As an addition to see if common propensity would indeed show a stronger association with CADM2, we also meta‐analyzed factors that were made up of similar indicators in different cohorts. We used similar procedures for these analyses as for the separate phenotypes in MAGMA.