led to the ‘‘trophic cascade’’ theory, which suggested that increased precipitation caused an abundance of new vegetation (i.e. food for P. maniculatus), resulting in an increase in reservoir populations and increased risk of exposure (Engelthaler et al. 1999; Hjelle and Glass 2000; Parmenter et al. 1993). However, using remote sensing techniques to examine this process, Glass et al. (2000) demonstrated that the relationships are not that straightforward. Although areas exhibiting high levels of vegetation growth correspond to regions of high HPS risk (Glass et al. 2000; Yates 2002), the inverse was not true: regions of high HPS risk did not always exhibit high rates of vegetation growth (Glass et al. 2000). Instead, the combination of soil moisture, soil type and vegetation structure were identified as being the most accurate proxy for HPS risk (Glass et al. 2000). Long-term ecological studies on population variability of P. maniculatus are offering new clues about the possible influences of climate on the host. For instance, although total populations are typically at a minimum during the early spring, prevalence of SNV within populations tends to be greatest during this time; this is believed to be the result of communal nesting during the winter months

123

604

(Mills et al. 1999). Consequently, the infected population is highly correlated with total population numbers during the antecedent year (Mills et al. 1999; Yates et al. 2002). This is important, since it is the infected population, not the total population that is highly correlated with incidence of HPS (Mills et al. 1999; Yates et al. 2002). It has also been hypothesized that mild conditions during the winter may increase survival rates and result in greater HPS risk during the subsequent year. 3.3 Climate and human exposure Although its indirect impacts are strong, climate has little direct impact on exposure to SNV, since the virus is extremely volatile and is only infectious for minutes in the most favorable conditions. As a result, climate has no known or suspected effect on the physical mechanisms that expose humans to SNV. However, seasonal activities such as planting, harvesting, and ‘‘spring-cleaning’’ may increase the likelihood of exposure to the virus (Mills et al. 1999). 3.4 Temporal and spatial variability As of September 2006, 453 cases of HPS have been identified in the United States from Washington to Florida, although an overwhelming majority of these cases have occurred in the western US (http://www.cdc.gov/ncidod/ hanta). HPS cases have also been identified in parts of South and Central America, and numerous less severe forms of hantavirus that have similar ecologies are found worldwide. In the southwest US, a vast majority of the cases occur at elevations greater than 1,800 m, although cases also occur in basins (Engelthaler et al. 1999). The cases occurring in the valley basins frequently occur during the winter and early spring, whereas in the higher elevations they tend to occur in late spring and early summer. Short-term local climate variability appears to cause substantial fluctuations in the incidence of HPS in the southwest US. Above-average precipitation during the 1991–1992 winter and spring accompanying an El Nin˜o event is believed to a have triggered the initial outbreak of HPS in the United States (Engelthaler et al. 1999). This outbreak may have lasted more than 2 years before incidence rates decreased to their baseline levels (Engelthaler et al. 1999). 3.5 Future research Several uncertainties have made it difficult to accurately asses HPS risk. Observations have demonstrated that finescale spatial distributions of P. maniculatus can be be very inconsistent, and ‘‘environmental islands’’ may exist within larger areas of otherwise unsuitable environments (Glass et al. 2000). Furthermore, the density of infected

123

Stoch Environ Res Risk Assess (2007) 21:601–613

hosts also appears to be modulated by biological characteristics of the initial population in combination with timing of favorable conditions (Yates et al. 2002). It is also likely that infected rodents shed the virus at a much higher rate at the beginning of their infection, increasing the risk of exposure to humans during this period. Therefore, although the general principles of the trophic cascade hypothesis are probably correct, other factors, including the subtleties of host population dynamics, variability of shedding of the virus by the host, vegetation structure, timing of environmental events, and microclimates, play a pivotal role in the prevalence of the disease and need to be understood further to improve predictions of disease risk.

4 Insectborne disease Insectborne agents with the largest mortality and morbidity burden are either parasitic protozoans (i.e, malaria, leishmaniasis), viruses (i.e., West Nile Virus, dengue fever), or parasitic helminths (i.e., lymphatic filariasis, onchocerciasis—river blindness). Many agents only infect people when humans enter a disease transmission cycle between an insect and another vertebrate disease host (zoonosis). These agents cannot reproduce to a high enough level in humans to facilitate infection of another vector. This is not the case for malaria, dengue, and most lymphatic filariasis agents where people are the primary hosts and a vector transfers the agent between humans. Multiple agents have both a human adapted and a non-human and insect transmission cycle (i.e., leishmaniasis, African sleeping sickness) or an agent efficiently adapted to multiple vertebrate hosts such as yellow fever. 4.1 Malaria Plasmodium falciparum in sub-Saharan Africa causes a large fraction of the 0.7 to 2.7 million human malarial deaths per year (http://www.cdc.gov/malaria). Significant numbers of malaria cases from P. vivax, ovale, and malariae are recorded in tropical and subtropical locations across Asia, South, and Central America. The parasite is transmitted to female Anopheles mosquito vectors feeding on the blood of infected humans and can be transmitted to a susceptible human after maturation within the mosquito. The parasite develops and sexually reproduces in mosquitoes and propagates asexually in humans. 4.2 Climate and agent propagation The prevalence of malaria is modulated by the population dynamics and interactions between humans, mosquitoes and the Plasmodium parasite. Temperature, humidity, and

