- Email: [email protected]
Inflated perception of responsibility for harm in obsessive–compulsive disorder
Anxiety Disorders 15 (2001) 259 ± 275
Inflated perception of responsibility for harm in obsessive±compulsive disorder Edna B. Foaa,*, Nader Amirb, Karen V.A. Bogertc, Chris Molnarc, Amy Przeworskic a
Department of Psychiatry, University of Pennsylvania, Philadelphia, PA b Department of Psychology, University of Georgia, Athens, GA, USA c Department of Psychiatry, Allegheny University of the Health Sciences, USA Received 9 April 1999; received in revised form 23 May 1999; accepted 21 April 2000
Abstract The present study examined the proposition that inflated responsibility is implicated in obsessive ± compulsive disorder (OCD). Compared to non-anxious control participants (NACs), and an anxious control group with generalized social phobia (GSPs), we predicted that individuals with OCD (OCs) would exhibit a greater urge to rectify situations involving potential risk, would report more distress upon leaving such situations unrectified, and would feel more personal responsibility if the unrectified situations resulted in harm. Fifteen OCs, 15 NACs, and 15 GSPs completed the Obsessive ± Compulsive Responsibility Scale (OCRS), which included low-risk, OC-relevant, and high-risk situations. Compared to NACs and GSPs, OCs reported more urges, distress, and responsibility in low-risk and OC-relevant situations; no group differences were detected on high-risk situations. GSPs and NACs differed only in their responsibility in OCrelevant situations, with GSPs reporting higher responsibility than controls. Our results suggest that compared to OCs, NACs, and GSPs can better differentiate between situations that merit concern and ones that do not. D 2001 Elsevier Science Inc. All rights reserved. Keywords: Obsessive-compulsive disorder; Cognitive distortions; Inflated responsibility
* Corresponding author. Department of Psychiatry, University of Pennsylvania, 3535 Market Street, Philadelphia, PA 19104, USA. Tel no.: 215-746-3327; fax: 215-746-3311. Email address: [email protected] (E.B. Foa). 0887-6185/01/$ ± see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 0 8 8 7 - 6 1 8 5 ( 0 1 ) 0 0 0 6 2 - 7
260
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
1. Introduction Clinical observations suggest that individuals with obsessive ± compulsive disorder (OCs) suffer from excessive personal responsibility for harm to others. Foa, Steketee, and Young (1984) noted that individuals with agoraphobia generally fear harm that might come to themselves, whereas OCs are often preoccupied with harm that may come to others. Such clinical observations led Salkovskis (1985, 1989) to propose that an exaggerated sense of responsibility for negative thoughts (e.g., ``I might stab my child'') and anticipated negative events (``I will run over a pedestrian if I'm not careful'') underlie the psychopathology of OCD. Similarly, Rachman (1993) suggested that many patients with obsessive ± compulsive disorder (OCD), especially those with checking compulsions, experience urges to ritualize only in those circumstances in which they assume personal responsibility for safety. Specifically, Salkovskis (1985) argued that OCs' distressing thoughts are typically related to feelings of responsibility for harm to others. In order to reconcile the discrepancy between his proposition and the observation that some OCs are not concerned with others' safety, Salkovskis extended the customary use of the concept of responsibility to include feelings of liability for one's own harm. In a later exposition of this model, Salkovskis, Richards, and Forrester (1995) further emphasize the central role of the responsibility-neutralizing link in the etiology of OCD. They suggested that obsessions of OCs are not only characterized by a heightened sense of responsibility, but are also distinguished by the belief that not acting in a potentially harmful situation is akin to directly causing the harm. To explore the relationship between responsibility and neutralizing, Freeston, Ladouceur, Thibodeau, and Gagnon (1992) administered the Cognitive Intrusions Questionnaire to college students. Factor analysis of responses on the questionnaire yielded five factors, only one of which was significantly related to questions of compulsive activity. This factor, the valence factor, represented the person's evaluation of the intrusion as a function of responsibility and guilt. The highest loading item on this factor was ``To what extent would you feel responsible if the thought content were to happen?,'' therefore, the authors interpreted the results as demonstrating a link between responsibility and compulsive behavior. Rachman, Thordarson, Shafran, and Woody (1995) examined the relationship between responsibility and OCD symptoms in a sample of college students. To this end, they devised a questionnaire that measured four aspects of responsibility: responsibility for harm, responsibility in social contexts, a positive outlook towards responsibility, and thought±action fusion. Only the thought± action fusion aspect correlated with OC symptom severity as measured by the Maudsley Obsessive ± Compulsive Inventory (MOCI; Hodgson & Rachman, 1977). Consequently, the authors concluded that inflated responsibility is not a general characteristic of OCD, but, rather, may be specific to personally relevant situations.