Hierarchical convergence of PTSD symptom clusters: A comparison of the numbing, dysphoria, and dysphoric arousal models of PTSD

Previous research has found support for 3 intercorrelated models of posttraumatic stress disorder (PTSD): the Numbing model (King, Leskin, King, & Weathers, 1998), the Dysphoria model (Simms, Watson, & Doebbelling, 2002), and the Dysphoric Arousal model (Elhai et al., 2011). Researchers have not yet examined whether the symptom clusters represented by these 3 models would converge together in the same manner at the second-order level. The aim of the present study was to compare the efficiency of 1- and 2-factor hierarchical structures for each of the 3 PTSD models. Two large samples (total N = 3,178) of college students who reported a traumatic event completed a measure of PTSD. Confirmatory factor analysis showed that Elhai et al.’s (2011) Dysphoric Arousal model fit the data best at the first-order, intercorrelated level. Examination of the fit of the hierarchical structures showed important differences among the models: the Numbing model fit a 2-factor second-order structure more efficiently than it fit a single-factor hierarchical structure, whereas the Dysphoria model was better explained at the second-order level by a single-factor, General PTSD structure. The Dysphoric Arousal model showed poor fit statistics with both the 1- and 2-factor hierarchical structures, suggesting that Elhai et al.’s (2011) model does not fit either the General PTSD representation or a 2-factor hierarchical conception of PTSD. The differences in fit of the 3 models at the hierarchical level have implications for the measurement, theoretical understanding, and treatment of PTSD.