Chapter 12: Vector-borne Diseases and Climate Change Protecting our Health from Climate Change: a Training Course for Public Health Professionals Chapter 12: Vector-borne Diseases and Climate Change

Vector-borne Disease Mortality Distribution Vector-borne diseases (VBDs) are among the major microbial causes of morbidity and mortality in the world today affecting nearly half of the world’s population, the majority of whom reside in developing countries located in the tropical and subtropical areas of the world. In countries that provide statistics to the WHO, VBDs collectively account for more than 1.5 million human deaths per year (Hill et al., 2005). However, many diseases are under-reported, especially those that are rarely fatal like Onchocerciasis and Lymphatic filariasis. When considering disease burden, it is important to also evaluate the morbidity it causes. Measures of DALY – or disability adjusted life years – are one measure that is often used to assess this. One DALY is defined as one lost year of healthy life and is a measurement of the gap between the current health of a population and an ideal situation where everyone lives to old age in full health. The geographical distribution of VBDs is largely reflective of the geographical distribution of both vectors and their reservoir hosts (if they have reservoir hosts). Many vectors are cold-blooded arthropod species unable to regulate their own internal temperature and therefore highly dependent on climate for survival and development. WHO, 2005 Majority of Vector-borne Disease (VBD) burden borne by developing countries Disproportionate amount in Africa

Vector-borne Disease What is VBD? Types of VBD transmission: Human-vector-human (Anthroponotic Infections) Vector Humans Malaria Dengue Yellow fever Animal-vector-human (Zoonotic Infections) Vector Animals Humans Lyme disease Hantaviral disease Most arboviral diseases (e.g., WNV) VBDs are diseases that are spread by arthropod or small animal vectors. Vectors act as the main mode of transmission of infection from one host to another and as such form an essential stage in the transmission cycle. Two main types of VBD transmission exist: Anthroponotic infections – or human-vector-human transmissions, where humans are the only reservoir of the disease Zoonotic infections – or animal-vector-human transmission, where animals are the main reservoir of the disease and humans are considered secondary or spillover hosts and do not generally contribute to the disease transmission cycle as their levels of circulating pathogen are often too low to help maintain transmission. The type of transmission of a VBD has implications for control strategies. Anthroponotic infections can theoretically be eradicated if all human cases of the disease can be treated, whereas zoonotic diseases are much more difficult to control since all animal reservoirs of the disease would need to be treated.

Vector-borne Diseases of Concern Pathogen Vector Transmission Protozoan Malaria Plasmodium falciparum, vivax, ovale, malariae Anopheles spp. Mosquitoes Anthroponotic Leishmaniasis * Leishmania spp. Lutzomyia & Phlebotomus spp. Sandflies Zoonotic Trypanosomiasis * Trypanosoma brucei gambiense, rhodesiense Glossina spp. (tsetse fly) Chagas disease * Trypanosoma cruzi Triatomine spp. There are many VBDs of concern, especially in developing countries, a number of which are on the WHO list of neglected tropical diseases (including Leishmaniasis, Trypanosomiasis, Chagas, Dengue, Lymphatic filariasis, and Onchocerciasis) because they occur in areas where poverty is the most significant risk factor for their occurrence. * WHO neglected tropical disease Hill et al., 2005

Vector-borne Diseases of Concern (cont.) Pathogen Vector Transmission Viral Dengue * DEN-1,2,3,4 flaviviruses Aedes aegypti mosquito Anthroponotic Yellow fever Yellow fever flavivirus Encephalitis (West Nile, Lyme, etc.) Flavi-,alpha- and bunyaviruses Mosquitoes and ticks Zoonotic Firlarial nematodes Lymphatic filariasis * Brugia malayi, timori, Wuchereria bancrofti Anopheles, Culex, Aedes mosquitoes Onchocerciasis * Onchocerca volvulus Simulium spp. blackflies The agents causing these diseases are protozoa, bacteria, viruses and filarial nematodes and are transmitted by a range of arthropod vectors. * WHO neglected tropical disease Hill et al., 2005

Vector-borne Disease Dynamics Susceptible population Migration (forced) Vector environment There are 3 crucial elements which must co-exist for the occurrence of VBD: the susceptible population, the vector (most often arthropods), and the disease pathogen (e.g., bacteria, virus, parasite). In areas where VBD most frequently occurs, conditions must be suitable for vectors and pathogens, which implies physiologically suitable conditions for vector, host, and pathogen survival and reproduction/replication. There are a number of areas in the world where conditions may be suitable for all three components; however, other factors have acted to prevent or eradicate disease transmission in these areas, perhaps as a result of improved health care services or vector control measures. Global climate change is likely to affect all 3 of these components both directly and indirectly. As an example of direct effects: Arthropods are highly sensitive to changes in temperature and precipitation as they cannot regulate their own internal temperatures and are therefore critically dependent on climate for survival and development (Githeko et al., 2000). Changes in climate may accelerate the development time of some arthropod species, for example. Similarly, many pathogens are climate sensitive as well, and changes in climate could result in increased reproduction rates of some pathogens. Some example of indirect effects might include: Changes in livelihood conditions due to climate change, which could affect nutritional status of individuals, thereby potentially increasing susceptibility to disease. The next few slides will cover indirect and direct effects of climate change on human health in more detail. Vector Pathogen Survival, lifespan Reproduction/breeding patterns Biting behavior Survival Transmission Replication in host

Climate vs. Weather Effects Average trend of weather patterns for a given location (averages over a long time period) Constrains the range of infectious disease E.g., malaria in Kenyan Highlands Weather Day-to-day climate conditions for a given location (shorter time periods, highly variable) Affects the timing and intensity of outbreaks E.g., dengue outbreak in Sumatra However, before moving on, a quick clarification: How are climate and weather related? Climate is the average weather in a location over a long period of time, while weather is the daily climate conditions at a given location. Weather will tend to fluctuate on a daily basis while climate is the long-term average of trends in a location. With respect to VBDs, climate acts to constrain the range of these diseases by limiting the geographical range of suitable habitats for vector and pathogen survival. In other words, climate determines what regions can potentially support one disease or another based on the physiological requirements of the vector, reservoir host, and/or pathogen. For example, the range of endemic malaria was believed to be constrained in altitude by the temperature threshold requirements of its Anopheline mosquito host; however, recent changes in climate are thought to have begun to shift these thresholds higher in altitude and be contributing to an increased incidence of endemic malaria in the Kenyan highlands. Weather, on the other hand, will affect the timing and intensity of a disease outbreak. For example, increased droughts followed by bursts of intense precipitation have been linked with mosquito-borne disease outbreaks, such as dengue, in countries like India and Bangladesh. The link between climate and malaria in the Kenyan Highlands example will be discussed in further detail later in the presentation as will the link between weather and dengue outbreaks. Epstein, 2001; Patz, 2002

