introduced the Health Anxiety Inventory - University of Wyoming ...

92 abramowitz et al.BAI, ASI-R, and BVS failed to demonstratediagnostic utility (range in AUC=.32–.42).DiscussionClinical observations suggest that health anxiety ispresent in a variety of psychological disorders, yetvery little empirical research on health anxietyexists in the literature. One reason for this dearth ofresearch is that very few measures have beendeveloped to assess the construct of health anxiety.The SHAI is a novel instrument that assesses thecognitive and behavioral features of health anxiety.It possesses good psychometric properties amongindividuals with HC and among nonclinicals. Thepresent study, however, is the first to evaluate thismeasure in an anxiety disorder patient sample. It isalso the first study to empirically consider the roleof health anxiety in the anxiety disorders. Giventhat the innate function of anxiety is to protect theorganism against injury and harm, it is no surprisethat health concerns emerge as prominent foci offear and worry across a number of these conditions.The results of the present study can thereforeinform the conceptualization and treatment ofclinical anxiety.Confirmatory factor analyses provided equivalentsupport for two- and three-factor models of theSHAI reported in previous studies by Salkovskis etal. (2002) and Abramowitz et al. (in press). On thebasis of parsimony, we elected to examine the twofactormodel originally reported by Salkovskis et al.(2002) assessing (a) the perceived likelihood ofacquiring a serious illness, intrusive thoughts, andbody vigilance (Illness Likelihood) and (b) catastrophicbeliefs about the anticipated burden ofhaving a serious illness (Negative Consequences).Our analyses revealed satisfactory convergent anddivergent validity for the total SHAI and the twofactors. Specifically, the total score and IllnessLikelihood factor were moderately to stronglyassociated with theoretically overlapping constructs(i.e., body vigilance, fear of cardiovascular andrespiratory symptoms, worry) and weakly relatedto constructs considered theoretically distinct fromhealth anxiety (i.e., social anxiety, obsessive-compulsivesymptoms, and fear of cognitive dyscontrol). TheNegative Consequences factor demonstrated significantrelationships (weak to moderate in magnitude)with fears of cardiovascular symptoms and worry.These results indicate that the SHAI possesses goodreliability and validity in a clinical sample.The findings from our regression highlight thetriumvirate of anxiety sensitivity, body vigilance,and worry as integral to health anxiety and HC.This is consistent with cognitive-behavioral modelsof HC which posit that intense health anxiety arisesas a result of the tendency to catastrophicallymisinterpret harmless bodily sensations as threatening.As a result of putting oneself on a heightenedstate of alert for signs of the feared illness, theindividual becomes body vigilant. Such increasedattentional focus on internal sensations raises theindividual’s awareness of normal bodily perturbations(body “noise”; e.g., ordinary fluctuations ingastrointestinal sensations, vestibular functioning,and so on), initiating a self-sustaining vicious cycleof catastrophic misinterpretation→anxiety→bodyvigilance→notice internal sensations, and so on.This has the long-term effect of strengthening theillness worry, which might be compounded bycatastrophic thinking and an intolerance of uncertaintyregarding the negative consequences of beingill (e.g., Langlois & Ladouceur, 2004).The second aim of the present study was toexamine the role of health anxiety in the psychopathologyof anxiety disorders. Consistent withconceptualizations of HC as severe health anxiety(e.g., Taylor & Asmundson, 2004), patients withthis disorder reported higher SHAI total scoresrelative to those with anxiety disorders. The findingthat PD patients evidenced higher SHAI total scoresrelative to patients with other anxiety disorders isalso consistent with formulations of PD as involvinghealth anxiety (Schmidt et al., 2003). Although wefound that patients with OCD did not show elevatedSHAI total scores, OCD patients did not differsignificantly from PD and HC patients in theirbeliefs about the probability of becoming ill. This isconsistent with clinical and research observationsthat, like the concerns of patients with HC and PD,obsessional preoccupation in OCD often focuses onhealth and illness (Abramowitz et al., 1999). Incontrast, patients with HC demonstrated elevatedbeliefs regarding the negative consequences of illhealth relative to all of the other patient groups. Thispattern of results suggests that whereas HC, PD, andOCD all involve preoccupation with body sensationsand the probability of becoming ill, HC (morethan these other disorders) involves catastrophicbeliefs about the consequences of acquiring thefeared illness.These findings have direct relevance to theassessment and treatment of HC and anxietydisorders using cognitive-behavioral therapy. Inparticular, treatment providers should includeassessments of health-focused anxiety when workingwith anxious patients, and consider the role ofsuch concerns as antecedents of avoidance andsafety-seeking behavior. Moreover, within the frameworkof available treatment manuals, cognitivetherapy techniques can be incorporated to addressAuthor's personal copy