A dimensional perspective on the genetics of obsessive-compulsive disorder

Abstract

This review covers recent findings in the genomics of obsessive-compulsive disorder (OCD), obsessive-compulsive symptoms, and related traits from a dimensional perspective. We focus on discoveries stemming from technical and methodological advances of the past five years and present a synthesis of human genomics research on OCD. On balance, reviewed studies demonstrate that OCD is a dimensional trait with a highly polygenic architecture and genetic correlations to multiple, often comorbid psychiatric phenotypes. We discuss the phenotypic and genetic findings of these studies in the context of the dimensional framework, relying on a continuous phenotype definition, and contrast these observations with discoveries based on a categorical diagnostic framework, relying on a dichotomous case/control definition. Finally, we highlight gaps in knowledge and new directions for OCD genetics research.

Fig. 1. Liability threshold model. In orange the distribution of genetic-risk variants for OCD in the population from few to many.

In blue the distribution of OC symptoms in the population. The red dashed line depicts the diagnostic threshold. In a case/control GWAS (diagnostic framework) all individuals to the left (purple arrow) of the threshold would be considered a control, while all individuals to the right (green arrow), with a clinical diagnosis of OCD, would be considered a case. In a quantitative GWAS (dimensional framework) all individuals across the symptom- and genetic-risk-spectrum (yellow arrow) would be included, also weighting in subthreshold symptoms.

Fig. 2. OCD symptom sub-types.

OCD symptoms are classically defined by four subtypes, including (1) responsibility/fear of harm (obsessions regarding responsibility and checking compulsions such as asking for reassurance or checking that harm did not occur to someone), (2) contamination (obsessions about dirt or germs or other possible contaminants and cleaning compulsions), (3) symmetry/ordering (obsessions about symmetry or order and ordering, repeating or counting compulsions), and (4) taboo (violent, aggressive, sexual or religious obsessions with reassurance-reeking rituals, avoidance or checking for harm) [–4].

Fig. 3. Phenotypic associations between OCD and other psychiatric disorders, grouped by their DSM-5 categorization.

Comorbidity estimates between OCD and AN [95, 96], TS [88, 89], HD [128], TTM [91, 100], BDD [91, 100, 129], ExC [62, 109], PTSD [1, 100], ANX [1, 91], MDD [1, 91, 92], ASD [97], ADHD [1, 91, 99, 100], BP [100, 130], and SZ [130] are on the dashed lines, while the green boxes indicate the population prevalence of each disorder (OCD [2]; AN [131]; TS [132]; HD [82]; TTM [85, 86]; BDD [83, 84]; ExD [87, 105, 133]; PTSD [90, 134, 135]; ANX [1, 90]; MDD [1, 90]; ASD [97, 98]; ADHD [90]; BP [136, 137]; and SZ [138, 139]).

Fig. 4. Heritability estimates and genetic correlation estimates between OCD and related psychiatric disorders.

For SZ, MDD, BP, ADHD, TS, AN, ANX, and PTSD heritability estimates (green boxes) and genetic correlations (black lines) were calculated using genome-wide data [80, 81, 117, 118, 121]. Whereas heritability estimates (pink boxes) and genetic correlations (orange lines) for HD [76, 113] and BFRBs [106] were calculated using twin-data.