(including medical care) Genetic/ constitutional Environmental Determinants of Human Disease Social and economic policies Living conditions Livelihoods Institutions (including medical care) Individual/ population Social relationships Health Individual risk factors Environmental determinants of health are generally external factors which can be causally linked to changes in health status. For example: An individual’s genetic makeup may determine their susceptibility or resistance to a disease. While the ability to make a living and adequately feed oneself may affect an individual’s immunity to infection. Malnutrition may acerbated one’s ability to carry out daily tasks and make a living in the first place. Social relationships may affect access to resources, including transport to health care facilities While social and economic policies may determine the availability of such health care services. All of these factors contribute both individually and collectively to individual and population health. Pathophysiologic pathways Genetic/ constitutional factors

(including medical care) Genetic/ constitutional Environmental Determinants of Human Disease (cont.) Climate? Social and economic policies Institutions (including medical care) Living conditions Livelihoods Individual/ population Health Individual risk factors Social relationships (Note to presenter: although multiple potential indirect effects have been listed, may want to only describe one or two of these.) Climate is in itself a contributor to individual and population health and changes in climate are likely to have consequences on human health. According to the Intergovernmental Panel on Climate Change or IPCC, human beings are exposed to climate change through changing weather patterns (such as more intense and frequent extreme weather events) and indirectly through changes in water, air, food quality and quantity, ecosystems, agriculture, and economy. As an example of indirect effects: Extreme weather events could cause social or economic disruption as a result of destruction of infrastructures Climate change could also result in poor growing conditions for crops or livestock causing increased household vulnerability to poverty, hunger, disease, mortality, or displacement – all with resultant implications on human health and well being Individual risk factors could also include location of housing in areas with higher risk of disease transmission Genetic factors could include higher susceptibility to certain diseases, which are exacerbated by climate change Pathophysiologic effects could include altered physiological responses such as changes in hormone regulation in response to stress Social relationships may be altered by climate change – perhaps as a direct result of absence due to death or migration While the ways in which institutions react in the face of climate change will impact resources available for community members to cope with climate change effects. It is important to keep these other factors in mind when planning disease intervention strategies. Genetic/ constitutional factors Pathophysiologic pathways

Relationship Between Human and Animal Health Vector Human disease Animal disease Phthiraptera (Lice) Typhus Hemioptera (Bugs) Chagas’ disease Siphonaptera (Fleas) Plague, Q fever Psychodidae (Sand flies) Leishmania Canine leishmania Culicoidae (Mosquitoes) MANY RVF Simulidae (Black flies) Onchocerciasis Ceratopogonidae (Midges) Bluetongue Glossinidae (Tsetse flies) Sleeping sickness Animal tryps Tabanidae EIA, Sura, T. vivax, Muscidae Mastitis Muscoidae Screwworm, fly strike Ticks Snails Bilharzia Fascioliasis Another thing to keep in mind is the close relationship between human and animal health. Many of the pathogens which affect animals can also affect humans. Climate effects which are likely to increase the burden of animal disease, especially livestock disease, are likely to have both direct and indirect effects on human health as well. Links between animals and humans are such that vectors which can carry pathogens that are infectious to animals may also be infectious to humans and vice versa. For example: Animal trypanosomiasis and sleeping sickness are caused by same pathogen (Typanosoma brucei) and carried by the same vector (the tsetse fly). If climate change effects result in an increased incidence of the disease in animal populations, as these are mostly cattle, this will have implications for human livelihood. Additionally, this will likely increase the risk of human incidence of the disease as a result of an increased abundance of the disease in an animal reservoir.

Direct Effects of Climate Change on Vector-borne Disease Climate change has the potential to Increase range or abundance of animal reservoirs and/or arthropod vectors (e.g., Lyme, Malaria, Schistosomiasis) Enhance transmission (e.g., West Nile virus and other arboviruses) Increase importation of vectors or pathogens (e.g., Dengue, Chikungunya, West Nile virus) Increase animal disease risk and potential human risk (e.g., African trypanosomiasis) In terms of direct effects, climate has the potential to: Increase the range or abundance of both animal reservoirs and arthropod vectors. There is some emerging evidence of this occurring with lyme disease in North America, malaria in the Kenyan Highlands, and bluetongue in Europe. 2. Climate change may also prolong the length of the transmission cycles of disease or the transmission season of diseases West Nile virus (WNV), which has recently appeared in North America, has an amplification cycle involving mosquitoes and avian reservoir hosts. Human risk of infection is highest late in the summer when mosquito population densities are highest. Warmer spring and fall temperatures could increase the transmission season of the disease, thereby shifting the risk of human infection of the disease earlier in the summer. 3. Climate could also increase the likelihood of successful importation of disease vectors and animal host reservoirs. For example, the global spread of the Asian tiger mosquito, Aedes albopictus, which has been linked to the sale of used tires around the world, was linked to an outbreak of chikungunya virus, a dengue-like virus in Italy in 2007. Importation of a suitable animal reservoir is believed to be one of the possible methods of introduction of WNV to North America in the late 1990s. 4. As mentioned previously, climate change effects resulting in increased animal incidence of disease are likely to increase the risk of human disease as well. Greer et al., 2008

