Fig. 1

Hazard ratios and the 95% confidence interval of clinical malaria for IgG1 and IgG3 levels for children at 15 months. A Cox proportional hazards regression model was used to test the protective effect of IgG responses previously associated with the phenotype of asymptomatic infection clearance at 15 months of age in the whole sample (CAIG + NCIG). Hazards ratios obtained from Cox-regression model adjusted by environmental risk of exposure to malaria, considering exposed children with no asymptomatic infection at the time of Ab measurement. The number of infants included in Cox-regression model was 207 (13 infants had no Ab measurement at 15 months of age and 19 infants were excluded due to presence of asymptomatic infection at the time of Ab measurement). 66, 81, 50, 9 and 1 infants had, respectively, 0, 1, 2, 3 and 4 malaria attacks between 15 and 18 months of follow-up