Stoch Environ Res Risk Assess (2007) 21:601–613

precipitation are historically recognized for influencing the life-cycle of the causative parasite and disease vector (Swaroop 1946; Pampana 1969). A primary step in the sequence of disease is the generation of larval Anopheles breeding sites near residential areas which consist of turbid water bodies generally created by human activities (burrow pits, drainage channels, tire tracks, etc.) and filled with water (http://www.cdc.gov/malaria; Mutuku et al. 2006). Depending on the local geography, precipitation events may activate or sustain Anopheles breeding habitats and potentially increase adult mosquito abundance. The Anopheles’s rapid development and high fecundity allow it to develop in novel habitats and temporarily increase its adult population size. Conversely, extreme or persistent precipitation may disrupt or extinguish Anopheles in their aquatic life stages. The development rate of the coldblooded mosquito is a complex function of resource availability, mosquito density, and temperature. Between 18 and 32°C, members of the Anopheles gambiae complex (henceforth An. gambiae) have development-times that non-linearly decrease with increasing average temperatures in a laboratory experiment (Bayoh and Lindsay 2003). Complete metamorphosis from a fertilized egg to adult therefore takes 9–12 days (www.cdc.gov/malaria). 4.3 Climate and exposure Successful Plasmodium transmission to a susceptible human host can be conceptualized as being dependent upon Plasmodium development, infectious mosquito longevity, biting frequency, and host preference. For example, if an adult Anopheles ingests a malarial parasite, the parasite must replicate in the mosquito’s midgut and disseminate to her mouthparts before another susceptible human can be infected. Generally, this extrinsic incubation period takes 9–21 days at average temperatures of 17–32°C (Patz and Olson 2006). Notably, optimal average temperatures for both An. gambiae and Plasmodium development are approximately 21–27°C. Thus, mosquitoes are cold-blooded and Plasmodium development is strongly influenced by ambient air temperatures. Anopheles similarly tend to live longer at average temperatures between 20 and 25°C and relative humidities above 60% (Martens et al. 1995). Anopheles human biting rates are strongly associated with soil moisture in the preceding 2–4 weeks, temperature, and concurrent monthly satellite derived measurements of vegetation health and density (Patz et al. 1998). 4.4 Temporal and spatial variability Competition, dispersal ability, genetic variability, and the physical environment limit the geographic extent of primary malaria vectors (Lindsay and Martens 1998; Coetzee

605

et al. 2000). For example, An. gambiae inhabits areas with 30–320 cm of precipitation, maximum average temperatures between 25 and 42°C and minimum temperatures between 5 and 22°C (Lindsay et al. 1998). Potential shifts in the geographic range of Anopheles in response to multiple climate change scenarios are currently debated (Lindsay and Martens 1998; Rogers and Randolph 2000; Ebi et al. 2005). It is similarly unclear to what extent global warming is responsible for the resurgence of malaria at higher elevations (Hay et al. 2002; Pascual et al. 2006). Incomplete knowledge of the current distribution of Anopheles, malaria transmission and reproductive potential, the relative importance of precipitation, and a multitude of societal and political changes limit the accuracy of long-term projections. Reliable forecasts of malaria activity have been integrated into mosquito management as early as 1921 in Punjab, India. Experts made seasonal forecasts based on summer precipitation, famine and prevailing economic conditions, previous malaria transmission intensities, and childhood malaria prevalence (Swaroop 1946). The World Health Organization’s Malaria Early Warning System incorporates similar information on disease surveillance, food security and vulnerability, and environmental and climatic conditions (WHO 2001). Strong linkages between December–February precipitation, sea surface temperatures (SST) and malaria epidemics in Botswana also enabled short-term forecasts of malaria transmission (Thomson et al. 2005). Averaging multiple General Circulation Model runs initialized with slightly different boundary conditions will likely increase malaria forecast lead times (Thomson et al. 2006). Significant inter-annual associations between malaria incidence, in situ precipitation, and river height anomalies exist in Peru, Guyana, and Colombia (Gagnon et al. 2002). Relating these hydroclimatic indices to the El Nin˜o Southern Oscillation (ENSO) pointed to temporally lagged influences of ENSO events on malaria epidemics. These studies suggest that regions with unstable malaria transmission and established climate relationships can have relatively accurate seasonal predictions without complete knowledge of malaria transmission dynamics. Secular trends in new malaria cases are historically influenced by developing parasite resistance to human anti-malarial drugs, mosquito pesticide resistance, and funding for mosquito control and public health infrastructure. How climate may influence malaria transmission in climatically favorable tropical regions with year-round and high intensity transmission is under studied. Vector ecology such as competition, predation, or dispersal and human land use and adaptations may be more important determinants of malaria risk than climatic variables.

123

606

4.5 Future research Predictions must be accurate and relevant and the information must be effectively communicated and disseminated to decision makers and other stakeholders. To improve management of vectorborne disease, precipitation estimates derived from doppler radar should be pursued. Although, social science studies of decision making, institutional ethnographies, and public and health policy that improve the use of forecasts are as important as climate prediction research. For example, participatory research engaging subsistence farmers through workshops and forecast public education campaigns improved harvests in Zimbabwe (Patt et al. 2005), similar strategies may diminish the risk of malaria transmission.

5 Waterborne disease Countless pathogens such as Vibrio cholerae, Escherichia coli, Giardia, and Salmonella are found in surface water and often negatively affect human health through the contamination of food and drinking water. Many of these risks can be averted through proper sanitation and sewage systems, although even advanced water treatment systems can become overwhelmed during flood events and fail to adequately sanitize water supplies (Curriero et al. 2001). Cholera is one of the most dangerous and common waterborne diseases and has many well-established links to climate. 5.1 Cholera Vibrio cholerae is a bacterium endemic to tropical and subtropical coastal, brackish, and estuarine microbial communities around the globe and is responsible for the human disease cholera (Huq et al. 1984). Generally, V. cholerae is transmitted through contaminated food and water in communities that do not have access to proper sewage and water treatment systems (http://www.who.int). 5.2 Climate and agent prevalence Numerous climate variables indirectly affect the prevalence of V. cholerae in the environment by creating favorable conditions for zooplankton, primarily crustaceans and copepods, which are the primary hosts of V. cholerae (Huq et al. 1984, 2001; Colwell 1996). Zooplankton populations rely heavily upon phytoplankton populations, which provide food and nutrients and advantageously modify the pH levels of the water (Lipp et al. 2002; Lobitz et al. 2000). Due to the relationship between zooplankton and phytoplankton populations, dynamics