Temperature Effects on Vectors and Pathogens Survival decrease/increase depending on the species Changes in the susceptibility of vectors to some pathogens Changes in rate of vector population growth Changes in feeding rate and host contact Pathogen Decreased extrinsic incubation period of pathogen in vector at higher temperatures Changes in the transmission season Changes in geographical distribution Decreased viral replication Temperature can affect both the distribution of the vector and the effectiveness of pathogen transmission through the vector. Gubler et al. (2001) list a range of possible mechanisms whereby changes in temperature impact on the risk of transmission of VBD: Temperature may act to: Increase or decrease vector survival Change the rate of vector population growth Change the feeding behavior of vectors Change the susceptibility of vector to pathogens Change the incubation period of pathogens in vectors Change the seasonality of vector activity Change the seasonality of pathogen transmission Vector is infective. Gubler et al., 2001

Precipitation Effects on Vectors Survival: increased rain may increase larval habitat Excess rain can eliminate habitat by flooding Low rainfall can create habitat as rivers dry into pools (dry season malaria) Decreased rain can increase container-breeding mosquitoes by forcing increased water storage Heavy rainfall events can synchronize vector host-seeking and virus transmission Increased humidity increases vector survival and vice-versa Precipitation can also have a number of effects both the vector and pathogens. Gubler et al. 2001 highlight that: Precipitation effects could include: • Increased surface water thereby providing increased breeding sites for vectors • Decreased rainfall could also increase breeding sites by slowing river flow • Increased rain could increase vegetation and allow expansion in populations of vertebrate host • Flooding could eliminate habitat for both vectors and vertebrate hosts • Flooding could also force vertebrate hosts into closer contact with humans. Gubler et al., 2001

Precipitation Effects on Pathogens Few direct effects but some data on humidity effects on malarial parasite development Precipitation can also have a number of effects both the vector and pathogens Gubler et al. 2001 highlight that: Precipitation effects could include: • Increased surface water thereby providing increased breeding sites for vectors • Decreased rainfall could also increase breeding sites by slowing river flow • Increased rain could increase vegetation and allow expansion in populations of vertebrate host • Flooding could eliminate habitat for both vectors and vertebrate hosts • Flooding could also force vertebrate hosts into closer contact with humans. Gubler et al., 2001

Vector Activity Increased relative humidity increases activity, heavy rainfall decreases activity Increased activity increases transmission rates Humidity and precipitation can also have a significant role in vector activity. A greater relative humidity can increase vector activity, but heavy rainfall can actually decrease activity. Increased activity increases transmission rates. Ogden et al., 2005; Vail and Smith, 1998 National Geographic Ranger DJ

Vector Survival Direct effects of temperature on mortality rates* Temperature effects on development: at low temperatures, lifecycle lengthens and mortality outstrips fecundity* * Non-linear (quadratic) relationships with temperature Relationship between temperature and vector mortality is quadratic: mortality rates increase at high and low temperatures. Temperature effects on development may affect mortality rates: particularly high rates of development of mosquitoes can result in small adults with poorer survival. This is one example where the terms in epidemiological models of VBDs interact with one another. Another important interaction is the dependence of transmission coefficients for tick-borne pathogens on the numbers of vectors feeding on the host. The understanding of such interactions is, however, largely rudimentary. When relative humidity is low, ticks have to make more frequent, energy-expensive trips to the litter layer to rehydrate. High “monsoon” rainfall knocks ticks off the herbage and prevents them from finding a host. Lower humidity ↑ the energy requirement for host seeking by ticks shortening their lives.* Lower rainfall ↓ breeding areas for mosquitoes, compounded by density-dependent intraspecific competition amongst larvae. More complex community-associated changes (habitat structure, predator abundance). Tsetse mortality, Rogers and Randolph, 2003

Vector and Host Seasonality Vector-borne zoonoses mostly maintained by wildlife Humans are irrelevant to their ecology Vectors and their hosts are subject to seasonal variations that are climate related (e.g., temperature) and climate independent (e.g., day-length) Seasonal variations affect abundance and demographic processes of both vectors and hosts Many VBDs are zoonotic and have life cycles that are fully maintained in wildlife. In these diseases, seasons often play a very important role in the relationships between vectors and hosts.

Vector and Host Seasonality (cont.) Vector seasonality due to temperature affects development and activity → transmission Host demographic processes (reproduction, birth and mortality rates), affected directly by weather and indirectly by resource availability → VBD epidemiology Both vectors and hosts have seasonal variations in their life cycles driven by seasonal changes in climate and climate independent effects such as day length. Vectors can be affected by the way in which temperature can change from season to season, with resultant impacts on their development, activity, and disease transmission role. The lifecycle and activity level of the host can be affected as well, affecting how fast infected or immune animals die and how fast uninfected animals are borne, with resultant impacts on the epidemiology of vector-borne zoonoses.

Evidence Reviewed by the IPCC Emerging evidence shows: Altered the distribution of some infectious disease vectors (medium confidence) Altered the seasonal distribution of some allergenic pollen species (high confidence) Increased heatwave-related deaths (medium confidence) Evidence for climate change has been controversial as we need to rule out, or account for, the effect of climate-independent factors before ascribing climate as a determinant of observed changes in VBD. Trend analysis has been nagged by the scarcity or consistency of long-term health records and numerous confounding factors. Despite these difficulties, some studies have developed innovative ways to examine long-term data and provide some evidence of the effects of climate warming on human health. The latest IPCC report has stated with medium confidence that evidence exists showing an altered distribution of some infectious disease vectors. IPCC AR4, 2007

Evidence of Climate Change Effects Some specific disease examples: Malaria — East African highlands Lyme disease — Canada Schistosomiasis — China Bluetongue Europe Source: CDC Source: USDA Source: Davies Laboratory Source: DEFRA In the next section, we will be looking at evidence of climate change effects. We will review some specific examples that provide some potential evidence of the effects of climate change on VBD. Our examples will include malaria, lyme disease, schistosomiasis, and bluetongue disease.

