Hoarding is characterized by difficulty discarding or parting with possessions, resulting in clutter that impedes the normal use of living spaces [Frost and Hartl, 1996]. Although hoarding can be a symptom of multiple neurological and psychiatric disorders, such as obsessive-compulsive disorder (OCD), mounting evidence suggests that once other primary causes are ruled out, hoarding can also be a discrete disorder [Mataix-Cols et al., 2010]. The DSM-V taskforce is currently considering the creation of a new diagnostic category named hoarding disorder (www.dsm5.org). We have recently shown that hoarding is highly prevalent and heritable, with genetic factors accounting for approximately 50% of its variance [Iervolino et al., 2009]. Very few studies have examined the genetic architecture of hoarding, and their results have been largely inconsistent [Zhang et al., 2002;Samuels et al., 2007;Pertusa et al., 2010].
We conducted a genome-wide association study for hoarding traits in a large cohort of Caucasian twins (see [Iervolino et al., 2009] for more details). A sub-sample of 3,410 participants had been genotyped and was included in this study. Of these participants, 2,350 were singletons (either MZ as MZ twins are genetically identical, only one member of the twin pair was genotyped-or DZ without the co-twin), predominantly female (91.8%), with a mean age of 56.8 years (SD ΒΌ 12.6; range ΒΌ 17-85). All participants completed the Hoarding Rating Scale-Self-Report (HRS-SR [Tolin et al., 2010]), a brief self-administered instrument consisting of five items (clutter, difficulty discarding, excessive acquisition, distress, and impairment). Each item is scored from 0 (none) to 8 (extreme) with a total score ranging from 0 to 40. Scores above 14 indicate severe hoarding with sensitivity and specificity of 0.97. HRS scores were obtained for each member of a MZ twin pair but only one twin was genotyped. In order to minimize measurement error and get a better estimate, the HRS-SR scores of each MZ twin pair were treated as biological duplicates and averaged. The mean HRS-SR score in our cohort was of 2.83 (SD ΒΌ 4.2; range ΒΌ 0-35).
Subjects were genotyped using either Illumina 317 K (n ΒΌ 1,348) or 610 K (n ΒΌ 2,062) BeadChips. All subsequent analyses were