h¯
° '
f , f Z ¾ fn
Towards a global health research agenda on “CLIMATE CHANGE AND HEALTH”
An international research seminar was arranged at the 24th of November 2008, by Umeå Centre for Global Health Research, Umeå University. The possible impacts on global health from climate change were in focus of the presentations and discussions. Below the different presentations are summarized, followed by concluding remarks.
Professor Stig Wall, Epidemiology and Public Health Sciences, Umeå University, introduced the seminar by pointing at two central research
directions within the Umeå Centre for Global health Research, valid for the issue of the day. One was the development of theory and methodology for the practice and governing of global public health. The other was the assessment of health systems facing new health threats. The Centre has a strong infra structure for global health research with more than 2 000 000 people in surveillance systems and 38 field sites in 19 countries, making research possible where it is most needed. Professor Wall concluded that climate change and health has been identified as a theme to focus on within the Centre.
“All hands on deck: Climate change and global health” Rainer Sauerborn, professor of Tropical Public Health, Heidelberg University, Germany
If climate were the only thing changing, it would be easier to understand the links with health, to anticipate the health effects and manage them. But environmental, epidemiological, demographic and social changes intertwine, thus making the task to tease them apart a monumental one. In contrast to the likely and huge impact on health, the nexus of climate change and health has engaged few researchers worldwide. Indeed, the research base is fragmentary with huge gaps. By necessity, international reviews, such as the IPCC’s fourth assessment report, have been very cautious in assigning causality. This should remind us of the dictum: “the absence of evidence of an association is not evidence of the absence of an association”. Research has focused on problems in industrialized countries leaving climate impacts and health sector responses in developing countries largely un-researched. Another gap is the dearth of research on the impact of climate change on non-communicable diseases and on aging populations. In contrast, research on the direct health effects of climate change, e.g. heat waves, has come further with a growing body of scientific evidence, already influencing policies. Climate change impacts on agriculture in (sub-) tropical countries will threaten future harvest yields in developing countries and hence the nutritional status particularly of children. Standardized health data of high quality is available for example in about 30 INDEPTH sites. Health data and meteorological data can be traced back for at least 10 years, which allows studying the climate-health links retrospectively. At the same time we should use such longitudinal health research settings to add environmental and meteorological data collection. Retrospective, current and prospective data collection should be used to corroborate the causal links between climate change and health and to study the effects and costs of adaptation strategies. This is an important part of research for global health. In addition, teaching and training are of highest priority in order to win over and prepare young minds to study the issues. Teaching must definitely include new research methods, such as various types of modelling and forecasting, and contents from other than health fields, such as meteorology, agriculture, economics and more. Both the health impact of climate change and the capacity to deal with it are inequitably distributed around the globe, to the detriment of developing countries. It is therefore important for research and training to identify particularly vulnerable population groups and propose measures to address inequities within and between countries. This should include an assessment of the economic dimension of climate damage and the costs of health systems responses. The poorer a population, the more it actually lives on its environment. Protecting the environment is therefore an important strategy to protect the livelihood of the poor. Seen under this angle, mitigation of and adaptation to climate change are not only important from a public health vantage point, but are clearly poverty reduction strategies. In summary, there is an urgent need to increase research and teaching in this field. Health professionals should engage in the specifics of the climate change issues and policies. Protecting our and our children’s health is the unifying motivation of citizens to accept the costs of mitigation
and adaptation. Health researchers and professionals therefore need to develop the health arguments within this debate by improving the evidence base and communicating it effectively back to decision-makers at all levels.
