Effects of long-term use of insecticide treated bednets on species composition and knockdown resistance in Malaria vectors in Asembo location of Western Kenya

Abstract

Insecticide-treated bed nets (ITNs) for personal protection against malaria vectors have been in use for the control of malaria in most tropical areas where malaria transmission is intense. However, the emergence of insecticide resistance in Anopheles vectors of malaria may impair the effectiveness of this control measure. Insecticide resistance has been reported in several countries, Kenya included. Effects of long-term use of ITNs on species composition and knockdown resistance (kdr) were investigated in Asembo Location of Western Kenya where a large scale ITN program has been implemented since 1997. Mosquitoes were sampled from a 12km transect in February and May 2004 from Asembo ITN intervention area and the adjacent Seme non-ITN intervention area. Indoor collection of adult mosquitoes from human dwellings was done using mouth aspirators and larval collection from habitats around homesteads using dippers. The mosquitoes were identified morphologically as Anopheles gambiae s.l. and as An. arabiensis or An. gambiae s.s. by polymerase chain reaction (PCR). Analysis of the presence of knockdown resistance (kdr) allele was done using kdr diagnostic PCR. Results showed that there was significant reduction in the proportions of An. gambaie s.s. in Asembo with relatively higher proportion of An. arabiensis compared to Seme (Binomial regression GenMod procedure, P<0.0001). The kdr allele was present in An. gambiae s.s. only but absent in An. arabiensis. There was no statistical difference in kdr allele frequency trend along the transect (logistic regression, df=l, X2=2.7664, P=0.0963). Overall, the frequency of kdr gene was 25% in Asembo and 16.7% in Seme despite lack of wide spread use of ITNs in Seme. Nevertheless, these kdr allele frequencies were not statistically different from each other (Fisher Exact test, F=17.9614, df=2, P=0.05274). The genetic differentiation based on the kdr allele was estimated but was also not significant with a low genetic variability (Fst value) of 0.0098, which corresponded to migration index (Nm value) of 25.23. The difference in kdr phenotypes between Asembo and Seme mosquitoes was not significantly different (ANOVA, F=8.71E-16, df=1.4, P=1.000). These results suggest free interaction between the Asembo and Seme mosquito populations. The observations of this study suggest that widespread use of ITNs in Western Kenya has led to an increase in kdr frequency spreading to adjacent Seme non-ITN intervention area. Another impact of long-term use of ITNs was seen in altered species proportions (less An. gambiae proportion and more An. arabiensis in Asembo than Seme). Though the kdr allele responsible for knockdown resistance in mosquitoes was absent in An. arabiensis, this species is known to be exophilic hence may not be effectively controlled by use of ITNs.