Episodes per child/year showing continuity of being ‘more susceptible’ over time.

Note:*Age in years at the start of the studyYear 1 = September 1998 to September 1999Year 2 = October 1999 to October 2000Year 3 = October 2000 to November 2001Year 4 = November 2001 to November 2002Year 5 = December 2002 to September 2003.M. susceptible = ‘more susceptible’. These are the children who experienced >2 episodes of clinical malaria above the total predicted from the Poisson regression model.