123

Stoch Environ Res Risk Assess (2007) 21:601–613

between climate and phytoplankton, although indirect, are important mechanisms that regulate the prevalence of the bacteria in the environment. Phytoplankton populations are primarily affected by nutrient levels, SST and solar radiation (Lipp et al. 2002). Wind, tidal forces and other physical processes can affect phytoplankton growth through the transport of deep nutrient-loaded water to the surface (Colwell and Huq 2001). However, heavy precipitation can dilute nutrient levels and decrease phytoplankton production (Colwell and Huq 2001). A decrease in incidence of cholera during the monsoon season in Bangladesh has been attributed to this phenomenon (Lipp et al. 2002). SST maxima of 30°C have been demonstrated to augment the propagation of V. cholerae in laboratory experiments, especially in the presence of copepods (Huq et al. 1984), conditions that appear to exist in the natural environment as well. For example, increased cholera incidence in Bangladesh has been observed during periods of high SST (Lobitz et al. 2000). In addition, sufficient amounts of solar radiation are necessary for photosynthesis among phytoplankton (Colwell and Huq 2001). Favorable combinations of these elements can result in phytoplankton blooms and greater abundance of zooplankton, which ultimately result in greater prevalence of V. cholerae in aquatic environments. 5.3 Climate and human exposure Substantial populations of V. cholerae are necessary, though not sufficient, to cause epidemics. Human exposure and infection usually occur when contaminated water or seafood is ingested, and climate variables have been demonstrated to make this more likely. Remote sensing and cholera case data have shown that increases in sea surface heights (SSH) are positively correlated with cholera incidence (Lobitz et al. 2000). It is hypothesized that an increase in SSH might cause inland intrusion of contaminated water and zooplankton, increasing the likelihood of contact between humans and V. cholerae (Lobitz et al. 2000). SSH is modulated by seasonal thermal expansion and contraction of sea water, ocean currents, tidal forces, and the piling up of water resulting from persistent winds and other atmospheric phenomenon. 5.4 Temporal and spatial variability Prior to 1961, cholera endemic areas in the 20th century had been reduced to coastal areas of Bangladesh and southeastern India. However, a large geographical expansion of endemicity has occurred since that time, and currently encompasses regions of Africa and South and Central America (Colwell 1996; Lipp et al. 2002).

Stoch Environ Res Risk Assess (2007) 21:601–613

Although favorable conditions for V. cholerae exist in many locations throughout the world, cholera is only a major health risk in only a handful of locations. This elucidates the importance of other factors, such as cultural and socio-economic conditions, when trying to understand and predict incidence of cholera and other infectious agents. Climate modulates the seasonal and inter-seasonal variability of cholera by affecting both the prevalence and the spatial distribution of the agent in the environment through numerous mechanisms that affect the seasonality of cholera incidence unique for each geographical location. For example, a bimodal distribution of cholera incidence occurs in Bangladesh, with a peak in early spring followed by a larger peak in fall following the monsoon, whereas, a single peak occurs in Calcutta, India in late spring, and in South America the peak occurs during the summer (Lipp et al. 2002). In addition to seasonal variability, endemic areas experience inter-seasonal increases in cholera incidence that have been closely linked to climate variability. Cholera incidence is associated with ENSO, particularly since the intensification of ENSO events beginning in 1976 (Rod et al. 2002). ENSO is not directly responsible for the increase, but complex interactions and feedbacks between its dynamics and the ecology of endemic areas result in favorable conditions for the propagation and transmission of the bacteria. For instance, in the Indian Ocean ENSO alters SST, monsoon intensity, and river discharge into lowland areas (Rod et al. 2002). Moreover, variability of the intensity and direction of trade winds associated with ENSO can result in the reorganization of local ocean currents, culminating in increased or diminished upwelling of deep nutrient-loaded water (Colwell 1996). These changes in wind intensity and direction also affect SSH, and may lead to the intrusion of contaminated sea water into inland areas. Thus, since the prevalence and human exposure of V. cholerae are influenced by many climate variables, there is a pronounced variability in cholera incidence that is highly correlated with climate variables. 5.5 Future research Long-term monitoring of SST, SSH, and chlorophyll concentrations (proxy for plankton concentrations) using remote sensing techniques have begun to verify and fine-tune the established relationships between cholera outbreaks and the environment. If these relationships are better understood, prediction of climatic conditions favorable for phytoplankton blooms and the corresponding V. cholerae, may provide crucial time needed to implement public health measures, and possibly short-circuit outbreaks of cholera.

607

6 Airborne (non-infectious) disease The major non-infections airborne sources of disease are related to air pollutants. Air pollution can include air toxics such as heavy metals and other poisonous substances, but the most ubiquitous pollutants are tropospheric ozone (ozone), PM, lead, carbon monoxide, nitrogen dioxide, and sulfur dioxide. 6.1 Air pollution Air pollution is derived from numerous chemical species and it impacts millions of people around the world. Health impacts range from mild conditions such as eye irritation to serious problems such as cancer or premature death. The U.S. Environmental Protection Agency (EPA) closely monitors six ‘‘criteria’’ air pollutants that are pervasive and considered harmful to human health: ozone, lead, carbon monoxide, nitrogen dioxide, sulfur dioxide, and PM. Ozone and PM will be discussed further here, as they are of particular concern in many U.S. cities and for the large number of people who live in areas with high concentrations. Car exhaust is the main source of ozone precursor pollutants; other sources include industry and consumer products such as paints and cleaners (EPA 1997). PM is a collective term used to describe solid particles and liquid droplets suspended in the air. Sources of PM include roads, industry, agriculture, construction, fossil fuel combustion, vegetation burning, and the processing of metals (Godish 1991). 6.2 Climate and agent prevalence Although air quality is influenced by a variety of sources, day-to-day variability in pollutant concentrations are most strongly influenced by changes in climate and weather variables (Davis and Gay 1993). These variables primarily affect pollutant levels by influencing emissions, chemical reactions that form pollutants, and transport processes (Bernard et al. 2001). The emission and formation of ozone and PM are influenced by a multitude of indirect and direct processes. For instance, photochemical reactions driven primarily by ultraviolet radiation result in the formation of ozone from its precursor pollutants, nitrogen oxides and volatile organic compounds (VOCs). Multiple climatemediated processes influence emissions of PM. High wind speed and low atmospheric moisture have been found to be correlated with PM, likely due to their role in the resuspension of particulates in the atmosphere (Wise and Comrie 2005). In addition, low humidity in conjunction with wind increase the likelihood and intensity of wildfires that contribute to PM. Furthermore, both ozone and PM concentrations can be influenced by climate-induced

