آزمون رفتاری ترس از آلودگی در اضطراب بیش از حد سلامت

Background and objectives Health anxiety is characterized by a preoccupation with the possibility of having a serious health condition or disease. Contemporary conceptualizations of health anxiety have improved in recent years to incorporate a fear of acquiring an illness; however, there is limited experimental data demonstrating the presence of fear of contamination among health anxious individuals. Method The present study utilized behavior approach tasks (BATs) to examine the degree to which contamination fear is present in elevated health anxiety. Participants were 60 undergraduate students who reported elevated health anxiety, contamination fear, or no anxiety about either health or contamination. Participants completed four BATS from which avoidance, anxiety, and disgust ratings were derived. Results Health anxious and contamination fearful individuals exhibited a similar degree of avoidance during the BATs. Contamination fearful participants reported significantly more anxiety and disgust relative to the non-anxious controls, but not the health anxious participants. Health anxious participants did not report more anxiety or disgust than the non-anxious participants.

Health anxiety is characterized by an exaggerated fear response in the presence of physical or cognitive cues that are perceived as indicative of threat toward the physical well-being of the individual. These cues motivate safety seeking behaviors aimed at providing assurance of good health and subsequently reducing anxious arousal. The Diagnostic and Statistical Manual of Mental Disorders (5th Edition; DSM-5; American Psychiatric Association [APA], 2013) has assigned the labels illness anxiety disorder and somatic symptom disorder to clinical manifestations of health anxiety. The preoccupation with threat toward physical health occurs in the absence of a diagnosed physical illness in the case of illness anxiety disorder. When health-focused anxiety occurs in the presence of a diagnosed physical illness, but the anxiety response is determined to be in excess of what is appropriate for the disorder, the DSM-5 specifies a diagnosis of somatic symptom disorder. A more thorough discussion of these disorders is beyond the scope of the present study, and indeed is unlikely to meaningfully contribute to the understanding of health anxiety in the absence of taxonomic examinations of these diagnostic labels. Thus, the present analysis approaches health anxiety as a psychological construct that underlies the terms used in the current diagnostic nomenclature. Theoretical and statistical analyses have provided support for the differentiation of health anxiety into two constructs: disease conviction and disease phobia ( Bianchi, 1973, Cote et al., 1996, Fergus and Valentiner, 2010, Kellner, 1986 and Pilowsky, 1967). The former term refers to an individual's often strongly held belief that they presently have a disease and is distinguished from a preoccupation that they will have a disease at some point in the future. Disease phobia refers to the fear of currently having an illness. This includes a fear of the consequences of having an illness, or more generally as a fear of simply being physically ill. Additionally, an individual may present with catastrophic thoughts about the personal costs of having an illness, the potential negative effect of illness on longevity, and a belief that illness will necessarily result in death. For example, an individual presenting with disease phobia may fear the repercussions that having cancer has on the ability to live a normal life, as well as the fact that the illness is likely to result in a long and especially painful death. Importantly, the emphasis in disease phobia is on the fear that the individual already has the illness, rather than the fear that one may acquire the illness at a later point. Contemporary theoretical models of health anxiety have furthered the concept of disease phobia and disease conviction as distinct factors in health anxiety (Barsky, 1992, Taylor and Asmundson, 2004 and Warwick, 1990). Nonetheless, health anxiety remains an evolving construct as improved conceptual models continue to better delineate its underlying factors (Abramowitz, 2008 and Noyes, 2005). Indeed, the cognitive-behavioral conceptualization of health anxiety has advanced the functional understanding of health anxiety which has subsequently led to improved treatment (Taylor & Asmundson, 2004). This model in particular has an advantage over other approaches (e.g., a categorical classification system based on symptom topography) by focusing on the functional relation between internal events, such as catastrophic cognitions, and behaviors that may serve to relieve these cognitions. Additionally, the cognitive-behavioral model of health anxiety is a flexible model which allows for the presence of constructs that are often excluded by the artificial boundaries presented by traditional