اضطراب خصلتی، حساسیت انزجار و ساختار سلسله مراتبی ترس

This paper describes an evaluation of Taylor's (1998) hierarchic model of fears and its relationship to trait anxiety and disgust sensitivity (DS). In Study 1 (N = 420), a confirmatory factor analysis supported a hierarchic structure of fears. Next, an analysis using structural equation modeling indicated that trait anxiety is associated with claustrophobic and social fears, whereas DS is associated with all four fear subtypes examined (claustrophobic, social, blood–injection–injury and animal). However, trait anxiety and DS did not account for all variance shared by fear subtypes. The addition of a generalized “fear factor” accounted for significant residual shared variance between the four fear subtypes, beyond that accounted for by trait anxiety and DS. Study 2 (N = 213) generally replicated these results. Findings suggest that the hierarchic structural model of fears would benefit from inclusion of trait anxiety and DS as higher-order contributors to fearfulness.

There has been a growing interest in empirically derived taxonomic models of mental disorders in recent years (Brown, Chorpita, & Barlow, 1998; Hettema, Prescott, Myers, Neale, & Kendler, 2005; Kendler, Myers, Prescott, Neale, & Eaves, 2001; Kendler, Prescott, Myers, & Neale, 2003; Krueger, 1999, Vollebergh et al., 2001, Watson, 1999, Watson, 2005 and Zinbarg and Barlow, 1996). In such models, symptoms or disorders that co-vary are thought to possess a common vulnerability and are placed within the same diagnostic class. Conversely, symptoms or disorders that are quantitatively unrelated are not placed within the same diagnostic class, and provide a demonstration of the heterogeneity of psychopathology. Furthermore, shared variance between diagnostic classes suggests higher-order classes, indicating a hierarchic structure of mental disorders. For example, major depressive disorder, dysthymic disorder, and generalized anxiety disorder have been repeatedly empirically linked, intimating a diagnostic class labeled variably as “Anxious-Misery” (Kendler et al., 2003, Krueger, 1999 and Vollebergh et al., 2001) and “Distress Disorders” (Watson, 2005). Such empirically derived taxonomic models can be contrasted to the classification system in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV-TR; American Psychiatric Association, 2000), which has been criticized by several researchers for its basis in phenomenological similarity rather than empirical evidence (e.g., Brown & Chorpita, 1996; Watson, 2005). Taylor (1998) proposed a hierarchic structural model of fears based on a review of factor analyses of fear surveys and behavioral-genetic studies in the fear and anxiety literature (e.g., Kendler, Neale, Kessler, & Heath, 1992; Staley & O’Donnell, 1984). According to this model, there is a “hierarchy of causal factors” consisting of general and specific mechanisms that influence one's proneness to developing fears. At the lowest level of the hierarchy are factors that are uniquely associated with specific fears (e.g., blood, snakes, and elevators). Intermediate factors are vulnerabilities associated with a set of specific fears. For example, fears of insects, snakes, and bats are thought to co-vary due to a common vulnerability to developing animal fears. Taylor identified four intermediate-level fear subtypes in his model, based on the findings of factor analytic studies of fear questionnaires: social, animal, blood–injury–illness (BII) and situational (or agoraphobic) fears (Arrindell, Pickersgill, Merckelbach, & Ardon, 1991). In addition to the four lower-order fear subtypes, Taylor's (1998) hierarchic model of fears suggests that higher-order factors (e.g., neuroticism) represent a general proneness to developing fears. Taylor noted that several behavior genetic studies have demonstrated a two-tier hierarchy of genetic factors: specific genetic contributions to each fear or phobia type and a general genetic factor associated with all fears or phobias (e.g., Kendler et al., 1992). In addition, Taylor reviewed two factor analytic studies that supported a hierarchical structure of fears using fear surveys (Staley & O’Donnell, 1984; Zinbarg & Barlow, 1996). Subsequent evidence for a hierarchical model of fears was provided by a recent study (Cox, McWilliams, Clara, & Stein, 2003) that used exploratory and confirmatory factor analyses of 19 common fears using data from the National Comorbidity Survey (Kessler et al., 1994). Although the hierarchic model of fears is helpful in providing a descriptive framework for classification, a significant drawback is the ambiguity surrounding the meaning and identification of higher-order factors that are likely to have a bearing on fear proneness. Results of behavioral genetic studies and factor analytic studies such as Cox et al. (2003) infer that a unitary, higher-order mechanism is responsible for the shared variance between fear subtypes. However, there are indications that fear proneness is affected by an amalgamation of several additive factors. For example, a recent factor analytic study of a fear questionnaire (Cutshall & Watson, 2004), significant and positive partial correlations were found between fear subtypes, even after controlling for neuroticism. To explain this observation, the authors contended that unmeasured factors might account for the common variance between fears that are unrelated to neuroticism. The implication of this finding is that the “general fear factor” identified by Cox et al. may in fact represent multiple, differentiable vulnerabilities to the development of fear. Certainly, clarifying such sources of covariance between fears would go far in developing our understanding of the structure of fears. The purpose of the current study was to test a hierarchic structural model of fears and its relationship to two likely contributors to fear proneness, trait anxiety and disgust sensitivity. Trait anxiety refers to individual differences in anxiety proneness in response to stressful situations (Spielberger, 1983). Trait anxiety, as well as its more broadly