“Climate change and health - Infectious diseases / Health consequences of climate dependent infections” Birgitta Evengård, professor, Div. of Infectious Diseases, Department of Clinical Microbiology, Umeå University
Basic facts on infectious diseases are that they remain being the second-leading cause of death world-wide and the third-leading cause of death in the US. There are 5-10 times more microbes living on and in every human being, than there are cells in our bodies. Microbes comprise 60% of the bio mass upon the planet. They have had 3.5 billion years to train adaptation to various environments and conditions, and adapts quickly. It is worth reminding that infections will continue to have an impact on human health which probably will get more severe through increased resistance to antibiotics. The climate change results in an increased risk for infections. One concrete example is raised temperatures making flooding more common impairing infra structure, resulting in inadequate cooling of food due to shortage of electric power or leakage of microbes into drinking water systems. In the north the Arctic region is the least researched considering climate change and health. In 2000 a Nordic society for circumpolar health was founded and there is also an international network on circumpolar health focusing on the health of the indigenous populations of the Arctic. The network has announced a working group on climate change and health. This is action driven by political will. Professionals and researchers should move up and go ahead. Except from the Sámi population the indigenous populations are described particularly vulnerable because of their close relationship to and dependence on the land, sea and natural resources. Direct threats are increasing extreme events followed by higher incidence of injury and mortality. The indirect effects include increased mental and social stress due to change of lifestyle, potential changes in bacterial and viral diseases, and access to quality water sources. Waterborne infections will become more frequent. Early signs of effects from climate change are seen in the outskirts of echo systems both at the northern borders and on high altitudes. Species of birds and insects have expanded north and vectors increase in numbers spreading infections as Borrelios and TBE. There is a change in biology with change happening on hotspots asking for awareness. Health status indicators are needed to create comparability of health status data between countries. A carefully planned strategy is needed at the community and regional level, to monitor and document environmental change – pilot studies, data analysis and evaluations. A critical need exists for the monitoring of wildlife diseases and human-wildlife disease interactions. Collaboration on regional and national levels is essential with surveillance systems and sharing of data. A change of climate, even if occurring in ”hot spots” does not consider borders. Success depends on the quality of leadership.
“Climate change and health in Europe – Collaborative projects and activities” Bertil Forsberg, Associate Professor of Epidemiology, Occup Env Medicine, Umeå University
We have since many years studied health effects of air pollution and in time-series studies controlled for weather to specify the independent effect of pollutants. More recently the effects of weather and climate change have become more interesting. For example, daily mortality increase with the ground level ozone concentration and the ozone production depends much on temperature. In time-series studies the health effects associated with season and weather (heat and cold) may always be potential confounders which must be adjusted for to find the true effect from air pollution. It is as important to understand how temperature more directly affects morbidity and mortality to accordingly create warning systems. Some networks and working groups have started collaboration and reporting on effects on health from climate change. We are a partner of the EU funded PHEWE (An European project on Assessment and Prevention of acute Health Effects of Whether conditions in Europe) which has published several papers on effects from hot and cold weather on mortality and morbidity with results from 15 European cities. A rise in temperature is found to increase respiratory hospital admissions significantly. We are also a partner in the national collaboration between FOI, Umeå University, Uppsala University and KTH funded by the Swedish Environment Protection Agency and called “Climatools”. A theoretical overview in Swedish called “Hälsopåverkan av ett varmare klimat” has been published. Our work will also result in a toolbox to decide on different adaptations for a new climate. As part of “Climatools” a “state of the art” document will be presented on climate change and health as a basis for vulnerability analyses. Epidemiological time-series analyses will be conducted to study the effects of temperature on mortality and hospital admissions in Sweden. A model will be developed and adaptation strategies and methods relevant to the health sector will be investigated and communicated. As experts for the Swedish Climate change assessment “Klimat- och sårbarhetsutredningen” we built a model to predict mortality related to a warmer climate. Economists estimated costs and revenues from climate change considering these health effects, but also effects on building constructions and needs for heating and cooling. The highest cost was estimated from increased mortality due to more warm days. We are also a partner in HENVINET, a European network funded under the EU 6th framework programme. The network aims at coordinating actions on environment and health with a special focus on four health endpoints: asthma and allergies, cancer, neurodevelopment disorders and endocrine disruptors. HENVINET has collaboration with the European Respiratory Society, and organized a workshop on climate change and health in Belgium in March 2008. One aim is to propose a long term agenda for research and advocacy on climate change and health. Areas of agreement and research gaps are to be identified but also disagreements amongst researchers made transparent. We are also a partner in a new EU project proposal called “Climatetrap” aiming at training, risk assessment and analyses of preparedness in the public health sector.