123

608

changes in human behaviors, such as the use of air conditioning, which in turn impact the amount of pollution emitted. Once the pollutants are in the atmosphere, climate variables affect their distribution. For instance, atmospheric high-pressure systems can lead to subsidence inversions, in which increasing temperature with height in the atmosphere causes strong stability. The resulting low wind speeds and limited vertical dispersion of pollutants increases ground-level pollution and can cause severe air quality problems. Precipitation plays a pivotal role in the removal of pollutants, especially particulates, from the atmosphere. This can occur through rainout (in-cloud capture of pollutants) or washout (raindrops collect pollutants as they fall and carry them to the surface). 6.3 Climate and human exposure Since humans are constantly immersed in the reservoir (i.e. the atmosphere), any person living in an area with air pollution is at risk for exposure. In particular, people working outdoors and/or engaged in strenuous physical activity outdoors are particularly susceptible due to increased penetration of pollutants into the airways (Bernard et al. 2001). Sensitive groups, such as the elderly, children, and people with respiratory diseases or compromised immune systems, are more likely to be adversely impacted at lower pollution levels or suffer more severe outcomes at high levels (EPA 1997). Meteorological conditions are predictors of both mortality / morbidity rates and air pollution. For instance, people spend more time indoors during certain seasonal conditions, which can result in increased infection rates from other illnesses or reduced exposure to air pollutants (Burnett et al. 2001; Schwartz 2005). Likewise, spatial autocorrelation and clustering of health responses have been found in air pollution-health studies and have been linked to both non-climatic factors such as socio-economic status and to ecological covariates (e.g., climatic conditions) which also demonstrate spatial patterns (Cakmak et al. 2003; Krewski et al. 2003). For this reason, time-series studies of the impact of air pollution on morbidity / mortality normally control for confounders such as weather variables in their regression equations (Rainham et al. 2005), and many investigators have used spatial analysis methods to account for spatial impacts on the associations between pollutants, weather, and mortality / morbidity (Cakmak et al. 2003; Krewski et al. 2003). However, since variations in pollutant concentrations are dependent on meteorological conditions and higher levels of air pollutants are linked to worsened health impacts in a dose-response relationship, changes in climatic conditions that impact ambient pollutant levels (e.g., warmer temperatures for ozone) will have health implications (Rainham et al. 2005).

123

Stoch Environ Res Risk Assess (2007) 21:601–613

6.4 Temporal and spatial variability Since climate has such a strong influence on air pollution, there are pronounced fluctuations in the concentrations of ambient air pollutants on diurnal, seasonal, and inter-annual timescales. Ozone displays strong seasonal and diurnal cycles, with peaks in the summer months and during daily traffic rush-hours. PM also displays a seasonal cycle in many areas that is primarily related to local moisture and wind conditions. Longer-term impacts from broad-scale circulation variability associated with climatic events such as El Nin˜o and global climate change will likely vary by location. These forms of variability can impact air quality both directly (by influencing the fate, toxicity, or behavior of pollutants) or indirectly (by altering the anthropogenic or biogenic production of pollutants). Tropospheric ozone concentrations are generally predicted to rise in the future due to increases in temperature, solar radiation, and stagnation (Bernard et al. 2001; Leung et al. 2005; Mickley et al. 2004). However, projected increases in water vapor content could enhance ozone destruction, leading to no change (Brasseur et al. 1998) or even a net decrease (Stevenson et al. 2005). Since the long-range transport of PM is influenced by synoptic scale meteorological patterns, it follows that changes in these synoptic patterns could influence the frequency of PM air pollution events (Buchanan et al. 2002). PM is closely linked to precipitation, but climate model precipitation predictions are not consistent across models and will likely vary regionally. Projected climate change impacts on air quality are further complicated by the potential for extreme climatic events, such as storms, heat waves, and droughts, which may increase in frequency and intensity (Crane et al. 2005). It is also possible that pollutant concentrations will become more highly variable, making prediction more difficult (Crane et al. 2005). Socio-economic and technological changes will clearly also have major implications for future air quality. 6.5 Future research Advances in both short-term and long-term modeling of climate–air quality relationships would be advantageous for the mitigation of pollution-related health effects. In the short term, forecasting of pollution conditions based on predicted weather patterns are already operational (e.g., AIRNow 2006). These models could be improved by better understanding of regionally-specific weather controls on pollutant concentrations. Long-term prediction of air quality trends would also benefit from a better understanding of climate–pollutant relationships, particularly the interaction between complete synoptic climate patterns and the resulting pollutant conditions. Improved regionally-

Stoch Environ Res Risk Assess (2007) 21:601–613

specific climate model predictions would also enable a better assessment of future conditions. Finally, studies assessing the combined impact of changing climatic and socio-economic conditions on air quality would provide a more complete picture of potential human health outcomes.

