All analyses were stratified by residency
(permanent and migrant) using
the Mantel-Haentzel method. To adjust
the difference in exposure duration for
case-children (1 week) and controlchildren
(8 weeks) for the 1-time or
rare exposures, we divided the frequency
reported by control respondents
by 8 and then used logistic regression
to calculate odds ratios (ORs)
on the basis of a 1-week period.