Evidence: Malaria in Kenya Kenya Division of Malaria Control, 2009 Highlands Endemic Malaria Image source: CDC A study by Pascual et al. (2006) reviewed temperature data for the past 50 years in East Africa to examine the role of climate in exacerbating incidence of endemic malaria in the Eastern highlands of Kenya where increases in malaria have been observed since the 1970s. Their analysis found evidence for significant warming at all sites and an applied dynamic model suggested that biological responses, such as those by the vector and pathogen, would also be magnified by at least 1 order of magnitude under climate warming. The map you see in the slide shows the different areas of Kenya and the different incidence rates. The bright red portion of the map shows an area with endemic malaria. The pink area on the coast also shows an area with endemic malaria. The aqua section of the map that abuts the red endemic area is the “Highlands” area, where incidence has been increasing. Legend Arid/Seasonal Endemic Coast Highland Lake Endemic Low risk

Evidence: Lyme Disease  2007 Example of the emergence of Lyme disease in Canada. It is very difficult to collect the data necessary to make inferences directly linking climate change to VBD patterns. As we will discuss in a little more detail in the case study, the tick vector which carries the Lyme disease bacteria interacts with a number of different animal reservoir hosts, across multiple seasons. Increasing climatic suitability is predicted to increase with climate change. Empirical studies indicate that suitable habitat exists north of the current range of the tick vector, and that hosts are already dispersing the tick northwards. Therefore, a number of climatic and climate-independent factors are operating simultaneously to facilitate the spread of the tick. This is the first in a series of three slides that shows how Lyme Disease has continued to spread, as indicated by the red dots and triangles across Canada over time. Climate change is one key factor in this spread. Source: USDA Ogden et al., 2006a 1970  1997

Evidence, Lyme Disease: Canadian Locations as of 1997 This is the second in a series of three slides that shows how Lyme Disease has continued to spread, as indicated by the increase in the number of red dots and triangles across Canada over time (compare with previous and following slide). Climate change is one key factor in this spread. Source: USDA Ogden et al., 2006a

Evidence, Lyme Disease: Canadian Locations as of 2008 This is the second in a series of three slides that shows how Lyme Disease has continued to spread, as indicated by the increase in the number of red dots and triangles across Canada over time (compare with previous two slides). Climate change is one key factor in this spread. Source: USDA Ogden et al., 2006a

Temperature change from 1960s to 1990s Evidence: Schistosomiasis Temperature change from 1960s to 1990s 0.6-1.2oC 1.2-1.8oC Freezing zone 1970-2000 Freezing zone 1960-1990 Yang et al., 2005 Baima lake Hongze lake Planned Sth-to-Nth water canal The next example looks at Schistosomiasis in China. There has been a northwards extension of potential transmission (limited by “freezing zone”), in Jiangsu Province, due to a rise in the average temperature in January since 1960. The study by Yang et al. (2005) noted an increase in the reported incidence of Schistosomiasis over the past decade which may reflect the recent warming. The northwards expansion of the “freeze line” (which limits survival of water snails) puts 21 million extra people at risk. Yangtze River Shanghai Source: Davies Laboratory

Evidence: Bluetongue Disease Culicoides midge range previously restricted by Spain (south), Portugal (west), Greek islands (east) Now spread across southern Europe including France and Italy and moving northward Spatial congruence between Bluetongue incidence and climate changes support link Purse et al., 2005 Temperature change: 1980s vs. 1990s Culicoides biting midge Source: DEFRA Bluetongue is a devastating disease of ruminants spread by a midge vector. Cattle are the main reservoir of the disease. The biting midge Culicoides imicola was, until the late 1990s, restricted to the warmest corners of Europe: southern Spain and Portugal to the west, and the Greek islands to the east. In recent years, the insect has spread across much of southern Europe, including France and Italy, together with a significant move northwards. A devastating epidemic of bluetongue in southern Europe has followed in its wake. The spread of bluetongue and its vectors presents some of the strongest evidence to date that climate change is driving VBDs into new regions, as warming and disease spread have occurred at the same times in the same places. An incursion of Bluetongue virus (BTV) occurred in the UK in 2007. Further climate change is predicted to drive bluetongue risk even further north, putting the UK in 2030 at risk of widespread endemic disease. 26

Summary of Climate Change Effects Climate change has the potential to Increase range or abundance of animal reservoirs and/or arthropod vectors Lyme, Malaria, Schistosomiasis Prolong transmission cycle Malaria, West Nile virus, and other arboviruses Increase importation of vectors or animal reservoirs Dengue, Chikungunya, West Nile virus Increase animal disease risk and potential human risk African trypanosomiasis As a quick recap: The major ways in which climate change is likely to impact VBD include Increasing the range or abundance of animal reservoir and arthropod vectors Prolonging the transmission cycle of disease Increasing the likelihood of successful importation of disease vectors or animal reservoirs Increasing the animal disease risk and potential human risks of disease.

Emerging\Re-emerging Infectious Diseases Introduction of exotic parasites into existing suitable host/vector/human-contact ecosystem (West Nile) Geographic spread from neighbouring endemic areas (Lyme) Ecological change causing endemic disease of wildlife to “spill-over” into humans/domesticated animals (Lyme, Hantavirus, Nipah) True “emergence”: evolution and fixation of new, pathogenic genetic variants of previously benign parasites/pathogens (HPAI) Another way in which climate change can have an affect on VBD is through emerging infectious diseases, as well as re-emerging infectious diseases. There are different ways in which an infectious disease can be said to “emerge” or “re-emerge.” One way is through the introduction of exotic parasites into a suitable vector population that has contact with a suitable host population. An example of this is WNV. Another type of disease emergence or re-emergence is the spread of endemic transmission of a disease from one area to a new area that did not previously have endemic transmission levels. An example of this is Lyme disease moving from the United States into Canada. Other forms of disease emergence include ecological changes which might cause a disease to move or “spill over” from an animal population into a human population, effectively introducing a new pathogen into a suitable population. A fourth potential form of disease emergence involves genetic changes in pathogens. These mutations can create “new” variants of diseases.

Case Study I: Malaria Now we will move onto discussing several case studies of VBDs and how their transmission can be affected by climate. The first case study that we will discuss is malaria.