categorical classification systems. Several studies have shown potential overlap in the phenomenology of contamination-based obsessive-compulsive disorder (OCD) and health anxiety (Brady et al., 2013, Deacon and Abramowitz, 2008, Olatunji, 2009 and Sulkowski et al., 2011). Contamination fear is met with attempts to either prevent contact with the source of contamination (e.g., not using public restrooms), or if contact cannot be prevented, to neutralize the threat posed by the contaminant (e.g., washing a predetermined number of times after touching a potential source of contamination). The overlap between contamination-based OCD and health anxiety suggests that contamination fear may function as a potential mechanism that drives the catastrophic cognitions and avoidant behaviors typical of excessive health anxiety. The studies cited above underscore the importance of determining the degree to which contamination fear contributes to excessive health anxiety; however, these studies are primarily cross-sectional designs using self-report data. Additional findings using observable, behavioral data in an experimental format are needed to further this line of research. To our knowledge, no studies have employed behavioral methods to experimentally test the degree to which contamination fear is present or absent in excessive health anxiety relative to contamination-based OCD. 1.1. Present study The present study used a behavioral test to determine the degree to which contamination fear is present in health anxiety. Health anxiety is often conceptualized as a fear that an individual presently has an illness or is in the process of developing an illness of unknown etiological origin, rather than a fear of acquiring an illness ( Noyes, Carney, & Langbehn, 2004). However, an increasing number of descriptive, correlational studies suggest that contamination fear may be more closely related to health anxiety than previously recognized ( Olatunji, 2009 and Thorpe et al., 2003). This study tested the role of contamination fear in health anxiety in an experimental format using a behavioral approach task (BAT). The BAT methodology has been used substantially in the study of various forms of psychopathology, especially anxiety ( Deacon and Olatunji, 2007, Koch et al., 2002, Olatunji et al., 2007 and Steketee et al., 1996). The procedure followed in the present study provides a test of contamination fear in health anxiety by comparing avoidance of sources of contamination by individuals reporting elevated health anxiety to that of individuals reporting elevated contamination fear, and a non-anxious comparison group. We predicted that the health-anxious and contamination-fearful individuals would refuse a greater number of steps than the non-anxious controls (NACs), and that they would not differ between each other, after controlling for the effects of neuroticism and trait anxiety. Additionally, we predicted that the health-anxious group would report more subjective anxiety than the contamination-fearful and NACs during the BATs, and conversely, that the contamination-fearful group would report more disgust.

3.1. Participant characteristics BAT variables and other study measure means and standard deviations are presented for the individual group samples in Table 1. Groups did not differ on age, F(2,56) = 2.21, p = 0.12, but they did differ on gender, χ2(2, N = 60) = 14.20, p < 0.001. This was due to greater frequency of males in the NAC group. Gender was included as a covariate in all subsequent analyses. Table 1. Unadjusted means for BAT variables, SHAI, PI-COWC, STAI-T, and EPQ-NR by group. Measure Health anxious Contamination fear Non-anxious Control M SD M SD M SD BAT avoidance 6.4 2.76 7.45 1.93 3.4 2.58 BAT anxiety 2.67 1.79 3.04 1.64 0.55 0.8 BAT disgust 3.39 1.77 3.84 1.85 0.95 0.92 SHAI 26.45 10.14 12.85 5.72 5.75 4.02 PI-COWC 13.20 7.82 23.26 7.47 3.30 3.80 STAI-T 51.90 11.27 40.26 8.09 32.05 7.76 EPQ-NR 8.5 2.37 6.26 2.75 2.15 1.93 Note. BAT = Behavioral Approach Task; SHAI = Short Health Anxiety Inventory; PI-COWC = Padua Inventory-Contamination and Washing Subscale; STAI-T = State-Trait Anxiety Inventory-Trait subscale; Eysenck Personality Questionnaire-Neuroticism. Table options 3.2. Zero-order correlations between BAT variables and study measures Zero-order correlations between the indices for avoidance, anxiety, and disgust, and the total scores for the study measures are presented in Table 2. A Bonferroni-correction was applied to these data (0.05/10 correlations) resulting in an alpha level of 0.005. All correlations reached this level (all p's < 0.005) with the exception of the SHAI (p = 0.02), which failed to meet the Bonferroni-corrected alpha level for the correlation with avoidance; however, it was significantly correlated with the remaining study variables, suggesting that no variable was uniquely related to the outcome measures. Table 2. Summary of zero-order intercorrelations for BAT variables, SHAI, and PI-COWC. Measure 1 