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
261
To explore the role of responsibility in the performance of compulsions, Ladouceur et al. (1995) carried out two experiments in which they manipulated degree of responsibility in college students. In the first experiment, participants were asked to determine whether they had heard a sound before. They were allowed to listen to the sound as many times as they wanted (i.e., check) before responding. Although the manipulation was successful in producing more feelings of responsibility in the high-responsibility (HR) group, participants in this group did not exhibit more checking behavior than participants in the lowresponsibility (LR) group. The authors attributed the negative results to the absence of a relationship between the responsibility manipulation and the performed task. Therefore, in a second experiment, participants in the HR group were instructed to sort different kinds of pills seemingly intended for public consumption. Participants in the LR group were told that the pill-sorting task was designed to inform the experimenter about color perception. Participants in the HR condition exhibited more preoccupation with making errors and were more anxious during the task compared to those in the LR condition. However, no group differences emerged on self-report of doubting, urges to check, and time required to complete the task. At best, the studies described above demonstrate a limited relationship between responsibility and OCD-like symptoms. Moreover, it is important to note that these studies were conducted with nonpatient students, and therefore the significance of the findings for our understanding of OCD is unclear. To date, only two studies have utilized individuals diagnosed with OCD to study the role of responsibility (Lopatka & Rachman, 1995; Shafran, 1997). Both studies examined the hypotheses that inflating responsibility would increase urge to neutralize, the degree of discomfort experienced, and the estimated probabilities of bad outcomes. Deflating responsibility was hypothesized to decrease urge, discomfort, and estimated probabilities. In the study of Lopatka and Rachman (1995), participants were instructed to perform a task at home that usually evoked an urge to check under ``highresponsibility'' instructions (HRI), ``low-responsibility'' instructions (LRI), and ``controlled'' instructions (CI). Responsibility was manipulated using contracts in which the experimenter assumed complete responsibility for potential negative events following the task (LRI) or the participant assumed complete responsibility for such events (HRI). Although the HRI condition produced a higher degree of responsibility compared with the CI condition, it did not produce stronger urged to check, higher discomfort, or higher estimated probability of harm. In contrast, the LRI reduced the subjects' urges to check, discomfort, and estimated probability of harm. Thus, the hypothesis that responsibility plays a role in OCD was partially supported. In a later study with a sample of OCs with varied phenomenology, Shafran (1997) manipulated degree of responsibility by varying the presence/absence of the experimenter during a task. The manipulation was successful in increasing perceived responsibility for threat, urge to neutralize, discomfort, and estimated
262
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
