The percentage of underreporters was already
described in depth in a previous paper and was shown
to be low (< 2% of the children when using Goldberg
cut-offs adapted for children) [3]. Underreporters have
not been excluded from the study sample that was used
for the present analyses.