7 Weather-related disease Hazards such as landslides, lightning, flooding, and tornados affect populations around the globe in often catastrophic ways. In proportion to these events, there is less awareness of ‘heat’ as a hazardous extreme weather event (Souch and Grimmond 2004). However, heat waves have broad geographic and social impacts and have been linked to excess human mortality and illness (Smoyer-Tomic et al. 2003). Furthermore, future climate scenarios project that heat and heat waves will increase in severity and frequency with increasing global mean temperatures (McGeehin and Mirabelli 2001; Souch and Grimmond 2004). 7.1 Heat waves Heat waves, which are associated with more deaths in the United States on average than any other weather-related phenomenon, exhibit perhaps the most direct impact on human health (CDC 2002). Although extreme temperatures can affect health by inducing heat stroke and heat exhaustion, the increase in mortality associated with temperature extremes is primarily due to deaths involving cardiovascular and respiratory disorders that are exacerbated by high temperatures (Kalkstein and Greene 1997). Since extreme heat typically worsens existing health conditions, it is difficult to quantify its direct effects on health. Currently, examining the disparity of mortality during periods of extreme heat versus mortality during normal conditions offers the best method to approximate the effects of extreme heat on human health. 7.2 Climate and agent intensity In the case of heat waves, the primary agent is itself an element of climate: temperature. Several characteristics of heat waves influence their impact on populations: their frequency (both in a given summer and the number of events over a longer period of time); their duration; and their intensity (Smoyer-Tomic et al. 2003). As a general rule, higher temperatures result in greater mortality, though the persistence of elevated temperatures throughout the day and evening has been associated with increased mortality. If temperatures cool considerably during the evening hours, temperature associated mortality tends to decrease relative to times when high temperatures are sustained overnight

609

(Kalkstein and Davis 1989). In addition, the persistence of heat events has been demonstrated to impact associated mortality, such that the first day of a heat wave causes fewer deaths than subsequent days (Sheridan and Kalkstein 2004). However, temperature is not the only climatic variable that modulates the effects of a heat wave on human health. Other meteorological variables augment and diminish the effect of heat on a population. For instance, higher humidity and apparent temperatures worsen the effects of heat on a population, whereas wind and cloud cover tend to alleviate these effects on most populations (Kalkstein and Davis 1989). 7.3 Climate and human exposure Environmental and behavioral modifications can reduce human exposure to excessive heat, although behavioral modifications are generally more feasible (Smoyer-Tomic and Rainham 2001). Modifications including indoor airconditioning can reduce exposure to extreme temperatures and enable behavioral modifications such as scheduling outdoor activities during the coolest hours of the day. However, these amenities are not always available to lower income populations and illustrate how demographic, socioeconomic, and housing factors can affect a person’s ability to cope with extreme heat (Basu and Samet 2002). Other populations at high risk for heat-related health impacts include the elderly, individuals with compromised immune systems, and person taking certain types of medication that cause an extreme sensitivity to temperature fluctuations (Smoyer-Tomic and Rainham 2001). In addition, dense urban areas typically have higher heat indexes (a combination of temperature and humidity) than surrounding landscapes and may pose an increased risk of heat-related illness or mortality to urban populations (Basu and Samet 2002). 7.4 Temporal and spatial variability The degree to which populations are impacted by extreme temperatures depends largely on the geographic location of the heat wave. It is not the temperature that most impacts a population, rather how much the temperature differs from the mean climatological conditions for specific region. Temperate regions are much more susceptible to adverse health effects from heat waves than regions that consistently experience high temperatures (Kalkstein and Davis 1989). For example, the temperature at which heat associated mortality increases is under 30°C threshold in many temperate cities in the United States, whereas cities in the southwest US typically do not experience mortality increases unless temperatures surpass 40°C (Kalkstein and Davis 1989). This is likely due to the fact that populations

123

610

in regions that frequently experience adverse conditions are acclimated to those conditions and utilize regionally-specific technology that can mitigate the extreme conditions (e.g. home air conditioning; Kalkstein and Davis 1989; Kalkstein and Greene 1997). Intensifying climate variables such as humidity, wind, and cloud cover can mitigate the impacts of extreme temperatures as well. Furthermore, the adverse health impacts of a heat wave are highly correlated with the seasonal timing of the event. Specifically, heat waves that occur in the beginning of the summer have a more severe effect than heat waves that occur in the latter half of the season (Kalkstein and Davis 1989; Kalkstein and Greene 1997). This trend could indicate that as populations acclimatize to increasingly warm temperatures throughout the summer, their susceptibility to heat-related health impacts decreases (Sheridan and Kalkstein 2004). However, other factors could be at play as well, such as the timing of vacations during the latter half of the summer, an exodus out of cities and into rural areas during the summer months and a decline in susceptible populations because of mortality caused by heat waves earlier in the summer. 7.5 Future research Future impacts of heat waves could be more significant as the globe continues to warm and regional weather patterns may become more variable. Heat sensitive and at-risk senior citizens will continue to comprise a larger fraction of the global population. For example, 25% of the United States population is projected to be older than 65 years in 2100 (Ebi et al. 2004). Climate models suggest that extreme weather events associated with excess mortality will be increasing in frequency in upcoming decades (Kalkstein and Greene 1997). Some studies suggest that a 70–100% increase in temperature related mortality by the year 2050 is a conservative estimate (Kalkstein and Greene 1997). Excess heat mortality in most of the United States on the other hand significantly decreased in spite of increasing temperatures from 1969 to 1998; potentially due to increased air conditioning adoption, acclimatization, and improved government responses (Davis et al. 2003). Future research directions and policy responses to these changing impacts will also need to complement the suite of technological, socio-economic, political, and cultural variables that will further contribute to the degree of vulnerability experienced by populations during heat waves.