Case Study I: Malaria (cont.) Estimated incidence of clinical malaria episodes (WHO) 40% world population at risk 500 million severely ill Climate sensitive disease1 No transmission where mosquitoes cannot survive Anopheles: optimal adult development 28-32ºC P falciparum transmission: 16-33ºC Highland malaria2 Areas on the edges of endemic regions Global warming  El Niño3 Outbreaks McDonald et al., 1957 Approximately 40% of the world’s population lives in areas at risk for malaria. Every year about 500 million people become severely ill from malaria. Between 700,000 and 2.7 million – mostly children in sub-Saharan Africa – die each year of malaria. Malaria is an extremely climate sensitive disease. Clearly transmission does not occur in climates where mosquitoes cannot survive. Optimal larval development occurs at 28°C and optimal adult development between 28 and 32°C. Transmission cannot occur below 16°C or above 33°C as sporogony (the production of sporozoites which comprises dissemination and development of the parasite in the vector) cannot take place. The effect of global warming on malaria may be felt most in areas that are currently on the edges of the range of infected mosquitoes (Patz and Olson, 2006). These include many of the densely populated highland regions in Africa that are surrounded by lowland areas where malaria is endemic. Small changes may therefore lead to the exposure of many people to malaria. Many global warming scenarios include an increase in the frequency and intensity of the El Niño phenomenon (Patz et al., 2002) such as storms, heavy rain, droughts, and warm temperature. El Niño seasons have been associated, although not always, with outbreaks of malaria in many areas (ref Atul). Therefore it seems reasonable to speculate that the intensification of El Niño effects due to global warming will facilitate local epidemics of malaria. 1 Khasnis and Nettleman 2005; 2 Patz and Olson 2006; 3 Haines and Patz, 2004

Malaria Transmission Map This is a map of the current distribution of malaria transmission. Transmission zones are highly dependent on climate. Localized transmission is affected by regional factors (such as type of vegetation, health services, vector control), but the global ranges are predominantly affected by climate. The key areas of importance for climate change will be changes (loss or appearance) of incidence at the margins of these ranges, where climate suitability for the vector and/or pathogen are currently marginal, and where small shifts in climate may push the transmission potential of the disease above or below the required threshold. WHO, 2008b

Transmission Cycles of Malaria The transmission cycle of malaria follows the human-vector-human model of VBD transmission. A person gets malaria from the bite of an infected female mosquito. The mosquito bite injects young forms of the malaria parasite into the person's blood. The parasites travel through the person's bloodstream to the liver, where they grow to their next stage of development. In 6 to 9 days, the parasites leave the liver and enter the bloodstream again. They invade the red blood cells, finish growing, and begin to multiply quickly. The number of parasites increases until the red blood cells burst, releasing thousands of parasites into the person's bloodstream. The parasites attack other red blood cells, and the cycle of infection continues, causing the common signs and symptoms of malaria. When a non-infected mosquito bites an infected person, the mosquito sucks up parasites from the person's blood. The mosquito is then infected with the malaria parasites. The parasites go through several stages of growth in the mosquito. When the mosquito bites someone else, that person will become infected with malaria parasites, and the cycle will begin again. Malaria parasites can also be transmitted by transfusion of blood from an infected person or by the use of needles or syringes contaminated with the blood of an infected person. (Directors of Health Promotion and Education)

Particularly vulnerable: children, pregnant women Climate Impacts on Malaria What are some of the potential direct and indirect pathways of influence? Human Pathogen Environment Vector Anopheles mosquitoes Plasmodium Temperature Water availability Humidity Particularly vulnerable: children, pregnant women This slide will examine the role of climate impacts on the vector, the pathogen, and humans. Vector: The malaria vector is the Anopheles mosquito. Several key transmission variables can be affected by climate. Most important are the climatic requirements for survival. Different species require different temperature ranges. Key vector factors: Climatic requirements for survival Temperature ranges for different species Standing water and humidity requirements Insecticide resistance (e.g., DDT). Pathogen: The malaria pathogen is the Plasmodium parasite. Common species include P. falciparum, P. vivax, and P. malariae. The pathogen requires a certain temperature for reproduction. Human population: The human population can be affected by climate and the environment as well. Climatic change can affect the ability of the human population to access medical treatment. Climate can also affect patterns of human movement, contributing to the spread of transmission to new areas. Key human population factors: Poverty and other social determinants of health Acquired immunity Access to medical treatment Resilience capacity: knowledge of how environmental factors affect malaria (i.e water management).

Competent Vectors Kiszewski et al., 2004 Competent vectors are found in many different regions of the world. However, social factors play a large role in malaria distribution as well. Many areas that have competent vectors do not currently have malaria transmission, due to the role of social and economic variables. Some of the pertinent social and economic variables include political systems and political stability, socioeconomic conditions, access to health care services for prevention and treatment, health care education information, and general public health infrastructure. These factors will be discussed more in the next slide. Kiszewski et al., 2004

Malaria Endemicity (Current) The malaria stability index is a measure of potential transmission, based on the intrinsic factors of the pathogen and mosquito species, and prevailing climate. Malaria is more “stable” in Africa because the mosquito species there are the most competent vectors of the plasmodium parasite (and because the climate is appropriate). This shows that there is a difference between “potential” malaria transmission (based on broad ecological factors), and “actual” transmission, based on local ecologies and social factors. What might some of these “local” or “social” factors be? Local pools of water and mosquito habitat Urbanization Altitude Medical services Vector control Population exposure to vectors Population movement. “Climate change related exposures... will have mixed effects on malaria; in some places the geographical range will contract, elsewhere the geographical range will expand and the transmission season may change (very high confidence).” Kiszewski et al., 2004

Projections for Malaria Now we will move onto discussing several case studies of VBDs and how their transmission can be affected by climate. The first case study that we will discuss is malaria. The effect of climate change on malaria remains in debate There will likely be areas of decline and areas of emergence Impacts are likely to depend on localized factors and a combination of climate and socioeconomic conditions Risk will increase the most on the fringes of malarial transmission, but control is generally good in these areas Impacts will likely remain highest in currently endemic areas, where control is poor and vulnerability is high.

Recent Example: Improving Malarial Occurrence Forecasting in Botswana From annual time-series data: statistical relationship between summer (Dec-Jan) rainfall and post-summer annual malaria incidence (Thomson et al., 2006) Model applied, with good success, to previous meteorologically-modeled forecasts of summer rainfall This extended (by several months) the early-warning of post-summer malaria risk One recent example of disease incidence projection is that of malaria in Botswana. A study by Thomson et al. (2006) suggested that there was a potential relationship between summer rainfall and subsequent annual malaria incidence in Botswana. This study looked at annual time series data and found a statistical relationship between summer rainfall (December to January) and post-summer annual malaria incidence. This model was applied with relatively good success to previous meteorologically-modeled forecasts of summer rainfall and has predicted an extension of the post-summer malaria risk.

