Anxiety sensitivity and posttraumatic stress among traumatic event-exposed youth

Abstract

Total and factor scores of the Childhood Anxiety Sensitivity Index (CASI) were examined in relation to posttraumatic stress symptom levels within a community-based sample of 68 (43 females) traumatic event-exposed youth between the ages of 10 and 17 years (M_age=14.74 years). Findings were consistent with hypotheses; global anxiety sensitivity (AS) levels, as well as disease, unsteady, and mental incapacitation concerns, related positively to posttraumatic stress levels, whereas social concerns were unrelated to symptom levels. These results suggest that fears of the physical and mental consequences of anxiety are associated with relatively higher levels of posttraumatic stress subsequent to traumatic event exposure. Findings are discussed in terms of potential implications for the role of AS in developmentally sensitive etiological models of posttraumatic stress disorder (PTSD).

Introduction

Anxiety sensitivity (AS) is a cognitive vulnerability characteristic reflecting fear of anxiety and anxiety-related sensations (Reiss & McNally, 1985). Consistent with theory highlighting the role of cognitive factors in the maintenance of posttraumatic stress (Ehlers & Clark, 2000), individual differences in AS may evidence meaningful associations with posttraumatic stress symptom levels among traumatic event-exposed persons (Taylor, 2003). Specifically, AS may be relevant to posttraumatic stress in at least two ways. First, AS may amplify or enhance symptoms following traumatic event exposure (Fedoroff, Taylor, Asmundson, & Koch, 2000). For instance, compared with someone low in AS, a high AS individual may be relatively more frightened and upset by re-experiencing symptoms (e.g., flashbacks), which may exacerbate such symptoms. Second, persons high in AS may be more apt to avoid reminders of a traumatic event, thereby preventing extinction of learned trauma-related fear (Lang, Kennedy, & Stein, 2002) and/or affective processing of, and recovery from, a traumatic event (Brewin, Dalgleish, & Joseph, 1996). Indirect evidence for this perspective comes from work conducted with non-clinical samples where AS has been shown to predict avoidance of emotionally salient events (Wilson & Hayward, 2006).

Empirical evidence supports the perspective that AS (assessed using the Anxiety Sensitivity Index [ASI; Reiss, Peterson, Gursky, & McNally, 1986]) is related to posttraumatic stress symptoms among adults. Between-groups studies indicate AS is elevated among adults with, versus without, posttraumatic stress disorder (PTSD; Asmundson & Carleton, 2005; Hinton, Pich, Safren, Pollack, & McNally, 2005). For instance, Taylor, Koch, and McNally (1992) compared ASI scores between persons with six different types of anxiety psychopathology (i.e., panic disorder, PTSD, generalized anxiety disorder, obsessive–compulsive disorder, social phobia, or simple phobia). AS was comparably elevated among persons with PTSD (M=31.6; SD=12.8) and those with panic disorder (M=36.6; SD=12.3), which is a population that generally evidences high AS. These scores were significantly higher than those observed in the comparison clinical groups, as well as norms from non-clinical samples (M=17.8; SD=8.8; Petersen & Reiss, 1987). Importantly, individuals in the PTSD group did not meet the criteria for panic disorder, but the presence of other co-morbid anxiety disorders (e.g., OCD) was not methodologically or statistically controlled, which may have influenced the results. Independent research also has found elevated AS among persons with PTSD, even compared with traumatic event-exposed individuals without PTSD, suggesting AS is related to posttraumatic stress levels above and beyond associations with traumatic event exposure alone. Among female victims of interpersonal violence (IPV), total ASI scores were significantly higher in a PTSD-positive group (M=30.7; SD=16.5), compared with an IPV-exposed PTSD-negative group (M=17; SD=11.6; Lang et al., 2002). In terms of specific facets of AS (i.e., mental incapacitation, physical, and social concerns factors on the adult ASI; Zinbarg, Brown, & Barlow, 1997), IPV-exposed adults with PTSD, as compared with IPV-exposed adults without PTSD, endorsed higher physical and mental incapacitation concerns. Research also has focused on continuous relations between AS and posttraumatic stress among traumatic event-exposed adults.

