community based or high‐risk sample would be one approach to test whether the deviant sex ratio for a particular disorder is a result of sampling or diagnosis. For example, depressive symptoms present equally in boys and girls in childhood, and the nearly 2‐fold increase among depressive symptoms among women only appears postpubertally. 26 Therefore, sex differences in the course or trajectory of depression that have been shown in longitudinal cohort studies 27 , 28 may not be evident in a cross‐sectional examination.
