Fig. 3
Time-to-first Plasmodium falciparum infection in Nandi vs Kisumu based on the observed levels of maternal antibody. Infants were divided into four groups. First, the two extremes of antibody levels were compared (a); Infants from a region of high malaria transmission with high levels of MSP1-3D7 antibody and infants from a low transmission region with low levels of MSP1-3D7 antibody. Modelling was used to estimate the duration of protective effect, and subsequent frequency of infection. There was evidence for a longer period of inherited immunity in the high group (presumably due to maternal antibody), but also a higher infection rate once immunity has decayed. The remaining children with more homogenous antibody levels, but from regions of different malaria exposure were also compared using a parametric model of duration of protective effect and subsequent infection rate in these two groups (b). There was no significant different in the predicted duration of protective effect between the two groups. Using a parametric survival analysis, assuming a constant rate of infection, the infection rate is higher in children from a region of high transmission