Several studies have documented positive correlations between AS and posttraumatic stress symptom levels among traumatic event-exposed adults (e.g., Feldner, Zvolensky, Schmidt, & Smith, in press; Hagh-Shenas, Goodarzi, Dehbozorgi, & Farashbandi, 2005). Asmundson, Norton, Allerdings, Norton, and Larsen (1998) examined AS and posttraumatic stress among 121 adults who had experienced a work-related injury. Total ASI scores were positively related to overall symptom severity as well as symptom level within each of the three posttraumatic symptom clusters (i.e., re-experiencing, avoidance, and arousal). Similarly, among 59 traumatic event-exposed adults, higher AS was associated with increased negative posttraumatic cognitions (e.g., negative cognitions about the self and world; self-blame) as well as severity of posttraumatic stress symptoms (Nixon & Bryant, 2005). Subfactor scores on the ASI also have been evaluated in relation to posttraumatic stress; Feldner, Lewis, Leen-Feldner, Schnurr, and Zvolensky (2006) observed that total ASI (r=.28) and physical concerns (r=.31) scores significantly related to posttraumatic stress level among 61 trauma-exposed rural adults.

Collectively, theory and research suggest that among trauma-exposed persons, AS is positively related to posttraumatic stress. However, to the best of our knowledge, this association has not been examined among youth. It is important to address this gap for multiple reasons. First, adolescents are at significant risk for exposure to traumatic events; recent estimates suggest as many as 68% of youth have been exposed to a traumatic event by age 16 (Copeland, Keeler, Angold, & Costello, 2007). Recognition of the considerable rates of trauma and posttraumatic stress among youth has resulted in a pressing need to articulate developmentally sensitive etiological models in this domain (Salmon & Bryant, 2002; van der Kolk, 2005). To achieve this objective, factors thought to relate to posttraumatic stress must be empirically examined among youth, as such processes identified in adults cannot be assumed to exist or function identically at other stages of development (Cicchetti & Rogosch, 2002). Adolescence is a theoretically relevant developmental stage in which to study anxiety, as the biopsychosocial changes related to puberty (e.g., onset of dating and substance use, sleep deprivation, changes in the hypothalamic–pituitary–adrenal axis, increased emotional lability, and negative affectivity; Brooks-Gunn, Graber, & Paikoff, 1994; Carskadon et al., 2002; Hayward, 2003) occurring during this stage make it a “critical period” during which anxiety problems often emerge (Leen-Feldner, Reardon, Hayward, & Smith, 2008). Thus, this period of enhanced vulnerability to anxiety is particularly important to study in order to advance our understanding of recovery from posttraumatic stress across the lifespan (Hayward & Sanborn, 2002). Indeed, empirical work has suggested that posttraumatic stress problems may be more likely to occur subsequent to traumatic event exposure during adolescence compared with other developmental periods (e.g., childhood; Copeland et al., 2007). For these reasons, it is important and timely to increase our understanding of the link between AS and posttraumatic stress among adolescents.

The current study sought to examine the relation between AS, its subfactors, and posttraumatic stress levels among a sample of traumatic event-exposed adolescents. To examine these processes in a developmentally sensitive design, measures developed for, and validated among, youth were employed. In particular, the psychometrically sound Childhood Anxiety Sensitivity Index (CASI; Silverman, Fleisig, Rabian, & Peterson, 1991), adapted from the adult-focused ASI (Reiss et al., 1986) for use with youth, was employed to measure AS. A factor analysis of the CASI using several samples of American and Australian clinical and non-clinical youth suggested that the physical concerns facet identified on the ASI is comprised of two distinct factors, resulting in four factors in addition to a global factor: disease concerns, unsteady concerns, mental incapacitation concerns, and social concerns (Silverman, Goedhart, Barrett, & Turner, 2003). Consistent with research suggesting physical and mental incapacitation concerns are most consistently linked to symptoms of PTSD among traumatized adults (e.g., Feldner et al., 2006; Lang et al., 2002), the disease, unsteady, and mental incapacitation concerns factors, along with the global index of AS, were expected to relate positively to posttraumatic stress levels in the current sample of youth. Finally, in an effort to examine the unique association between AS and trauma symptoms, the incremental predictive validity (Haynes & Lench, 2003) of AS was examined relative to several factors associated with posttraumatic stress levels among youth, including gender, age (Kilpatrick et al., 2003), and time since traumatic event exposure (Zatzick et al., 2006). Further, in light of empirical evidence linking negative affectivity, or the temperamentally based proclivity for experiencing negatively valenced affective states (e.g., anger, sadness, anxiety), to trauma symptoms among youth (e.g., Weems et al., 2007), this factor also was included as a covariate. Overall, this approach allowed for evaluation of the specific association between AS and trauma symptoms after controlling for variance accounted for by these theoretically relevant factors.