2 3 4 5 1 BAT avoidance – 2 BAT anxiety 0.39** – 3 BAT disgust 0.44*** 0.88*** – 4 SHAI 0.30* 0.53*** 0.50*** – 5 PI-COWC 0.59*** 0.59*** 0.67*** 0.40** – Note. Correlations ≥ 0.39 are significant at the Bonferroni-corrected alpha level of 0.005. BAT = Behavioral Approach Task; SHAI = Short Health Anxiety Inventory; PI-COWC = Padua Inventory-Contamination and Washing Subscale. *p < 0.05. **p < 0.005. ***p < 0.001. Table options 3.3. Post-selection validation of group membership Multiple pairwise comparisons were performed to ensure appropriate group membership (health-anxious vs. contamination-fearful vs. NAC) after pre-study selection of participants. This analysis was performed using the participants' SHAI and PI-COWC total scores from the measures completed during the study. Six comparisons were made requiring a Bonferroni-correction to control for Type I error rate (0.05 divided by the six comparisons) resulting in an alpha level of 0.008. All group differences were significant at the 0.008 level, indicating appropriate group membership based on the two measures. 3.4. Effect of psychopathology on BAT avoidance The result of a one-way ANCOVA suggested a significant difference among groups on avoidance during the BATs, F(2, 54) = 7.95, p < 0.001, partial η2 = 0.23, after controlling for the effects of gender, neuroticism (EPQ-N), and trait anxiety (STAI-T). Three planned comparisons were performed to test the relation between the groups on avoidance. A Bonferroni correction was applied to these comparisons to account for multiple comparisons. Results of these comparisons showed that avoidance exhibited by the health-anxious (M = 6.7, 95% CI [5.29, 8.17]) and contamination-fearful (M = 7.25, 95% CI [6.1, 8.39]) groups was significantly greater than that of the NAC group (M = 3.23, 95% CI [1.69, 4.77]), but not significantly different from each other ( Fig. 1.). The observed power for the effect of group in this analysis was 0.94. Full-size image (14 K) Fig. 1. Adjusted mean avoidance scores on the BATs as a function of group. There were significant differences between the pathological groups and the non-anxious controls, but not between the health-anxious and contamination-fearful groups. Standard errors (±1 SE) are represented by the error bars attached to each column. HA = Health Anxious; CF = Contamination Fearful; NAC = Non-anxious Controls. Figure options 3.5. Effect of psychopathology on BAT anxiety and disgust Two separate one-way ANCOVAs were planned to test the effects of group on anxiety and disgust experienced during the BATs; however, violations of the assumption of homogeneity of variances was observed for both the anxiety and disgust ratings. A review of normality diagnostics and the graphical distribution of data between groups suggested that strong positive skewness of the anxiety and disgust ratings for the NAC group likely contributed to heterogeneity of variances. Logarithmic transformations of the mean anxiety and disgust ratings were conducted, consistent with the recommendation of Tabachnick and Fidell (2007). A subsequent one-way ANCOVA showed a significant difference in groups for anxiety, F(2, 54) = 6.1, p < 0.005, partial η2 = 0.18. Three planned comparisons were also conducted to test the group differences on anxiety. A Bonferroni correction was also applied to this analysis. These results demonstrated that anxiety reported by the contamination-fearful group (adjusted M = 0.54, 95% CI [0.44, 0.64]) was significantly greater than that of the NAC group (M = 0.25, 95% CI [0.12, 0.38]), but not significantly greater than that of the health-anxious group (M = 0.42, 95% CI [0.3, 0.55]). The health-anxious group did not report significantly greater anxiety than the NAC group (p = 0.39). The observed power for the effect of group in this analysis was 0.87. The ANCOVA for disgust revealed a significant difference among groups on disgust ratings, F(2, 54) = 6.29, p < 0.005, partial η2 = 0.19. Three planned comparisons were again conducted to test the group differences on disgust. These comparisons showed an effect on disgust similar to that of anxiety such that the contamination-fearful group (M = 0.62, 95% CI [0.53, 0.72]) was significantly greater than that of the NAC group (M = 0.34, 95% CI [0.21, 0.47]), but not significantly greater than that of the health-anxious group (M = 0.53, 95% CI [0.41, 0.64]). The health-anxious group did not report more disgust during the BATs than the NAC group (p = 0.23). The observed power for the effect of group in this analysis was 0.88. 