8 Conclusion Simplistic correlations between single climate variables and human health outcomes are rarely realistic due to the complex interactions that occur between agent, host,

123

Stoch Environ Res Risk Assess (2007) 21:601–613

reservoir, and the environment. Climate variables are an essential part of the environment, interacting with other components through complex mechanisms and feedbacks. Although climate variables typically have subtle or no measurable effects on disease systems, some systems are extremely sensitive to climate variability. The diseases reviewed in this paper exemplify some of these systems. For instance, precipitation is needed for propagation of the agent causing coccidioidomycosis, but a lack of precipitation creates conditions that are conducive for transmission of the disease to humans. Climate links to HPS are obscured by complex rodent dynamics that occur between the climate influence and transmission to humans. Climate affects the life-cycle and habits of mosquito-vectors associated with malaria by affecting the availability of water and creating advantageous temperatures amongst others. Cholera-climate dynamics modify the spatio-temporal occurrence of the disease over short distances, with seasonal distribution patterns varying between nations in close proximity, such as India and Bangladesh. Heat waves, while directly tied to climate, impact human health differently depending on geographic location and societal adaptations. Air pollution concentrations can similarly be exacerbated by either warmer or colder conditions, as well as changes in atmospheric stability. Owing to these many complexities and interactions, understanding the role climate plays in a specific human health outcomes requires detailed knowledge of the disease ecology at the particular time and place under study. Numerous climate-mediated processes can affect hosts, reservoirs, agents and other environmental components, augmenting or diminishing the risk of individual diseases. Oftentimes, a multitude of climate variables will play pivotal roles in a disease system, and furthermore, each climate variable can affect the system through several independent mechanisms. The impact of any single variable on the incidence of disease tends to depend on what component of the system it affects. Typically, climate variables that contribute to human exposure have a pronounced effect on the incidence of disease, whereas climate-mediated mechanisms that only influence the prevalence of the agent in the environment have more modest effects on disease incidence. The greater the role climate plays in a disease system, the more intimately the disease is tied to climate variability. Generally, short-term climate fluctuations affect the incidence of previously endemic disease within a specific geographic region, whereas longer-term variability has the capacity to impact the geographical extent of an agent and its associated disease. Consequently, the effect of climate on individual diseases is highly variable, but it can have pronounced effects on the temporal and spatial distribution of disease.

Stoch Environ Res Risk Assess (2007) 21:601–613

Climatic sensitive health outcomes oftentimes disproportionately impact politically-marginalized populations with less adaptive capacity and access to resources. The role of social, cultural, and political systems in influencing the distribution of health outcomes cannot be overstated; sub-Saharan Africa hosts 90% of all malaria cases, air pollution is related to a 20% lung capacity decrease in city children, and physiologically vulnerable and socially isolated elderly have elevated heat mortality risks (Gauderman et al. 2004; Basu and Samet 2002). Many populous regions of the world have environments which are suitable for the propagation and transmission of disease, but effective sanitation and risk-reducing behaviors can greatly diminish risk. For example, most waterborne illnesses are entirely preventable through the careful sanitation of water supplies, environmental regulations can limit emissions from industrial sources and diminish air pollution, and mosquito control has been effective in combating mosquito-borne illnesses. Many of these solutions are beyond the capacity of some societies due to political, socio-economical and cultural constraints. These factors must all be considered when the impacts of climate on human health are investigated. Ultimately, insights into disease ecologies in the context of coupled natural and human systems have the potential to lead to better prediction and preventive strategies, which will be crucial for protecting human health as the physical and social environments of the world continue to evolve.

References AIRNow (2006) AIRNow: Quality of Air Means Quality of Life. http://www.airnow.gov/ Accessed Nov 2006 Basu R, Samet JM (2002) Relation between elevated ambient temperature and mortality: a review of the epidemiologic evidence. Epidemiol Rev 24:190–202 Bayoh M, Lindsay S (2003) Effect of temperature on the development of the aquatic stages of Anopheles gambiae sensu stricto (Diptera: Culcidae). Bull Entomol Res 5:375–381 Bernard SM, Samet JM, Grambsch A, Ebi KL, Romieu I (2001) The potential impacts of climate variability and change on air pollution-related health effects in the United States. Environ Health Perspect 2:199–209 Brasseur GP, Kiehl JT, Mu¨ller JF, Schneider T, Granier C, Tie XX, Hauglustaine DD (1998) Past and future changes in global tropospheric ozone: impact on radiative forcing. Geophys Res Lett 20:3807–3810 Buchanan CM, Beverland IJ, Heal MJ (2002) The influence of weather-type and long-range transport on airborne particle concentrations in Edinburgh, UK. Atmos Environ 36:5343–5354 Burnett R, Ma R, Jerrett M, Goldberg M, Cakmak S, Pope C, Krewski D (2001) The spatial association between community air pollution and mortality: a new method of analyzing correlated geographic cohort data. Environ Health Perspect 109:375–380. doi:10.2307/3434784 Cakmak S, Burnett RT, Jerrett M, Goldberg M, Pope C, Renjun M, Gultekin T, Thun M, Krewski D (2003) Spatial regression