Malaria Projection: 2050 P. falciparum Biological model Other projections have focused on modeling disease risk based on changing habitat suitability of vectors and pathogens. A study by Martens et al. used simplified analytical techniques to model future malaria risk and concluded that climate change could greatly increase the number of human cases as a result of potential increases in the geographical range and altitude of disease vectors. Their approach was based on potential ranges of transmission, as determined by the transmission cycle and requirements of the parasite and vector. The main criticism of this model by Rogers and Randolph (2000) is that it is the relative, not absolute, increase in disease transmission that matters and that these models fail to take socioeconomic conditions and other factors which currently constrain the range of disease occurrence. Martens et al. 1999 Martens et al., 1999

Malaria Projection: 2050 Based on current distributions (statistical model) Rogers and Randolph (2000) similarly modeled changing habitat suitability for malaria transmission but predicted limited or negligible spread resulting from climate change. As per their criticism of the Martens et al. model, their model was based on actual/current ranges of transmission with the assumption that current distributions reflect “real” risk. However, because the assumptions that constrain current malaria risk in marginal areas are predominantly climatic, criticism of this model reveals concerns that it may under-predict future malaria risk. Rogers and Randolph, 2000

Climate Change and Potential Malaria in Zimbabwe: Baseline 2000 The next three slides show how malaria transmission in Zimbabwe has been projected to change over the next 50 years or so. Ebi et al., 2005

Climate Change and Potential Malaria in Zimbabwe: 2025 Projection Baseline 2000 2025 2050 Ebi et al., 2005

Climate Change and Potential Malaria in Zimbabwe: 2050 Projection Baseline 2000 2025 2050 There is a clear trend of increasing climate suitability for the transmission of malaria. Ebi et al., 2005

Case Study 2: Lyme Disease

Transmission Cycle of Lyme Disease Lyme disease is caused by a spirochete bacterium of the Borrelia burgdorferi species complex that is transmitted by tick vectors in the Ixodes ricinus complex. B. burgdorferi is maintained in the environment in a cycle between tick vectors and mammalian hosts across multiple seasons, with occasional accidental transmission into humans in whom it causes Lyme disease. The tick vector feeds on a blood meal once during each of its three life cycle stages: larva, nymph, and adult. Larval ticks bite actively in midsummer after which they drop off their host, spend a long period developing in the surface layers of the soil, and eventually molt into the nymphal stage. Nymphs overwinter until the following spring or early summer when they will emerge and seek a host to feed on. Once fed, the nymph drops from the host, spends a long period developing in the surface layers of the soil, and then molts one final time into the adult stage after which it will feed once again on a mammalian blood meal before reproducing and dying. Engorged adult females undergo a prolonged period of incubation in the environment prior to laying their eggs, and then the eggs undergo a long period of development before hatching into larvae. The protracted periods the ticks spend off the host means that much of the tick’s life is spent in the environment and exposed to ambient temperature and humidity conditions. Larval and nymphal stages maintain transmission cycles of the bacteria because both tick stages feed on the same hosts. Because of their size and the timing of their biting season, nymphs are the stage most likely to transmit the disease to humans, though infected adult ticks can also transmit the disease to humans. Stafford, 2007

Lyme Disease Distribution in the Unites States of America I. pacificus I. scapularis Ixodes scapularis is the primary tick vector present in the northeastern US, while Ixodes pacificus is the primary vector in the western US. If the geographic range of I. scapularis ticks moves northwards as is expected of many arthropods (Vickery, 2008), then this range expansion may result in Lyme disease becoming endemic in the most heavily populated parts of southeastern Canada (note that I. pacificus is already endemic in western Canada). I. scapularis is more adaptable being a generalist species in terms of its hosts and its habitat preferences. Generalists are more likely to be able to expand their range in response to climate change than are specialist species (Braschler and Hill, 2007).

Passive Surveillance: Migratory Bird Distribution of Ticks (I Passive Surveillance: Migratory Bird Distribution of Ticks (I. Scapularis) Ogden et al., 2006a, 2008

Hypothesis: Migratory Birds Carry I Hypothesis: Migratory Birds Carry I. scapularis Into, and Through, Canada Northern-migrating ground-feeding birds stop-over in tick-infested habitat Spring migration coincides with spring activity period of Ixodes scapularis nymphs Even if climate expands the geographic range of an environment suitable for vectors and VBDs, there has to be some physical movement of vectors into these new areas for range expansion to occur, and this process is not inevitable. In this case, the tick vectors are being moved by migratory birds moving north in springtime. The midge vectors of BTV in the Mediterranean basin moved northward due to midges being carried long distances by prevailing winds so the BTV was able to establish in more northern areas as a consequence of a warming climate (Purse et al., 2005). Nymphs feed continuously on birds for 4-5 days, then drop off into the habitat

Projections for Lyme Disease

Prediction of Potential Extent of I. scapularis Populations at Present The risk maps provide a prediction for where tick populations may be now (using actual recent meteorological data). Ogden et al., 2008

Validation of the Risk Maps Field validation indicates that the tick vector is emerging predominantly in the high risk areas of the risk maps. Significantly, when accounting for other factors determining tick population establishment (habitat and dispersion by hosts), model-derived relationships between temperature and suitability for tick population establishment have a strong relationship with the observed pattern. This provides field evidence for temperature to be a determinant of the expansion of the range of I. scapularis, and that recent warming could be driving recent emergence of this tick in Canada (Ogden et al., 2008). Ogden et al., 2008

Prediction of Potential Extent of I. Scapularis Populations by 2049 The field validation provides confidence that we can then develop risk maps for future climate scenarios (here using IPCC emissions scenario A2 to parameterise the general circulation model CGCM2). Ogden et al., 2008

Prediction of Potential Extent of I. Scapularis Populations by 2079 Ogden et al., 2008

Prediction of Potential Extent of I. Scapularis Populations by 2109 Ogden et al., 2008

Case Study 3: Dengue The next case study that we will discuss is dengue.

