A recent press release from the UK Department for International Development (DFID) suggested that millions more people in Kenya are susceptible to malaria as a result of mosquitoes colonising higher ground as global temperatures rise. ('New evidence of a link between climate change and malaria', 30.12.09 - see below). The press release was extensively covered in UK newspapers and elsewhere. Simple analysis shows that the claims of the press release are almost entirely without foundation. The battle against the severe threat from climate change is impeded, not helped, by government departments issuing alarmist and exaggerated alerts based on poor science.
All other things being equal, the spread of malaria is probably encouraged by higher global temperatures. The malaria parasite in the insect's body grows fastest at average daily temperatures of about 25 degrees and most parts of the world are well below this level. But temperature is only one factor in the spread of this terrible disease, possibly a small one. The presence of stagnant water in open sunlit fields after deforestation, increased population pressure, lack of availability of mosquito nets, or a reluctance to use them, may all contribute as much to the spread of malaria as increasing temperatures. Nevertheless, despite the plethora of more convincing explanations for varying levels of malarial illness, policy-makers and government departments continue to state, without any qualification, that malaria will become very much more prevalent in a warmer world.
The 2001 IPCC report also overstated the connection between climate change and malarial infections. Understandably, the top-selling books by climate sceptics published in the last few years all feast on the weak scientific evidence for this assertion. These books usually quote the specialist in insect-borne diseases, Professor Paul Reiter of the Pasteur Institute in Paris, who has strenuously and effectively attacked the idea that increasing temperatures will necessarily produce a rapid rise in the incidence of insect-borne diseases. Professor Reiter points out that malaria transmission is a complex matter and that rising temperatures are only weakly linked to an increasing incidence of malaria. (The illustration at the head of this article provides us with some sense of just how complex malaria is.) Why, he and others have asked, if temperature is so important, did the disease disappear from countries like Britain just as the climate was warming at the end of the 'Little Ice Age' during the eighteenth and nineteenth centuries?
However, the story that substantial increases in malaria will inevitably follow rising temperatures will not go away and DFID's recent press release is another example. The document claims that 'new research' in the highland areas around Mount Kenya has shown strong links between increasing temperatures and malaria incidence. When I asked for a copy of the scientific papers to back up this assertion, DFID said it was unable to provide this 'new research', stating instead that the press release was based on 'a cumulation of several studies over the past few years'.
Three papers were attached to this surprising response from DFID. The press release asserted that temperatures on the western side of Mount Kenya had risen two degrees Celsius in twenty years, prompting epidemics of malaria, but these research papers actually showed a much smaller increase. The specific claim that the Mount Kenya area has recently become vulnerable to malaria was backed up by interview data of a few years ago from a small number of families who declared a total of eight cases of malaria in the past five years compared to only three in the period of five to ten years ago. No medical analysis appears to have been carried out to determine whether the disease recorded was or was not malaria. Neither was any attempt apparently made to adjust for deficiencies in memory of events ten years ago.
What about the physical evidence of mosquitoes? The scientific papers sent to me by DFID write of finding a total of two mosquito larvae in pools on high ground near Mount Kenya. These larvae produced a total of 23 insects when incubated in a laboratory. The quality of this finding is never questioned. Most importantly, the background research papers do fully acknowledge that other events in the geographic areas under study, such as deforestation or increased pooling of stagnant water as a result of land-use changes, could well have been the primary cause of any growth in the number of mosquitoes and of malaria, not climate change. But these possibilities are unmentioned in the DFID press release.
A line-by-line analysis of the press release is provided below. It suggests that almost all the assertions in the DFID document are unsupported by any evidence or are downright wrong. Malaria may indeed be increasing in highland areas of Kenya but the evidence that the disease is driven by rapid climate change is completely unsupported. The research used by DFID and sent to me actually states that changes in land-use may be the key factors in encouraging the growth of this extremely unpleasant disease.
The climate sceptics often claim that the political establishment routinely exaggerates the effects of increasing temperatures. These sceptics say that very poor scientific evidence is used to create alarmist headlines about the impact of global warming. The DFID press release is a clear candidate for criticism of this type. It has used inadequate scientific research as the basis for an unqualified conclusion about the impact of climate change on insect-borne diseases. In this case, the press release goes even further by mis-stating many of the elements of the research on which it is based. Such abuse of science does substantial disservice to the attempt to energize the global community to take action to mitigate climate change.
(With many thanks to Professor Paul Reiter for his comments during the preparation of this article. Errors are my responsibility. If you are interested in reading more about the causes of malarial infections, please consult one of Professor Reiter's papers.)