611 models for large-cohort studies linking community air pollution and health. J Toxicol Environ Health 66:1811–1823 CDC (2002) Heat-related deaths—Four states, July–August 2001, and United States, 1979–1999. MMWR 51:567–570 Coetzee M, Craig M, Le Sueur D (2000) Distribution of African malaria mosquitoes belonging to the Anopheles gambiae complex. Parasitol Today 16:74–77 Colwell RR (1996) Global climate and infectious disease: the cholera paradigm. Science 274:2025–2031 Colwell RR, Huq A (2001) Marine ecosystems and cholera. Hydrobiologia 3:141–145. doi:10.1023/A:1013111016642 Comrie AC (2005) Climate factors influencing the coccidioidomycosis seasonality and outbreaks. Environ Health Perspect 6:688– 692 Crane M, Whitehouse P, Comber S, Ellis J, Wilby R (2005) Climate change influences on environmental and human health chemical standards. Hum Ecol Risk Assess 11:289–318 Cromley EK, McLafferty S (2002) GIS and public health. Guilford Press, New York Curriero FC, Patz JA, Rose JB, Lele S (2001) The association between extreme precipitation and waterborne disease outbreaks in the United States, 1948–1994. Am J Public Health 8:1194– 1199 Davis RE, Gay DA (1993) A synoptic climatological analysis of air quality in the Grand Canyon National Park. Atmos Environ 27:713–727 Davis RE, Knappenberger PC, Michaels PJ, Novicoff WM (2003) Changing heat-related mortality in the United States. Environ Health Perspect 111:1712–1718 Del Casino VJ (2004) (Re)placing health and health care: mapping the competing discourses and practices of ‘traditional’ and ‘modern’ Thai medicine. Health Place 10:59–73 Ebi KL, Teisberg TJ, Kalkstein LS, Robinson L, Weiher RF (2004) Heat watch/warning systems save lives: estimated costs and benefits for Philadelphia 1995–98. Bull Am Meteorol Soc 85:1067–1073 Ebi K, Hartman J, Chan N, McConnell J, Schlesinger M, Weyant J (2005) Climate suitability for stable malaria transmission in Zimbabwe under different climate change scenarios. Clim Change 73:375–393 Elliott P (2000) Spatial epidemiology: methods and applications. Oxford University Press, New York Engelthaler DM, Mosley DG, Cheek JE, Levy CE, Komatsu KK, Ettestad P, Davis T, Tanda DT, Miller L, Frampton JW, Porter R, Bryan RT (1999) Climatic and environmental patterns associated with hantavirus pulmonary syndrome, four corners region, US. Emerg Infect Dis 1:87–94 EPA (1997) Ozone: Good Up High, Bad Nearby. http://www.epa.gov/ oar/oaqps/gooduphigh Epstein P, Mills E (2005) Climate change futures: health, ecological and economic dimensions. The Center for Health and the Global Environment, Harvard Medical School, Cambridge, p 142 Gagnon A, Smoyer-Tomic K, Bush A (2002) The El Nin˜o-Southern Oscillation and malaria epidemics in South America. Int J Biometeorol 46:81–89 Gauderman WJ, Avol E, Gilliland F, Vora H, Thomas D, Berhane K, McConnell R, Kuenzli N, Lurmann F, Rappaport E, Margolis H, Bates D, Peters J (2004) The effect of air pollution on lung development from 10 to 18 years of age. N Engl J Med 351:1057–1067 Glass GE, Cheek JE, Patz JA, Shields TM, Doyle TJ (2000) Using remotely sensed data to identify areas at risk for hantavirus pulmonary syndrome. Emerg Infect Dis 3:238–247 Godish T (1991) Air quality. Lewis Publishers, Chelsea Haggett P (1994) Geographical aspects of the emergence of infectious diseases. Geogr Ann 76:91–104

123

612 Hay S, Cox J, Rogers D, Randolph S, Stern D, Shanks G, Myers M, Snow R (2002) Climatic change and the resurgence of malaria in the East African Highlands. Nature 415:905–909 Hjelle B, Glass G (2000) Outbreak of hantavirus infection in the four corners region of the United States in the Wake of the 1997– 1998 El Nin˜o-Southern Oscillation. J Infect Dis 181:1569–1573. doi:10.1086/315467 Huq A, West PA, Small EB, Huq MI, Colwell RR (1984) Influence of water temperature, salinity, and pH on survival and growth of toxigenic Vibrio cholerae serovar 01 associated with live copepods in laboratory microcosms. Appl Environ Microbiol 2:420–424 Huq A, Sack R, Colwell R (2001) Cholera and global ecosystems. In: Aron J, Patz J (eds) Ecosystem change and public health: a global perspective. Johns Hopkins University Press, Baltimore, pp 327–347 Kalkstein L, Davis R (1989) Weather and human mortality: an evaluation of demographic and interregional responses in the United States. Ann Assoc Am Geogr 1:44–64 Kalkstein L, Greene S (1997) An evaluation of climate/mortality relationships in large US cities and the possible impacts of climate change. Environ Health Perspect 1:84–93 Kolivras KN, Comrie AC (2003) Modeling valley fever incidence based on climate condition in Pima county, Arizona. Int J Biometeorol 47:87–101 Kolivras KN, Comrie AC (2004) Climate and infectious disease in the southwestern United States. Prog Phys Geog 3:387–398. doi:10.1191/0309133304pp417ra Krewski D, Burnett R, Goldberg M, Hoover B, Siemiatycki J, Jerrett M, Abrahamowicz M, White W (2003) Overview of the reanalysis of the Harvard six cities study and American Cancer Society study of particulate air pollution and mortality. J Toxicol Env Heal A 66 (16–19):1507–1551 Leung LR, William J, Gustafson I (2005) Potential climate change and implications to US air quality. Geophys Res Lett 32(L16711). doi:10.1029/2005GL022911 Lindsay S, Martens W (1998) Malaria in the African highlands: past, present and future. Bull World Health Organ 76:33–45 Lindsay S, Parson L, Thomas C (1998) Mapping the range and relative abundance of the two principal African malaria vectors, Anopheles gambiae sensu stricto and An. arabiensis, using climate data. Proc R Soc Lond B Biol Sci 265:847–854 Lipp EK, Huq A, Colwell RR (2002) Effects of global climate on infectious disease: the cholera model. Clin Microbiol Rev 15:757–770 Lobitz B, Beck L, Huq A, Wood B, Fuchs G, Faruque ASG, Colwell R (2000) Climate and infectious disease: use of remote sensing for detection of Vibrio cholerae by indirect measurement. Proc Natl Acad Sci USA 4:1438–1443. doi:10.1073/pnas.97.4.1438 Martens W, Niessen L, Rotmans J, Jetten T, McMichael A (1995) Potential impact of global climate change on malaria risk. Environ Health Perspect 103:458–464 Mayer JD (2000) Geography, ecology and emerging infectious diseases. Soc Sci Med 50:937–952 McGeehin MA, Mirabelli M (2001) The potential impacts of climate variability and change on temperature-related morbidity and mortality in the United States. Environ Health Perspect 109:185– 189 Mickley LJ, Jacob DJ, Field BD (2004) Effects of future climate change on regional air pollution episodes in the United States. Geophys Res Lett 31(L24103). doi:10.1029/2004GL021216 Mills JN, Ksiazek TG, Peters CJ, Childs JE (1999) Long-term studies of hantavirus reservoir populations in the southwestern United States: a synthesis. Emerg Infect Dis 1:135–142 Mutuku F, Alaii J, Nabie Bayoh M, Gimnig J, Vulule J, Walker E, Kabiru E, Hawley W (2006) Distribution, description, and local

