Is prevalence of diarrhea a better predictor of subsequent mortality and weight gain than diarrhea incidence? uri icon