Climate Variability and Dengue Incidence Aedes mosquito breeding (Argentina)1: Highest abundance mean temp. 20ºC, ↑ accumulated rainfall (150 mm) Decline egg laying monthly mean temperature <16,5ºC No eggs temp. <14,8ºC Other studies: Virus replication increases ↑ temperature2 Transmission of pathogen ≠ >12ºC3 Biological models: small ↑ temperature in temperate regions  increases potential epidemics4 Dengue is an important mosquito-born disease, with about 2.5 billion people at risk worldwide. The Aedes spp. mosquito vectors are well adapted to the urban environment and thrive well in a warm, humid environment. Viral replication in the vector increases with temperature, with expected temperature-related effects on transmission. Minimal transmission temperature for the dengue virus is 12°C. Dengue hemorrhagic fever (DHF) outbreak in southern Sumatra was accompanied by more extreme weather due to El Niño effects (Corwin et al., 2001). Linked to future climate change projections, a small rise in temperature in temperate regions will increase the potential for future epidemics, given a susceptible population and introduction of the virus. 1Vezzani et al., 2004; 2Watts et al., 1987; 3Patz et al., 2006; 4Patz et al., 1998

Dengue Transmission Map This is the current area at risk of dengue transmission. The lines represent January and July isotherms and demarcate the area where the vector for dengue exists. Climate change is likely to shift these isotherms farther north above the equator and farther south below the equator, thereby potentially increasing the size of the area at risk of dengue transmission. WHO, 2008b

Transmission Cycle of Dengue The typical transmission cycle of dengue follows the human-vector-human model, similar to malaria. However, there is also the potential for dengue to move from an animal transmission cycle into a human transmission cycle. Whitehead et al., 2007

Example of Weather Effects: El Niño Global warming intensify El Niño Several studies found relationships between dengue epidemics and ENSO (El Niño Southern Oscillation) Drought conditions: increase water storage around houses  elevated Aedes aegypti populations Enhanced breeding opportunities when rainfall accumulates following drought (Kuno et al., 1995) Several studies have found relationships between dengue epidemics and ENSO (El Niño Southern Oscillation). ENSO is a global scale pattern of climate variation that accounts for up to 40% of temperature and rainfall variation in the Pacific. Both drought conditions and the rainfall accumulation following a drought contribute to augmentation of the vector population. ENSO= global scale pattern of climate variation accounting for up to 40% of temperature and rainfall variation in Pacific Hales et al., 1999

Case Study 4: African Trypanosomiasis T. b. gambiense T. b. rhodesiense The last case study that we will discuss is African trypanosomiasis, commonly known as “sleeping sickness” in humans.

Source: David Rogers, Oxford Case Study 4: African Trypanosomiasis (cont.) Trypanosomiasis Trypanosomosis, spread by tsetse flies, imposes a huge burden on African people and livestock Many aspects of the vectors’ life cycles are sensitive to climate, and spatial distributions can be predicted using satellite-derived proxies for climate variables African trypanosomiasis is found only in Africa, where it contributes a large disease burden to both people and livestock. WHO estimated that in 2000 that some 50 to 60 million people in Africa were exposed to the bite of the tsetse fly. At that time WHO considered that close to 300,000 children, women, and men on the African continent were affected by the disease, a figure which is much larger than the 27,000 cases diagnosed and treated that year (WHO, 2008a). The vector is the tsetse fly. The vector is sensitive to climatic factors. Source: David Rogers, Oxford

African Trypanosomiasis Distribution T. b. gambiense T. b. rhodesiense This map shows the distribution of African trypanosomiasis. The line divides the areas where two different species of the vector are found. WHO, 2008a

African Trypanosomiasis Transmission T.b. gambiense The transmission of African trypanosomiasis follows an animal-vector-animal transmission cycle with occasional spillover into humans. T.b. gambiense sleeping sickness is transmitted from human to human by the tsetse fly and is the most common form of transmission of this disease. As some animals can host the human pathogenic parasite, transmission can occasionally take place directly from animals to humans, which is believed to be one of the potential mechanisms of the long-term maintenance of the disease in endemic areas. The transmission of T.b. rhodesiense sleeping sickness primarily involves domestic and wild animals. However, intensified human to human transmission occurs during epidemics. The effects of climate change on African trypanosomiasis have not been extensively studied to date, but it is likely that indirect climate change effects on human and animal health will exacerbate the impact of this disease in areas where it is endemic. (WHO, 2008a) T.b. rhodesiense

Different Approaches to Modeling Will climate change affect VBD risk? Focus has been on human-vector-human transmitted diseases (e.g., malaria and dengue) Results of simplified modeling (e.g., Patz et al., 1998; Martens et al., 1999) Climate change could greatly increase numbers of human cases (increase geographic range and altitude) Results of statistical pattern matching (e.g., Rogers and Randolph, 2000) Climate change could have a small effect on numbers of human cases (small changes to geographic range/altitude) There is some debate regarding the results of projection modeling, and at times it can be difficult to answer the question, “Will climate change affect vector-borne disease risk?” As vector-borne diseases often have complex transmission cycles and a multitude of factors which can affect the risk of disease, it is difficult to untangle the effects of climate change from other effects. Many approaches to modeling disease risk have been proposed and a number of studies have attempted to focus on the simpler of the two main transmission cycles: the human-vector-human cycles as shown with Malaria projections previously. Studies by Patz and Martens have used simplified modeling techniques and concluded that climate change could greatly increase the numbers of human cases as a result of potential increases in the geographical range and altitude of disease vectors. Studies by Rogers and Randolph have used statistical pattern matching approaches and concluded that climate change will likely have small effects on the numbers of human cases as a result of small changes to the geographical range or altitude of disease vectors. Simple analytical models Usually predict significant spread due to climate change Are based on potential ranges of transmission, based on the transmission cycle and requirements of the parasite and vector Likely over-predict because they poorly reflect local and socioeconomic conditions and their evolution, and poorly model current distributions Statistical models Usually predict limited or negligible spread due to climate change Are based on actual/ current ranges of transmission (assume that current distributions reflect “real” risk) May under-predict because they assume that constraints in marginal areas are predominantly climatic Most models are subject to particularly high error in peripheral edges of transmission ranges, where climate is marginal and where socioeconomic factors will play a significant role in determining malaria risk. Models generally focus on endemic malaria, and rarely consider sporadic transmission. Even if total area/people at-risk remains the same (i.e., some areas decrease and some increase), new at-risk regions may be less prepared to deal with transmission than historically-affected regions.