“INDEPTH, a dynamic environment”* Stephen Tollman, University of the Witwatersrand and Medical Research Council, Johannesburg, South Africa
Those who are affected the most have the least control. From a southern perspective the question is not simply what the priorities are but which crisis to focus on at the moment! Sweden is probably the largest system for longitudinal health and socio demographic surveillance in the world. Creating such infra structures in some of the more marginalised, vulnerable poorly resourced parts of the world is a challenge. It’s not just that health and population scientists have done relatively little. It’s also that the empirical evidence base is rather weak. It may be that the just published report from the WHO Commission on the Social Determinants of Health creates an opportunity to link several critical agendas - climate on the one hand but with health and equity on the other. The INDEPTH Network is a dynamic environment for creating an effective research infrastructure in middle and low income settings. Currently 37 sites in 19 countries are participating in the network: 25 sites in Africa, ten in Asia, one in Oceania and one in Latin America. They function both as independent R&D sites and as well-linked elements of the INDEPTH Network. Health and socio-demographic surveillance systems (HDSS) focus on geographically defined communities with entry registration at a specific point in time (baseline) followed by prospective follow up of the demographic, the health and the socioeconomic dynamics of these defined communities, including all vital events (births, deaths, migrations). These research platforms can be linked to other infrastructures, like geographic information systems, weather systems and climate systems. This ability to link the systems together may be very powerful: over time, researchers get the opportunity to think sequentially and thus the possibility to ask questions about cause – as well as build capacity. There is thus great potential for creating research and development infrastructures so as to ask questions on climate change and health. Of course, it is a challenge to understand the influence of climate change on ecological & social processes
such as health, biodiversity loss or migration/displacement. Examples of future research focuses in INDEPTH are: climate variability and malaria transmission (modeling and risk mapping); the effect of climate parameters on meningitis epidemics; environmental degradation, climatic change and food security; relationships between climate variability and the incidence of other infectious diseases; and capturing the effects of industrialisation on environment and health. The power of longitudinal comparison across multiple settings allows for interpretation and the ability to create models. Linked data platforms addressing population, environment and climate in low and middle income settings can provide much needed policy-relevant knowledge for mitigation and adaptation.
* International Network for the Demographic Evaluation of Populations and Their Health (www.indepth-network.org)
“The direct effects of high ambient temperatures on patterns of mortality and hospital admissions in Sweden” Joacim Rocklöv, Statistician, Occupational and Environmental Medicine, Umeå University
Heat exposure is associated with severe health consequences, especially in the elderly population. In the U.S. heat related mortality is the number one weather related killer. That compares to other weather related events like thunderstorms and tornados also causing increased morbidity and mortality. The heat wave in south and middle Europe in 2003 was associated with about 70000 excess deaths. During such events the most susceptible group are the elderly population, especially those living alone and not hospitalized. Numerous risk groups has been studied and shown to be susceptible when analyzing the patterns of mortality and morbidity associated with the heat wave of 2003 in, for example, France. Some of the susceptible groups are elderly women, people with certain medication, people living alone, people with heart- and lung disease and obese people. Also the rates of hospital admissions increase at extreme events, but have shown contrasting pattern at short-term variations of temperature. People seem to adapt to changes in temperature to some extent but adaptation probably also costs in terms of impaired health and excess mortality at least in the transition period of changing climate. There are results supporting that humans have an early programming to the climate they are brought up in, and will therefore be more susceptible to temperature exposure in the elderly years if climate have changed substantially. The general methodology used when studying the association between temperature and mortality and morbidity is Poisson time series regression with an ecological study design using contrasts in exposure over time with focus on short term variability in exposure and outcome. The shortterm exposure response relationship may not be sufficient to describe the effect of prolonged exposure, e.g. a heat wave. Here we need to include additional variables describing the risks during such events. To be able to study short-term variability the seasonality is removed from the time series. However, in the full climate change context some uncertainty is with this removal since we are uncertain that temperature is not the driving some seasonality. However, doing so we assume the determinants of seasonality are other than temperature like daylight and differences
in behaviour etc. The source of seasonality will be interesting to gain more knowledge of and is an area that has got increased attention in the climate change and health forums during the last year. There are direct effects of high ambient temperatures on mortality and hospital admission patterns, also found in Sweden. The most influential lag time is the first two days where immediate effects are seen. Global warming increase the probability of extreme heat events compared to a ”today’s normal” reference and might therefore constitute a public health threat among the elderly population. The effects are somewhat different but act similarly among different geographical regions of Sweden with the greater risks per degree C seen in the north. High levels of relative humidity and temperature may increase the heat load substantially, especially in larger cities. Also hospital admission increase strongly during extreme warm summer in Sweden and might put new demands and needs on the health sector in a future climate.