123

Stoch Environ Res Risk Assess (2007) 21:601–613 knowledge of larval habitats of Anopheles gambiae S.L. in a village in Western Kenya. Am J Trop Med Hyg 74:44–53 Pampana E (1969) A textbook of malaria eradication. Oxford University Press, London Pappagianis D (1988) Epidemiology of coccidioidomycosis. Curr Top Med Mycol 6:199–238 Parmenter RR, Brunt JW, Moore DI, Ernest S (1993) The hantavirus epidemic in the Southwest: rodent population dynamics and the implications for transmission of hantavirus-associated adult respiratory distress syndrome (HARDS) in the Four Corners region. Department of Biology, University of New Mexico, Albuquerque, New Mexico; Sevilleta Long-Term Ecological Research Program (LTER); Publication No:41 Pascual M, Ahumada JA, Chaves LF, Rodo´ X, Bouma M (2006) Malaria Resurgence in the East African highlands: Temperature trends revisited. Proc Natl Acad Sci USA 15:5829–5834. doi:10.1073/pnas.0508929103 Patt A, Suarez P, Gwata C (2005) Effects of seasonal climate forecasts and participatory workshops among subsistence farmers in Zimbabwe. Proc Natl Acad Sci USA 102:12623–12628 Patz J, Olson S (2006) Malaria risk and temperature: influences from global climate change and local land use practices. Proc Natl Acad Sci USA 103:5635–5636 Patz J, Strzepek K, Lele S, Hedden M, Greene S, Noden B, Hay S, Kalkstein L, Beier J (1998) Predicting key malaria transmission factors, biting and entomological inoculation rates, using modeled soil moisture in Kenya. Am J Trop Med Hyg 3:818–827 Patz JA, Campbell-Lendrum D, Holloway T, Foley JA (2005) Impact of regional climate change on human health. Nature 7066:310– 317. doi:10.1038/nature04188 Rainham D, Smoyer-Tomic K, Sheridan S, Burnett R (2005) Synoptic weather patterns and modification of the association between air pollution and human mortality. Int J Environ Heal R 5:347–360 Rodo´ X, Pascual M, Fuchs G, Faruque ASG (2002) ENSO and cholera: a nonstationary link related to climate change? Proc Natl Acad Sci USA 20:12901–12906. doi:10.1073/ pnas.182203999 Rogers D, Randolph S (2000) The global spread of malaria in a future, warmer world. Science 289:1763–1766 Schwartz J, Litonjua A, Suh H, Verrier M, Zanobetti A, Syring M, Nearing B, Verrier R, Stone P, MacCallum G, Speizer FE, Gold DR (2005) Traffic related pollution and heart rate variability in a panel of elderly subjects. Br Thorac Soc 60:455–461 Sheridan S, Kalkstein L (2004) Progress in heat watch-warning system technology. Bull Am Meteorol Soc 12:1931–1940 Smoyer KE (1998) Putting risk in its place: Methodological considerations for investigating extreme event health risk. Soc Sci Med 47:1809–1824 Smoyer-Tomic KE, Rainham DC (2001) Beating the heat: development and evaluation of a Canadian hot weather health-response plan. Environ Health Perspect 109:1241–1248 Smoyer-Tomic KE, Kuhn R, Hudson A (2003) Heat wave hazards: an overview of heat wave impacts in Canada. Nat Hazards 28:463– 485 Souch C, Grimmond C (2004) Applied climatology: ‘heat waves’. Prog Phys Geogr 28:599–606 Stevenson D, Doherty R, Sanderson M, Johnson C, Collins B, Derwent D (2005) Impacts of climate change and variability on tropospheric ozone and its precursors. Faraday Discuss 130:41– 57 Swaroop S (1946) Forecasting of epidemic malaria in Punjab. Am J Trop Med 29:1–17 Thomson M, Connor S, Phindela T, Mason S (2005) Rainfall and sea surface temperature monitoring for malaria early warning in Botswana. Am J Trop Med Hyg 73:214–221

Stoch Environ Res Risk Assess (2007) 21:601–613 Thomson M, Doblas-Reyes F, Mason S, Hagedorn R, Connor S, Phindela T, Morse A, Palmer T (2006) Malaria early warnings based on seasonal climate forecasts from multi-model ensembles. Nature 439:576–579 Wise EK, Comrie AC (2005) Meteorologically adjusted urban air quality trends in the southwestern United States. Atmos Environ 16:2969–2980

613 WHO (2001) Malaria early warning systems: a framework for field research in Africa. World Health Organization, Roll Back Malaria Technical Support Network for the Prevention and Control of Malaria Epidemics, Geneva Yates TL, Mills JN, Parmenter CA, et al (2002) The ecology and evolution of an emergent disease: hantavirus pulmonary syndrome. BioScience 10:989–998

123