encompassing cousin neuroticism, has an established relationship with a variety of fears. For example, trait anxiety and neuroticism typically exhibit an association with agoraphobic (Chambless, 1985 and Kendler et al., 2002; Muris, Merckelbach, & Rassin, 2000) and social fears and phobias (Cutshaw & Watson, 2004; Shean & Lease, 1991; Stemberger, Turner, Beidel, & Calhoun, 1995). A similar relationship has been found for BII and animal fears (Olatunji, 2006 and Page, 1994), although associations with these fear subtypes have generally been smaller than for social fears (Cutshaw & Watson, 2004; Kendler et al., 2002) and occasionally non-significant (De Jongh et al., 1998 and Mulkens et al., 1996; Thorpe & Salkovskis, 1995). The observation that disgust sensitivity is associated with fears and phobias is more recent. Disgust sensitivity (DS) is defined simply as individual differences in sensitivity to the emotion of disgust (Haidt, McCauley, & Rozin, 1994; Rozin, Fallon, & Mandell, 1984). Disgust sensitivity is conceptualized as an enduring trait, has temporal stability (Rozin, Haidt, McCauley, Dunlop, & Ashmore, 1999), and is resilient to secondary treatment effects (de Jong, Andrea, & Muris, 1997). There is also some evidence that DS is moderately heritable, in that parent and child reports of DS are moderately and positively correlated (Davey, Forster, & Mayhew, 1993; Rozin et al., 1984; but see de Jong et al., 1997). For example, in one interesting study, Davey et al. found that parental food-related DS, but not parental fear of spiders, predicted the fear of spiders in offspring. Disgust is thought to be protective, by promoting the avoidance of things that may indicate contamination and disease, such as feces or maggots (Matchett & Davey, 1991). In that sense, both the emotions of disgust and fear have functional value by motivating avoidance of potentially dangerous stimuli or situations (Woody & Teachman, 2000). Furthermore, an early study demonstrated a causal link between the experience of disgust and a subsequent development of fear (Webb & Davey, 1992). Consequently, the relationship between phobic behavior and DS has received a good deal of attention in recent years (e.g., Phillips, Senior, Fahy, & David, 1998; Woody & Teachman, 2000). For example, DS has been found to have a positive relationship to animal fears (Arrindell, Mulkens, Kok, & Vollenbroek, 1999; Davey, 1994, Davey et al., 1993 and Klieger and Siejak, 1997; Sawchuk, Lohr, Tolin, Lee, & Kleinknecht, 2000) and BII fears (de Jong & Merckelbach, 1998; Muris, Merckelbach, Schmidt, & Tierney, 1999; Sawchuk et al., 2000). Furthermore, DS tends to be higher in individuals reporting animal phobias (de Jong et al., 1997 and Merckelbach et al., 1993) and BII phobias (Schienle, Schäfer, Walter, Stark, & Vaitl, 2005; Tolin, Lohr, Sawchuk, & Lee, 1997). Interestingly, associations have also been found between disgust sensitivity and other anxiety-relevant constructs, such as obsessive-compulsive disorder (Mancini, Gragnani, & D’Olimpio, 2001; Muris, Merckelbach, Nederkoorn, Rassin, & Horselenberg, 2000; Tsao & McKay, 2004), agoraphobia (Muris, Merckelbach, Nederkoorn, et al., 2000; Muris, Merckelbach, & Rassin, 2000), claustrophobia (Davey & Bond, 2006), height fears (Davey & Bond, 2006) and separation anxiety (Muris et al., 1999). In a review of the literature concerning disgust and fears, Woody and Teachman (2000) voiced concern that the relationship between DS and fears may be a spurious artifact due to the effects of trait anxiety on both DS and fearfulness. Certainly, several studies have reported a moderate and positive correlation between DS and trait anxiety or similar constructs such as neuroticism (Druschel & Sherman, 1999; Haidt et al., 1994, Muris et al., 1999 and Olatunji et al., 2005; Schienle, Schäfer, & Stark, 2005). Furthermore, there is evidence that the relationship between DS and fears is reduced after accounting for trait anxiety (Muris et al., 1999; cf. Mulkens et al., 1996). On the other hand, DS has been demonstrated to predict specific fears (Muris et al., 1999) and spider phobias (Olatunji, 2006) even after controlling for trait anxiety. Taken together, these findings support the contention that DS partially moderates the relationship between trait anxiety and fears (Baron & Kenny, 1986). In terms of Taylor's (1998) hierarchic model of fears, DS as a mediator of trait anxiety would be placed on an intermediate level between first-order fear subtypes and higher-level constructs such as trait anxiety. Although one study has examined the correlations and partial correlations between DS, trait anxiety, and fears (Muris et al., 1999), no study to date has examined the relationship among these constructs in a comprehensive structural model. In the current study, we examined the relationship of DS and trait anxiety to four common fear subtypes as measured by select items from the self-report Fear Survey Schedule-III (FSS-III; Wolpe & Lang, 1964). Fifteen items from the FSS-III with one addition (“long tunnels”) were chosen to represent the four fear constructs in Taylor's (1998) model: social fears, BII fears, animal fears, and claustrophobic fears (e.g., Arrindell et al., 2003). These 16 items were chosen to maximize factor homogeneity, based on a literature review and experience with the FSS-III in our lab. Claustrophobia is considered a component of the more broadly inclusive and relatively less-stable agoraphobia or “situational” factor (Arrindell et al., 1991 and Taylor, 1998) and was preferred in this study in order to reduce interpretive ambiguity. The first aim of this study was to replicate prior studies (e.g., Cox et al., 2003) that have identified a hierarchic structure of fears. It was expected that a confirmatory factor analysis (CFA) of items from the FSS-III would support a hierarchic model comprised of 16 feared situations, four first-order fear factors (social, BII, animal, and claustrophobia) and one higher-order factor that is common to all fears, i.e., a “general fear factor.” The conceptual path diagram is illustrated in Fig. 1.