“Change: living conditions, diseases and climate” Peter Byass, Professor of Global Health, Epidemiology and Public Health Sciences, Umeå University
This is a slightly more philosophical view on the theme. Rather than presenting evidence, one can reflect on the nature of change and in particular interacting changes that work with or against one another. The political rhetoric of “change – we can believe in” seems very relevant also to climate change and health; the need for a change that people in general can believe in using arguments that convinces people in terms of what changes there will be in their lives. With a simple population-based view of the world, the three entities of climate, living conditions and health and diseases are somehow associated with one another in a complex cycle; they are not independent. Using such a model a change in one place changes everything. Climate affects living conditions and the other way around. Living conditions clearly affects diseases; changed climate is a threat to health etc. It’s important to have an overall, overarching view of the issue. It’s not meaningful to look at limited, separated parts leaving the rest out. An example of the complexity of global disease patterns changing can be illustrated by conducting a PubMed search on “malaria incidence”. It gave 211 abstracts over the last decade, of which 80 reported changes over time, 15% reported increase, 14% no change and 71% a decrease. Does this mean that there has been a change in
malaria incidence, a global decrease? There are 107 malaria endemic countries in the world but only 27 countries were represented in the 80 abstracts. There was a problem of representativity. Evidence must be handled carefully. To emphasise the importance of change, looking at the UN’s population predictions shows there will be massive changes by 2050. A massive change over two generations both in terms of total population numbers, in terms of the balance between the different parts of the world and in terms of the proportions between the different age groups. That is the back ground in population terms against which climate change, the changing patterns of disease etc. should be dealt with. The whole complexity must be kept in mind. Population or climate change can’t be dealt with in isolation; neither can the epidemiological transition in terms of changed patterns of disease. These are all interacting factors.
“The Puumula virus outbreak in Northern Sweden 2007” Magnus Evander, Associate professor of virology, Clinical Microbiology, Umeå University
There are specific types of hantaviruses with specific rodents as vectors. In Asia the virus is called Hantaan and in Europe/Sweden it is called Puumula. Hemorrhagic fever with renal (HFRS) syndrome is caused by hantaviruses transmitted from rodents to humans. Humans inhale infectious aerosols produced from rodent saliva, excreta, urine. Hantavirus infections are world wide and no specific treatment or safe and effective vaccine is available. There is more than 200,000 cases/year of HFRS and thousands die every year, more than 50% of cases happen in China. 250 patients/year are registered in South- and North America from Hantavirus cardiopulmonary syndrome (HCPS), with more than 40% mortality. The symptoms of Nephropathia epidemica (NE) caused by Puumula virus are acute fever, fatigue, headache, nausea, vomiting, myalgia, abdominal- and back pain, and visual disturbances. The majority has signs of renal failure. Incubation time is 2-6 weeks. The vector in Sweden is the bank vole. People more at risk are farmers and people working in the forests. In Sweden there are normally 150-500 cases every year but in 2007 there was a peak with 2200 registered NE cases. ~30% of diagnosed cases are hospitalized and the fatality rate in humans are <0.5%. In Sweden NE is endemic in the four counties in the north. NE disease incidence follow bank vole cycles; high incidence comes with high numbers of bank voles. In 2007 the Puumula outbreak in Sweden was sudden, unexpected and large, with more than half of the cases in the north. The outbreak meant a heavy burden on health care and society. In Norrbotten and Västerbotten county most NE patients had classical HFRS symptoms and displayed the whole spectra from mild to severe disease. One out of three required hospital care, occasionally dialysis and intensive care and there were three known lethal cases. Four out of ten Puumula infected was female and the highest proportion infected patients found in ages 35-78. Possible reasons for the outbreak in 2007 were a peak in the bank vole population and a mild winter with less snow and more ice. Climate effects on bank voles brings that they get no protection under the snow from predators, there´s no food and shelter from cold weather under the snow. This could lead to more bank voles than normal moving indoors with increased risk for infection on humans. Bank voles are sensitive to changes in
temperatures and climate which can give changed patterns of the spread of the disease and it´s prevalence. The NE trend over years in Västerbotten County is increasing.
