Predicting genetic loading from symptom patterns in obsessive- compulsive disorder: a latent variable analysis

Area of Science:

• Psychiatry
• Genetics
• Computational Biology

Background:

• Familial and genetic factors influence some obsessive-compulsive disorder (OCD) symptom dimensions.
• Previous studies relied on exploratory analyses of symptom categories, not individual symptoms.

Purpose of the Study:

• To identify meaningful patient subgroupings in obsessive-compulsive disorder (OCD).
• To utilize a novel latent variable mixture model for subgroup identification.

Main Methods:

• Confirmatory factor analysis of a 65-item OCD symptom inventory from 398 probands.
• Latent variable mixture model analysis using five symptom factors to identify OCD subpopulations.

Main Results:

• Two statistically separate OCD subpopulations were identified.
• One subpopulation showed a significantly higher proportion of affected relatives and earlier onset.
• This group also presented with more severe OCD symptoms, psychiatric comorbidity, and impairment.

Conclusions:

• This research paradigm identified characteristics of individuals with familial OCD.
• Findings may aid genome-wide linkage and association studies for OCD.
• The approach may serve as a model for other symptom-based medical disorder studies.