Why is Malaria Declining While Temperature (Might Be) Going Up in the East Africa Highlands?

Stern et al. (2011) examined trends in temperature and malaria for the Highlands of East Africa which span Rwanda, Burundi, and parts of Kenya, Tanzania and Uganda to resolve controversies over whether (a) the area has warmed and (b) malaria has become more prevalent. For temperature, the authors used three time series obtained from the Climate Research Unit (CRU) for four locations. Temperature data for Kericho extend through 2009, but only through 2006 for the Highlands. For malaria, the researchers used (a) a data set on malaria cases through May 2010 at a hospital in Kericho, Kenya, and (b) data from over 5,200 surveys on the prevalence of malaria though 2009 in the Highlands.

Results indicate that temperature has increased significantly in the region, but with respect to malaria, "malaria in Kericho and many other areas of East Africa has decreased during periods of unambiguous warming." The paper does not attempt an explanation for the divergence in these trends.

There are two major flaws in this paper. First, it does not discuss whether the temperature data from the CRU for the specific locations are uncontaminated by microclimatic influences and, therefore, representative of the broader region. In this methodological oversight, it is far from unique. This shortcoming is endemic to many, if not most, studies of the impact of climate change. That is, they accept the underlying data without questioning its quality or whether they are stationary before undertaking statistical analysis or using them to model other processes. Thus, one should take this paper's conclusion that temperature is trending upward for the region with a pinch of salt. Hence the "might be" in the title of this review. Second, it does not discuss the differences between the three CRU versions.