Review of Trape et al., (2011), Malaria morbidity and pyrethroid resistance after the introduction of insecticide-treated bednets and artemisinin-based combination therapies: a longitudinal study, The Lancet Infectious Diseases, published online August 18th.
Jean-François Trape and colleagues have been causing quite a media stir with the article they recently published in The Lancet Infectious Diseases, an offshoot of the eminent medical journal. The research consisted of a longitudinal study, following the inhabitants of a village in Senegal for almost four years, recording all instances of malaria and treating cases with artemisinin-based combination therapies (ACTs), the WHO-recommended front-line medication for uncomplicated malaria cases. In addition, 18 months into the study, the villagers were provided with long-lasting insecticide treated bednets (LLINs); concurrently, throughout the study period Anopheles mosquitoes, the vectors of malaria, were collected from the village area and tested for resistance to the insecticides used in these LLINs.
The scientists noted an immediate decrease in the monthly incidence of malaria in the initial period following the introduction of LLINs, but then recorded a substantial, and statistically significant, increase in re-infections in the final three months of the study. Moreover, when broken down into age groups, this rebound of malaria cases was most notable in children aged 10-14 years, which is unusual given that normally younger children are considered most susceptible to malaria infection. Perhaps most concerning of all, more than a third of mosquitoes tested at the end of the study were resistant to the insecticide in the LLINs, and the prevalence of a gene known to correlate with resistance had increased from 8% to 48% over the four-year period.
This research is clearly highly important, timely, and worthy of deeper consideration. While LLINs have been shown to be highly effective in earlier studies to prevent malaria infection, fewer studies have focused on the long-term consequences of these interventions, which are being rolled out at great speed throughout malaria endemic areas. Clearly, insight into the effects on mosquitoes, as well as re-infection rates, is warranted. However, a few gentle caveats should also be made.
For one, although the observation of reduced immunity being responsible for higher rebound rates is of great interest, and a sound conclusion from the evidence, it requires much further testing before a causal relationship can be proven. Secondly, the authors compare a rebound period of three months, at the end of the study, with the preceding two years of LLIN use as well as the initial LLIN-free 18 months, in order to draw their conclusions about re-infection rates.
Malaria is highly seasonal; while this is acknowledged by the authors (one of the figures includes rainfall in the graph of Anopheles biting rates), it is unclear whether this was included as a covariate in the statistical analysis, and therefore controlled for as a variable. As it happens, that final three month period occurred during a season of high rainfall, usually associated with high numbers of mosquitoes and subsequently high biting rates. Similarly, the authors do not account for the possibility of other environmental factors influencing the rate of malaria during this final, and very short, time period. Had they compared time periods of equal length, and incorporating all seasons, it might be easier to ascertain whether the increase they observe is a true trend or a statistical artifact of sampling bias.
However, I am getting slightly off the point with technical grievances. The methods the authors use are sound; their conclusions valid, if perhaps in need of further justification. My main concern with the impact of this paper lies in its misinterpretation by the media. Already, I have seen one newspaper report (The Independent on Sunday August 21st, 2011 “Twenty Holiday Myths Exposed”) which erroneously stated “Research published last week suggested mosquito nets were not as effective as previously believed.”
I can only assume they are referring to the Trape paper; this is patently not the conclusion the scientists come to. Rather, they state that in one area of high malaria endemicity, bednets may reduce immunity, leaving people more vulnerable later, and secondly, that LLINs may result in increased resistance to insecticides in mosquitoes.
The authors make no statement about the efficacy of bednets in preventing infection. Moreover, the use of bednets in a highly endemic area, where people have no other option for malaria prevention, is a very different situation than what travelers face, visiting a malarial area for a matter of days or weeks, armed in conjunction with prophylactic medication and insect repellant sprays. The Independent‘s ignorant throw-away statement may encourage travelers to forsake bednets, which still are a key way of preventing malaria infection, thus leaving themselves at greater risk. I am sure Trape and his colleagues did not intend for this to be a consequence of their research.