“Using demographic surveillance to monitor population reactions to climate change” Anders Emmelin, Lecturer in Epidemiology, Epidemiology and Public Health Sciences, Umeå University
Climate change is a complex process with different expressions on the ground depending on altitude, latitude and degree of forestation/deforestation, among many other factors. Depending on the way local climate changes, the health consequences will vary. Changes in temperature and precipitation may change habitats for many infection vectors, reduced or untimely precipitation will influence food security negatively. On the other hand social and societal adaptation can modify the effects of climate change. The interactions with local practices and societal attempts to meet change will influence consequences locally. Not all consequences will be easily predictable. Demographic surveillance has the potential to be a strong tool in monitoring population reactions to climate change. In rural Ethiopia, the Butajira demographic surveillance site has recorded life events since 1987, events such as births, deaths, marriages, migration in/out in a sample of the population. Information about events has been collected through regular visits to each household by local interviewers. The altitude ranges from approximately 1,200 to 3,500 metres and within an area of 25*25 kilometres. There is great variation locally in climate conditions with secondary effects on agriculture. Data from the demographic surveillance showed high peaks in mortality for 1998 and 1999, mainly seen in the rural areas, both highland and lowland. Data on the population was combined with other data for the same period, such as rainfall data from the National Meteorological office local weather station in Butajira. Combining the data indicated a mortality pattern connected to rainfall – untimely rain and too little rain with severe effects on food security and health. The two mortality peaks were related to malaria (1998) and diarrhoea/malnutrition (1999). By the middle of 1999 it was recognised in the area that there was severe food shortage among the poorer households in the rural areas. It was the poor subsistence farmers that were affected the most. Health consequences of climate change should be monitored not only in affluent societies, but also among the most vulnerable. The INDEPTH network of demographic surveillance sites is already collecting population data in Africa, Asia and Latin America. These data need to be extracted and organised with a view to monitoring effects of climate change. Weather data, along with agricultural data, need to be added.
“Climate and meningococcal disease” Helena Palmgren, PhD, Infectious Diseases, Umeå University
Meningococcal disease is a global disease, endemic to almost all countries in the world. The disease is airborne without any vector involved in the transmission and has only humans as hosts. The disease is caused by the capsulated bacteria Neisseria meningitides or meningococci. Its manifestations are either bacterial meningitis, meningococcal septicaemia or a combination of both. The case fatality rate is between 70-80% untreated. With antibiotic treatment, the CFR is still high, 5-15% and another 5-10% of the patients suffer from lifelong neurological sequel. 220% of all investigated populations in the world carry Neisseria meningitidis on mucosal membranes in the upper respiratory tract. The role of meningococcal carriers in the epidemiology of the disease still needs to be better understood. Climate factors are important for both the distribution and the seasonality of meningococcal disease. Although the disease occurs all over the world, the high endemic area is the so called meningitis belt in sub-Saharan Africa The belt has been swept by epidemics for at least 150 years with an interval of every five to ten years. Historically the epidemics start in the east in Ethiopia, Sudan and Chad and then spread westwards to Niger, Mali, Burkina Faso and Senegal. Seasonality of the disease can be seen in many countries but is especially prominent in the meningitis belt. In the belt the highest disease incidence is found when temperatures are high and humidity low. The epidemic starts in the dry season and end abruptly as soon as the rain season begins. Ecological modelling on climate factors has shown that the two most important factors for meningococcal epidemics are low absolute humidity and land cover type. Absolute humidity also has a bearing on other factors that influence the epidemics like rainfall and dust profiles. The climate factors also influence carriage rates and social behaviour that might facilitate the spread. Early warning systems based on meteorological measures should be a very useful tool in enabling preventive measures like vaccination, in good time before epidemics start. Global warming may be behind some alarming trends in the meningitis belt and in the whole of Africa during the last 30 years, with both geographical extension of the belt and with new countries suffering from the disease. Climate factors also influence the incidence of meningococcal disease outside the meningitis belt. Large epidemics often occur in geographical areas with climate and land cover profiles similar to those in sub-Saharan Africa like epidemics that have occurred in India and China. But the impact of climate factors in geographical distribution of meningococcal disease is somewhat confused by the fact that the countries in Europe with the highest incidence are UK and Ireland, countries known for high rainfall and high humidity. The seasonality is seen in many countries, also in Sweden, with higher disease incidence in the seasons with low humidity. But in the northern hemisphere the lowest absolute humidity is found during winter, the cold season. It is not well understood how meningococcal disease is affected by climate factors. One plausible explanation is that low absolute humidity is affecting the mucosal membranes in the upper respiratory tract so that Neisseria meningitides more easily can both bind to and penetrate the membranes. This has not been very well investigated, mainly due to a lack of animal models for the disease. This is a research field that needs to
be further analysed. It is also important with an extended collaboration between medical and meteorological experts to be able to create better early warning systems in the meningitis belt and to predict the future effect of global warming on meningococcal meningitis epidemiology.
