Figure 1.

Sensitivity analysis of adjustment for an unmeasured dichotomous confounder on the prevalence odds ratio for migraine*. * Assumes a confounder prevalence of 20% in the non-IBS cohort.