Seroepidemiological and parasitological evaluation of the heterogeneity of malaria infection in the Gambia

Abstract

Background: As countries make progress in malaria control, transmission may be reduced to such an extent that few cases occur, and identification of the remaining foci of transmission may require a combination of surveillance tools. The study explored the usefulness of parasite prevalence, seroprevalence and model-estimated seroconversion rates for detecting local differences in malaria transmission in a West African country.

Methods: Age-stratified cross-sectional surveys were conducted during the wet season in 2008 and the following dry season in 2009 in The Gambia. In each season, 20 village communities were sampled from six diverse areas throughout the country. A total of 7,586 participants were surveyed, 51% (3,870) during the wet season. Parasites were detected by thick film slide microscopy, and anti-MSP1-19 antibodies were detected by ELISA using eluted dried blood from filter papers.

Results: Overall parasite prevalence was 12.4% in the wet season and 2.2% in the dry season, with village-specific parasite prevalence ranging from 1.4 to 45.9% in the wet season and from 0.0 to 13.2% in the dry season. Prevalence was highest in the eastern part of the country. Serological indices also varied between villages, indicating local heterogeneity in transmission, and there was a high correlation between wet and dry season estimates across the villages. The overall prevalence of anti-MSP119 antibodies was similar in the wet (19.5%) and in the dry (19.6%) seasons.

Conclusion: The study illustrates the utility of measuring both parasite prevalence and serological indices for monitoring local variation in malaria transmission, which are more informative than single measures as control intensifies and malaria declines. Measurements of seropositivity have the logistical advantage of being relative stable seasonally so that sampling at any time of year may be conducted.

Figure 1

Map of The Gambia showing the six areas and 20 study villages samples in the wet season (September to November) of 2008 and dry season (March to May) of 2009. The villages which have health centres are shown in italics.

Figure 2

Parasite prevalence (%, 95% CI) and anti-MSP1₁₉ seroprevalence (%, 95% CI) in each of the 20 study villages from six different areas of The Gambia, from the wet season sampling in 2008. Sample sizes and numbers, as well as tests for inter-village heterogeneity within areas, are given in Additional files 1 and 2.

Figure 3

Correlations among serological measures across the 20 different villages sampled throughout The Gambia. (A) Seroprevalence measurements from the wet and dry seasons are similar (dotted line shows x = y) and very highly correlated. (B) Derived lambda parameters from the age-seroprevalence distributions tend to be slightly higher for the wet season estimate, with outlying data for two labelled villages. (C) Wet season correlation between seroprevalence and derived lambda parameter, showing two labelled villages with an elevated lambda estimate. (D) Dry season correlation between seroprevalence and derived lambda parameter. Estimates of the lambda parameter with 95% confidence intervals for each season are given in full in Additional file 3.