Environmental, entomological, socioeconomic and behavioural risk factors for malaria attacks in Amerindian children of Camopi, French Guiana

Abstract

Background: Malaria is a major health issue in French Guiana. Amerindian communities remain the most affected. A previous study in Camopi highlighted the predominant role of environmental factors in the occurrence of malaria. However, all parameters involved in the transmission were not clearly identified. A new survey was conducted in order to clarify the risk factors for the presence of malaria cases in Camopi.

Methods: An open cohort of children under seven years of age was set up on the basis of biologically confirmed malaria cases for the period 2001-2009. Epidemiological and observational environmental data were collected using two structured questionnaires. Data were analysed with a multiple failures multivariate Cox model. The influence of climate and the river level on malaria incidence was evaluated by time-series analysis. Relationships between Anopheles darlingi human biting rates and malaria incidence rates were estimated using Spearman's rank correlation.

Results: The global annual incidence over the nine-year period was 238 per 1,000 for Plasmodium falciparum, 514 per 1,000 for Plasmodium visa and 21 per 1,000 for mixed infections. The multivariate survival analysis associated higher malaria incidence with living on the Camopi riverside vs. the Oyapock riverside, far from the centre of the Camopi hamlet, in a home with numerous occupants and going to sleep late. On the contrary, living in a house cleared of all vegetation within 50 m and at high distance of the forest were associated with a lower risk. Meteorological and hydrological characteristics appeared to be correlated with malaria incidence with different lags. Anopheles darlingi human biting rate was also positively correlated to incident malaria in children one month later.

Conclusions: Malaria incidence in children remains high in young children despite the appearance of immunity in children around three years of age. The closeness environment but also the meteorological parameters play an important role in malaria transmission among children under seven years of age in Camopi.

Figure 1

Groups of mosquito collection sites. (a) Ilet Moulat (on the upstream Oyapock riverside), (b) Saint Soit (on the downstream Oyapock riverside), (c) Camopi main hamlet (at the confluence of the Oyapock and Camopi Rivers) and (d) hamlets on the Camopi riverside.

Figure 2

Inter-annual variations of malaria incidence rates per 1000 person-years attributed to P. falciparum and P. vivax (including and excluding relapses). Incidence rates were calculated for the population 0 - 7 years old.

Figure 3

Intra-annual variations of incidence rates for P. falciparum and P. vivax (including and excluding relapses) during the period study (2001-2009) in children under 7 years of age.

Figure 4

Incidence rates for P. falciparum, P. vivax (including and excluding relapses) and mixed infections by age on the study period (2001-2009).

Figure 5

Intra-annual malaria incidence rates in children and (a) monthly cumulated rainfall and river level, and (b) minimum and maximum temperature in Camopi on the study period (2001-2009). Incidence rates have been calculated for the two species combined and P. vivax relapses were excluded.

Figure 6

Spearman's rank order correlation coefficients between residual series of malaria incidence and meteorological factors. Correlations for which associated P-value < 0.05 are indicated by black markers.

Figure 7

Human biting rates of anopheline species collected between 2003 and 2006 in different groups of hamlets of the study site. (a) Ilet Moulat (on the upstream Oyapock riverside), (b) Saint Soit (on the downstream Oyapock riververside), (c) Camopi main hamlet (at the confluence of the Oyapock and Camopi Rivers) and (d) hamlets on the Camopi riverside.