We tested whether (1) the TOCS captured OC dimensions, (2) if these dimensions were heritable as well as co-heritable, (3) if these dimensions had shared and unique etiological factors, and (4) whether shared variance between the dimensions was best explained by a common pathway (single global latent trait) or independent pathway model. To address these questions, we factor analyzed the TOCS in a population-based sample (n = 16,718) of youth to identify OC dimensions. We used structural equation modeling in 220 twin pairs to examine the heritability of the individual dimensions, their co-heritability and test the fit of the independent and common pathway models. Current evidence indicates that the best fitting model is unclear in adults15,16 and unknown in youth. The expectation was that if a common pathway model fit best, then shared etiological factors were mediated by an underlying latent trait, whereas if an independent pathway model fit best, then shared etiological factors influenced dimensions directly. If the variance shared by multiple dimensions is captured by a global latent trait, research designs should focus on that latent trait. Conversely, if unique genetic factors influence each OC trait dimension directly, research designs should focus on the individual OC dimensions.
