Does Low Participation in Cohort Studies Induce Bias?

For the 3 chosen associations, the ORs were not biased by nonparticipation. The results are reassuring for studies based on the Danish cohort and similar cohorts of pregnant women. The methodology used to compute confidence intervals for the relative odds ratios performed well in the scenarios considered.