We compared the fit of the ACE common pathway (AIC = 8259.76, df = 2604, p = 0.02) and independent pathway (AIC = 8243.88, df = 2595, p = 0.22) models to the correlated factor model (AIC = 8265.35, df = 2568) for the OC trait dimensions. The independent pathway model was the most parsimonious. Model fit was unchanged by removing the Hoarding dimension (data not shown). As shown in Fig. 1, the majority of shared variance for each dimension was accounted for by genetic factors (Ac = 32–58%), except Cleaning/Contamination where common environment (Cc) accounted for the majority of the variance (36%). Genetic influences (As) accounted for the majority of dimension-specific variance only for Hoarding and Superstition (19–26%). For all other dimensions, unique environment influences (Es) accounted for the majority of the dimension-specific variance (Es = 17–38%). Variance estimates from the independent model are presented in Supplemental Table 4.