[Press release sent out by DFID with my notes in bold:]
New evidence of link between climate change and malaria
New research released today shows a direct link between rising temperatures and the spread of malaria.
a) The research is not new. The key scientific papers to which the press release refers were published in 2001 and 2006. The research was carried out by Dr Githeko of the Kenya Medical Research Institute and his colleagues. Dr Githeko contributed a summary of this earlier research and of other recent findings from around the world for a Commonwealth health ministers meeting in 2009.
b) DFID now acknowledges the material is not 'new research'. When questioned the Department said: 'It's more a cumulation of several studies over the past few years'.
c) The research does not show a 'direct link' between malaria and average temperatures. The 2001 paper actually says: 'Land use pattern and land cover might be the key factors affecting [...] malaria transmission in the region'.
Four million people living on the slopes of Mount Kenya in the Kenyan Central Highlands are now at risk of malaria, after warmer temperatures pushed the disease into high altitude areas where the population has little or no immunity.
a) The 2001 paper says: 'With the increased population on the highlands, enhanced human activities including deforestation, farming and livestock rearing could create more vector [i.e. malaria carrying mosquitoes] habitats'. It does not say climate change is causing malaria.
b) The 2006 paper says in reference to Western Kenya, over 200 km from Mount Kenya, that the researchers had found 'an association between rainfall and unusually high maximum temperatures and the number of inpatient malaria cases 3-4 months later'. It makes no reference to the slopes of Mount Kenya.
The findings, announced today by a research team from the Kenyan Medical Research Institute (KEMRI) partly funded by the UK Department for International Development (DFID), also show the number of cases recorded during malaria epidemics has increased by up to 600% over the past decade.
a) The findings, at least as provided in the back-up reports provided by DFID, do not show this. The only mention of malaria cases in this research is as follows: 'In Karatina and Naro Moru areas where An. Arabiensis [one of the malaria mosquitoes] was detected, the numbers of malaria cases recorded were eight in the past 5 years, three in the past 5-10 years and none in the past 10-20 years, indicating a trend of increasing malaria incidence over the past 10 years.'
b) Malaria may well be becoming an even more serious problem in Kenya, and rising temperatures may indeed contribute to the incidence of this disease. But DFID provides no evidence to support its assertion about the Mount Kenya region. Dr Githeko's 2009 summary paper does mention third-party research showing an 'eight-fold' increase in malaria in Western Kenya (not the Mount Kenya area) and this may be the basis for the DFID assertion.
While similar outbreaks elsewhere have been attributed to multiple factors such as drug resistance and land use change, the research team claim the only change that has occurred recently in the Kenyan Central Highlands that might lead to malaria is an increase in the mean annual temperature.
a) Dr Githeko's 2009 summary paper says that 'In the central highlands the only change that occurred and that lead [sic] to malaria was the mean annual temperature'. But in this and the other papers, he also pointed to variations in rainfall, increases in maximum daily temperatures, changes in land use, greater population density, deforestation, irrigation, and stagnant water as having an effect on malaria rates. Dr Githeko's comments are inconsistent, even within single papers.
The spread of malaria into the Central Highlands district followed a rise in average yearly temperatures from 17 degrees in 1989 to nearly 19 degrees today. The malaria parasite can only mature in temperatures above 18 degrees.
a) Prior to 1994, the mean annual temperature in the Central Highlands region is said by Dr Githeko to be 'between 17.4 degrees C and 18.2 degrees C'. Dr Githeko does not provide data for recent years. His record stops at around 2002.
b) The average annual temperature does not have to be above 18 degrees for the malaria parasite to mature. The parasite requires periods of higher temperatures to grow but it can mature in a few days of warm weather. The annual mean is not a particularly relevant measure. (Some northern European countries were infested by malaria until the twentieth century and their annual average temperatures were well below 18 degrees C.)
Before the 1990s malaria was completely absent in the Central Highlands district. However, as climate change pushed average temperatures over the 18 degree tipping point in the 1990s, malaria epidemics began to break out among the population.
a) As said before, mean annual temperatures rising above 18 degrees do not create a 'tipping point' for malaria. Recent years may have seen malaria increasing in the Mount Kenya region, partly perhaps as a result of global warming but, second, the DFID papers provide no evidence whatsoever of 'epidemics'.
Malaria has continued to spread in populations with little immunity to the disease. In 2005 malaria-carrying anopheles mosquitoes were discovered in Naru Moro, over 1,900 metres above sea level.
a) Dr Githeko found two (2) mosquito larvae in 2005 in Naru Moro, not mosquitoes. Larvae do not carry malaria. The mosquito carries malaria after it has ingested the blood of a person with the disease.
b) DFID is wrong to state that Dr Githeko's findings show that 'malaria has continued to spread in populations with little immunity' and to associate this conclusion with his finding that the two Naru Moro larvae showed this continuing spread.
The KEMRI researchers were funded by the Department for International Development (DFID) and the Canadian International Development Research Centre (IDRC).
[Press release then continues with quotations for journalists.]