“The effects of climate and environment on tularaemia” Anders Sjöstedt, Professor of Clinical Bacteriology, Clinical Microbiology, Umeå University
Endemic foci of tularemia in Sweden are Ljusdal and Örebro. A very small organism, Francisella tularensis, is the causative agent of tularaemia. It is one of the most infectious agents known, with infectious dose of 10 bacteria. It is spread by rodents, hares, mosquitoes, ticks etc. Humans get infected through the skin, mucous membranes, GI tract and lungs. There are two clinically important subspecies of the bacteria; tularensis (Type A) only found in North America and holarctica (Type B) the only subspecies in Europe. By far the most predominant form in most countries is ulceroglandular tularemia – a result of a vector-bite, in Sweden by mosquitoes. In Sweden the respiratory form has often been connected with farming activities. Hunting and export of animals for hunting purposes is a risk factor for spreading the disease. In many countries the true prevalence is hard to state. Physicians are not used to see the disease and might mis-diagnose it as something else. War is known to increase risk factors for disease like exposure to live rodents, contaminated grain supplies, degrade of garbage disposal and water-borne epidemics. There are several examples of tularemia as a war-related disease; Soviet Union, Karelia, Bosnia, Kosovo etc. The disease was very common historically in several countries and was originally identified almost hundred years ago. Since then the trend has decreased in many countries. In the US there were more than 14000 cases in 1915 – 1945 with a peak in 1939 with 2291 cases. Since 1995 there has been some 100 cases/year. In the 1940’s there was a 100000 cases/year in Soviet Union. Since the mid-1950s there have been a few hundred annual cases. Some factors that can have changed the historical trend are urbanization, “survival” hunting and ecological changes. There have been outbreaks of tularemia in Sweden, once in 1967 with more than 2000 human cases caused by mass exodus of lemmings and exposure trough traditional farming activities. The lemming population has decreased in the last twenty years and some Norwegian researchers have recently claimed that subtle climate change could be the reason for their decrease. Since the 1980’s, there have been much fluctuations in the annual number of tularemia cases in Sweden, with an annual incidence ranging from 10 – 800. In almost all cases tularaemia infected individuals in Sweden report that they have been by a lake or been fishing. Water seems to be a key factor involved in transmission and a hypothesis is that water itself can be the long term reservoir, and not animals. One hypothesis is that vectors like mosquitoes get in contact with the bacteria through water. In an ongoing research project samples are collected to be linked to observations of factors like the role of temperatures, nutrients, precipitation and wetland restoration to be able to model and predict outbreaks. Within the EU there are projects with restorations of wetlands – a factor that might contribute to the spread of arthropod-borne diseases. From the geographical variations seen it is probably so that the disease will be an emerging disease in parts of the
world where it has not been seen so far. Some examples of research questions are: Is the normal habitat of F. tularensis (type B) water, why does it spread from its main habitat irregularly, how do vectors become infected and why are there such dramatic geographical variations? It’s important to follow and understand the quite quickly changing epidemiology of the disease in connection to climate factors.
