15 November 2002

by Diarmid Campbell-Lendrum diarmid.campbell-lendrum@lshtm.ac.uk
and Richard Reithinger rreithinger@yahoo.co.uk
 

Several studies have attempted to predict the likely effects of future climate change on the distribution of vector-borne diseases. Rogers and Randolph suggest that such predictions will only be correct if the models on which they are based can give a reasonable explanation of the current distribution of disease. For example, their own analyses show that a statistical relationship with several climate variables predicts a malaria range that agrees with the current distribution in 78% of 0.5 grid cells. Applying predictions of future climate to their model, however, leads to insignificant changes in the distribution: areas that are likely to meet the multiple climate requirements for malaria transmission are balanced by those that become unsuitable for at least one parameter.

A study by Hales et al. applies similar analytical techniques to dengue, the most important viral vector-borne disease in the world, with an estimated 0.433 daily-adjusted life years. As for malaria, the current distribution of the disease is well explained by a statistical regression model using only climate variables. Vapour pressure, a measure of both temperature and precipitation, correctly predicts the reported presence or absence of dengue in 89% of 0.5 grid cells. In contrast to malaria, however, applying climate and human population projections for the 2080s to this simple model generates predictions of huge expansion in the global population at risk: an extra 1.5-2.5 billion people, or 15-25% of the world’s population, compared with the situation if climate change were not to occur. The reason why projections for dengue might differ from the ones for malaria is because dengue distribution appears to be less limited by non-climatic factors, for example living standards and control programmes. Also, most dengue transmission is due to a single, highly transportable vector, Aedes aegypti, which exploits a wide range of urban breeding sites and is notoriously difficult to control.

As Hales et al. point out, climate and climatic change are not the only factors that will determine the future distribution of dengue. Although strengthened control programmes and new technologies (e.g. mass-release of sterile Aedes mosquitoes) could reduce the dengue threat, it is likely that increasing trends in other determinants of dengue (e.g. urbanization, the production of household waste providing breeding sites for A. aegypti and the global interchange of dengue serotypes), will contribute to the expansion of dengue-endemic areas.

Given this complexity, projections of the impact of future climate change should be viewed as ‘best guesses’ of what is likely to happen if other factors remain constant, or change along predictable lines. This work should emphasise the need to (i) develop new interventions and make the best use for existing tools to control dengue now and in the future, and (ii) minimize risk factors for this disease wherever possible. The study suggests that anthropogenic climate change should be included in this list of risk factors.