4. Discussion 4.1. Summary of results The present study had two primary aims that further the understanding of health anxiety. The first aim was to clarify the relative effects of health anxiety and contamination fear on avoidance of contamination threat, as well as anxiety and disgust experienced in the presence of contamination threat. This was achieved by subjecting individuals reporting elevated health anxiety, contamination fear, or neither form of psychopathology to a series of BATs. The primary outcome variable for the BATs was the number of steps completed by the participants on each BAT, resulting in the creation of an index of avoidance. Avoidance was then compared across groups, the result of which failed to demonstrate differences between the health-anxious and contamination-fearful participants. Both pathological groups were significantly more avoidant during the BATs compared to the NAC group. This finding was consistent with the study hypothesis that health-anxious and contamination-fearful individuals would not exhibit differences in avoidance behavior in the presence of sources of contamination. Individuals with pronounced contamination fear, for whom the fear of illness by contamination is a source of motivation for avoidance (Rachman, 2004), did not differ in terms of avoidance of sources of contamination compared to those individuals with elevated health anxiety. This relation was found after controlling for the effects of gender, neuroticism, and trait anxiety. Moreover, these results were robust, even among an analogue sample. This result provides support for the conceptualization of health anxiety as a form of psychopathology partially characterized by a fear of acquiring illness via incidental contact with contaminants. A different pattern of findings was observed when evaluating anxiety and disgust experienced during the BATs. The results of these analyses only partially supported the study hypotheses. As predicted, individuals with elevated contamination fear were significantly more likely to experience disgust after completing the BAT steps, relative to the NAC group. This hypothesis was based on consistent findings in the literature demonstrating that contamination fear is partly a function of an individual's tendency toward, and aversion for, experiencing disgust (Rozin and Fallon, 1987 and Tolin et al., 2006). The finding that the contamination-fearful group, but not the health-anxious group, was significantly different from the NAC group on anxiety ratings during the BATs was surprising, and contrary to the study hypothesis. It was predicted that the health-anxious group would report more anxiety in response to the BATs relative to the NAC group; however, this effect failed to emerge. The hypothesis was based on the assumption that health anxiety, as the term implies, is primarily a problem of fear and anxiety, as well as studies showing a tendency for elevated anxiety and intolerance of this emotional response (Abramowitz et al., 2007a and Cox et al., 1999). The utilization of statistical controls to reduce the effects of third variables such as neuroticism, and emphasize the effect of contamination fear, may have artificially suppressed phenomenological variables that are central to the emotional response of these individuals, especially among the health anxious participants. Covarying for the effect of neuroticism may remove an important aspect of health anxiety that determines how the individual responds to contamination stimuli. While this may improve the understanding of how contamination fear functions in health anxiety, it may ultimately prevent understanding of health anxiety itself. Indeed, a post-hoc analysis of the means for the EPQ-N and STAI-T showed that the health-anxious group reported significantly higher neuroticism than the contamination-fearful and NAC groups. 4.2. Implications for clinical care The potential for improving conceptualization and clinical procedures in the treatment of health anxiety was a primary motivation for the present study. One important question addressed in the present study was whether consideration of fear of illness acquisition would be appropriate for a health-anxious population. The observation that individuals with elevated health anxiety were equally averse to contamination threat compared to contamination-fearful individuals suggests a need to consider fear of illness acquisition in treatment of health anxiety. Standard cognitive-behavioral conceptualization and clinical procedures exist for treatment of health anxiety and have been demonstrated effective for this population. For example, a cognitive-behavioral treatment approach for health anxiety might include imaginal exposure to having a feared illness (e.g., imagining that one has received a diagnosis of chronic obstructive pulmonary disease), exposure to illness-related cues (e.g., visiting a respiratory therapy clinic), and resisting the urge to seek additional tests (e.g., resisting the urge to schedule a chest X-ray). The present findings specifically support the inclusion of fear of acquiring illness through contagion as a consideration when formulating a case conceptualization and developing an exposure hierarchy. This may take the form of exposure to the possibility of acquiring an illness through contamination (e.g., exposure to the possibility of contact with an undetected fume by visiting a coal-burning power plant).