“Climate Variability, Migration and Rural Livelihoods: some theoretical and methodological considerations” Mark Collinson, University of Witwatersrand, Johannesburg, South Africa (Project team: Mark Collinson, Lori Hunter, Wayne Twine and Stephen Tollman)
Studying health patterns and how populations and households are changing particularly in their livelihood strategies can give an understanding of how the disease patterns are related to the social organization. We have a theme of work which is called how households respond to shocks and stresses. It relates to environment changes which has climate related drivers within it. We are looking at outcomes as health and other demographic issues, climate variability and migration. The environmental change has a direct and immediate impact on the health and well-being of rural households. Natural resources are often essential meeting basic livelihood requirements. The environment shapes the capabilities that poor rural households have. Arable farming, live stock husbandry, consumption and trade are all depending on the natural resources. They also act as a “buffer” against household shocks such as job loss and/or mortality. Changes in local vegetation cover hold a tremendous potential to impact livelihoods. The poorest and the most vulnerable will disproportionately experience the negative impacts of some of these environmental changes. As related to out migration, a decline in livelihood options can act as a “push” factor shaping migration trends. A hypothesis is that out migration is negatively associated with natural resource availability and positively associated with natural resource variability. Other factors has to be taken into account as relevant factors that drives migration and changes, influences at the individual level (e.g. age and sex), the household-level (e.g. socio-economic status) and village level (e.g. market proximity). The study setting Agincourt is situated in the north east of South Africa close to the border to Mozambique. South Africa is a middle income country but a highly unequal society. The rural area of which Agincourt is a typical one is where the poorest are located having a lot of historical drive behind it; the homeland history with poverty and segregation. ‘Rural’ in this setting means that people live far from resources. Only a minority can afford electricity. It is a savannah ecosystem and there are many factors interrelating between climate, weather and vegetation determined by geology, fire, herbivore and people affecting
how the system works. The condition of the vegetation in the setting is a proxy for these other complex systems that will not be modeled. Remote sensed data from the MODIS satellites are used. An index NDVI (“normalized difference vegetation index” – how green is the environment – its variability) can be derived from the data. What is there for the cattle to eat, for the subsistence farmers to grow from etc? This will be the proxy measure of the health of the environment with links to migration. The migration context is studied using HDSS. What drives the migration has to be understood using multilevel longitudinal models looking at individual, household and village factors linked to natural resource availability and variability. The seminar was closed by Professor Lars Weinehall, Epidemiology and Public Health Sciences, Umeå University, summing up the potential for climate change and health as a theme within the Umeå Centre for Global Health Research. Many researchers at Umeå University are dealing with aspects on climate and health and many resources can be identified with a potential for bringing both an arctic and tropical perspective. The centre has links to the INDEPTH network and contacts to researchers around the world. By organizing an Umeå body / a network for collaboration, important steps can be taken to broaden the research base and scientific expertise, important for coming forward in this momentous task.
Concluding remarks
The seminar showed many examples of the complexity of the issue and the need for research and prioritization of key research questions and actions. The seminar unfolded an agenda shoving a fertile ground to think further from, with a mix of very long perspectives and shorter locally anchored ones. There was the northern arctic perspective and the southern tropical perspective as well as examples from low income- and high income countries. Links and similarities between the northern and southern hemisphere appeared in connection to climate change and health; the research agenda is not as different as one could expect. Some diseases, for example Meningococcal disease, can be studied from both perspectives. Most of the similarities appeared in the field of infectious diseases in modelling, in linking disease parameters to social and environment parameters, in looking at early warning systems etc. The longitudinal effects must be followed in climate research regardless of setting. Climate change impact on human health is a very under researched area; in the north but even more in the south. Researchers from different disciplines will have to join in the development of methods, tools, teaching and training and other coordinated efforts but also into the mitigation debate on how health can be brought into the mitigation as well as how to influence policy. Public health researchers must actively reach out to disease specialists and climate/meteorologist professionals, to mention some. A theoretical base and new skills/ techniques are needed, as well as professional leadership and broad teams working over sectors and cross borders, connecting to existing monitoring systems and creating new ones. With stronger evidence base health arguments can be developed and communicated to citizens and policymakers at all levels.
Summary, December 2008, Maria Nilsson, Epidemiology and Public Health Sciences, Umeå University