Another obstacle to progress in identifying susceptibility loci is the fact that depression is a heterogeneous phenotype. Indeed, it is possible to meet DSM-IV or DSM-5 diagnostic criteria for a major depressive episode through at least 227 different symptom combinations.144 As currently described by DSM-5, MDD can manifest with or without: (1) anxious distress; (2) mixed features; (3) melancholic features; (4) atypical features; (5) mood-congruent psychotic features; (6) mood-incongruent psychotic features; (7) catatonia; (8) peripartum onset; and (9) a seasonal pattern.145 These subtypes of major depressive disorder could reflect different genetic contributions. Consistent with such a hypothesis, studies suggest that depression with a history of child maltreatment has a different onset, course, and response to treatment when compared a depression that arises among individuals without a history of abuse.146,147 Recent twin studies have also suggested that genetic liability to MDD reflects not one, but three distinct symptom dimensions (psychomotor/cognitive, mood, and neurovegetative symptoms).148 Thus, GWAS that simply examine “depressed” cases versus controls may decrease the ratio of “signal to noise” by combining multiple disorder subtypes that vary in their genetic