Limitations of Statistical Models Data quality and potential misclassification Explanatory variables climatic, land use (NDVI) and Fourier transformations (data dredging?) Pattern matching using “known” current distribution does not = “ecological” niche Ecological niche + societal-human factor → potential misclassification (false negatives) When evaluating model predictions, important to keep in mind that all modeling methods have their frailties due to: Oversimplification of the system Knowledge gaps Poor data Uncertainty that statistical associations = causal relationships.

Limitations of Statistical Models (cont.) Cannot use this model to obtain climate change projections and say that the effects of climate change are negligible Need to model climate change effects on ecological and societal-human factors simultaneously The two approaches shown previously (simple analytical model and statistical pattern matching) demonstrate the different range of projected climate change related risk on human-vector-human VBD.

Future Outlook? Two approaches (simple analytical model and statistical pattern matching) show different projected degree of effect of climate change on human-vector-human VBD risk The ideal is mechanistic models of transmission but these require a high number of parameters and detailed knowledge of the ecology of the diseases Both are useful techniques in assessing risk, but for human-vector-human VBD we need more “layers” The ideal modeling approach would be mechanistic models of transmission but these require a high number of parameters and detailed knowledge of the ecology of the diseases. Simple analytical model and statistical pattern matching techniques are both useful in assessing risk, but to accurately model human-vector-human VBD, we need more “layers.” .

Future Outlook? (cont.) Both techniques may be more useful (side-by-side) for projections of risk of VBD We need to develop risk maps using the precautionary principle (worst case) and overlay these with mitigating factors or conservative estimates We need to develop risk maps using the precautionary principle (worst case) and overlay these with mitigating factors or conservative estimates. Both techniques may be more useful (side-by-side) for projections of risk of VBD.

Perspective Can see potential associations with climate but causality difficult to confirm Need to consider non-climatic contributing factors Very long future time scale Data needed for accurate projections not readily available Further empirical field work required to improve projections Nevertheless, opportunities exist for adaptation Can see potential associations with climate but causality is difficult to confirm. Non-climatic contributing factors need to be considered as well (i.e., urbanisation, deforestation). Very long future time scale. Accurate projections difficult to make based on current knowledge and data available information to measure changing risk). Data needed for accurate projections not readily available. Nevertheless, opportunities exist for adaptation. Opportunities for adaptation that do not require detailed predictions are possible now.

Opportunities for Adaptation Surveillance Precautionary approach Mainstreaming response Enhancing health system capacity Anticipating new and emergent pathogens changing VBD burden Adaptation measures which can be implemented to reduce current and future disease threats include: Strengthening surveillance and public health Adopting precautionary approaches in health planning and disease monitoring Mainstreaming response to disease threats Enhancing health system capacity to handle current and anticipated future disease risks Anticipating the potential for new and emergent VBD pathogens and their potential to change the current VBD burden.

A New Approach to Risk Assessment Surveillance/control applied in retrospect (= too late?) Recognition & diagnosis Response to epidemic Pathogen emerges Disease in humans Forecasting Prediction Surveillance Risk assessment Risk identification Intervention Human disease prevention As we conclude, it is important to think of how prediction of future risk as related to climate change and vector borne disease can be used in the public health practice. Modeling future projections can be an incredibly useful tool for the public health practice as it can assist in the planning of resources for surveillance and intervention to reduce the human risk of disease. The top row indicates what has often happened in the past (and will likely occur in the future in many cases). As pathogens emerge and are recognized as capable of causing disease in humans, surveillance and control efforts are deployed to mitigate the disease epidemic. The bottom row indicates what ideally we should aim for. By assessing the risk of potential disease through the identification of potential pathogens and routes of disease spread, can generate forecasts of disease risk areas and implement surveillance efforts to monitor spread and plan intervention efforts in order to reduce the risk of human disease. The red circle identifies the scientifically difficult phase – modeling, laboratory, and field studies combine at this point.

Adaptations Include Precautionary approach to risk assessment Increased surveillance and monitoring (baseline + changing incidence) Improved tools for integrative risk assessment “Mainstreaming” through increased health system capacity Preparedness for new and emergent pathogens Despite the difficulties related to accurately predicting the changing risk of VBD, it is important to proceed with the implementation of adaptation measures on the ground now by adopting: Precautionary approach to risk assessment Increasing surveillance and monitoring of disease threats (baseline + changing incidence) Improving tools for integrative risk assessment “Mainstreaming” through increased health system capacity Continuing to prepare for new and emerging pathogens.

Future Directions Human infections are intricately linked to the global environment, and we should be aware that climate change has significant potential to change the epidemiology of infectious disease Physicians and health care planners need to be aware of these changing risks Study: multidisciplinary approaches Invite new partners Human infections are intricately linked to the global environment, and we should be aware that climate change has a significant potential to change the epidemiology of infectious disease. Physicians and other health care planners need to be aware of these changing risks. Future directions for research and practice related to climate change and vector borne disease will need to continue to include multidisciplinary approaches, by inviting new partners to contribute to the discussion and the debate.

Conclusions Climate change will affect the distribution and incidence of VBD globally Impacts will vary from region to region Current evidence suggests impacts on some diseases may already be occurring Risk assessments constrained by complex transmission cycles and multiple determinants In conclusion: It is likely that climate change will affect the distribution and incidence of VBD globally As stated by the IPCC, impacts will vary from region to region Current evidence uncovered to date suggests that impacts on some diseases may already be occurring Risk assessments are currently constrained by the complex transmission cycles of diseases and the multiple determinants of disease risk.

Conclusions (cont.) Current models produce differing results Non-climatic factors remain important determinants of risk Impacts may include unanticipated emergence of new pathogens In addition: The currently available models have produced differing results It is important to keep in mind that non-climatic factors remain important determinants of risk Some of the potential impacts of climate change may include the unanticipated emergence of new pathogens. Recommended further reading: 2007 IPCC report: The Physical Science Basis, FAQ 3.1 page 103 and FAQ 3.2 page 105