Within MVP there were three depression phenotypes investigated across five different analyses. We used 1) an ICD code-based algorithm to determine depression case status based upon investigation of the electronic health records (MDD, primary analysis), 2) self-reported physician diagnosis of depression as reported in the MVP baseline survey (SR-Depression), and 3) the 2-item PHQ scale of depressive symptoms in the past 2 weeks, included in the MVP baseline survey (depressive symptoms). Phenotypes in outside cohorts for UKB-PGC and 23andMe have been previously described.7–10 See Table 1 and Figure 1 for summary. For the ICD code-based algorithm in MVP, codes used to assess case status are presented in Table S3. Cases included people with at least one inpatient diagnosis code or two outpatient diagnosis codes for Major Depressive Disorder (MDD). Controls include only those without any inpatient or outpatient depression diagnosis codes for depression.