probability of threat. However, responsibility for thoughts and control over the threat were not affected. It is important to note that in both the Lopatka and Rachman's (1995) and the Shafron's studies (1997) perceived responsibility and estimated probability of threat were similarly affected by the responsibility manipulation. Therefore, it is unclear whether responsibility or estimated probabilities was the crucial variable in determining the urge to check and its associated discomfort. The study described in this paper aims to explore further the relationship between OCD and inflated personal responsibility for harm in a treatment-seeking population of OCs with checking compulsions by investigating whether inflated responsibility is a general characteristic of OCD or is specific to OCD concerns. We hypothesized that compared to nonanxious controls (NACs) and anxious controls with generalized social phobia (GSPs), OCs would report: (1) greater urge to rectify situations involving potential risk, (2) more emotional distress upon leaving such situations unrectified, and (3) more personal responsibility for harm if the unrectified situation later resulted in a negative outcome. In order to examine these hypotheses and the degree of specificity of these characteristics, we developed the Obsessive ± Compulsive Responsibility Scale (OCRS), which includes 27 situations varying in degree of risk and OC relevance. 2. Method 2.1. Participants Participants were 15 OCs with primary checking compulsions (OCs), 15 GSPs, and 15 NACs.1 OCs and GSPs were individuals seeking treatment for their disorder at the Center for the Treatment and Study of Anxiety. All OCs and GSPs were diagnosed with their respective disorder according to DSM-IV criteria (American Psychiatric Association, 1994). OCs had a score of 16 or higher on the Yale-Brown Obsessive± Compulsive Scale (Y-BOCS; Goodman et al., 1989); M = 24.5, S.D. = 7.4. GSPs mean SPAI (Turner, Beidel, Dancu, & Stanley, 1989) score of M = 132.7, S.D. = 16.4. NACs were students (undergraduate, continuing education, and graduate levels) and hospital staff at a large, urban university. Individuals who volunteered to participate received monetary compensation. NACs did not meet any DSM-IV Axis I disorder. The groups did not differ on age or years of education, but OCs scored significantly higher than NACs and GSPs on self1
A preliminary version of this study included individuals with the diagnosis of OCD with varied phenomenology (n = 24) and a group of NACs (n = 20). In the present version, we included only OCs who had primary checking rituals (n = 15), as the OCRS includes OC-relevant situations that trigger an urge to check only. We then dropped the last five participants in NAC group to minimize the effect of unequal variances across groups (Miller, 1986, p. 57).
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
263
Table 1 Sample characteristics Demographics % male Age Education OCI Distress
OCs (N = 15)
GSPs (N = 15)
NACs (N = 15)
53 male 34.1 years (S.D. = 8.4) 15.7 years (S.D. = 2.7) 70.9 (S.D. = 34.8)
53 male 35.5 years (S.D. = 10.0) 16.1 years (S.D. = 1.6) 22.5 (S.D. = 22.9)
73 male 33.7 years (S.D. = 12.2) 15.9 years (S.D. = 3.0) 6.1 (S.D. = 6.7)
Note: Y-BOCS: Yale-Brown Obsessive ± Compulsive Scale; OCI Distress: Obsessive ± Compulsive Inventory Distress total score.
report measures of OCD severity; the latter two of which did not significantly differ (Table 1). 2.2. Materials The initial version of the OCRS included 36 situations, 12 in each of 3 categories: low risk (LR), OC relevant, and high risk (HR). Based on pilot data from OCs and NACs, three items from each category were removed because they yielded responses that were inconsistent with those for the remaining items of the corresponding scales. The final version was comprised of 27 situations, 9 LR, 9 OC, and 9 HR. Examples of LR situations are: ``You see some stuffing coming out of a prize stuffed animal at a carnival'' and ``You see a piece of string on the ground.'' Examples of OC situations are: ``You see some nails on a road'' and ``You see a sharp wire in a parking lot.'' Examples of HR situations are: ``You see a person sitting alone in a diner is choking'' and ``You see a person faint in a supermarket.'' The instructions were as follows: ``This form describes situations that you may encounter in real life. Imagine what you would feel (not what you would do) if you were in the situation. . .'' Subjects were asked to rate, on three separate 0± 7 scales, the degree to which they would feel the urge to rectify the situation, the degree of distress they would feel upon leaving the situation unrectified, and the degree of personal responsibility they would feel if the unrectified situation later resulted in harm. The reliability of the OCRS was determined by examining the internal consistency of the three subscales in OC-relevant situations. Reliability was excellent: a's were .89 for the urge scale, .93 for the distress scale, and .94 for the responsibility scale. 2.3. Measures 2.3.1. The Y-BOCS (Baer, Brown-Beasley, Sorce, & Henriques, 1993; Goodman et al., 1989) This is a comprehensive interview measure of obsessive ± compulsive symptom severity (Goodman et al., 1989), which yields a total score based on various
264
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
aspects of psychopathology (e.g., distress and interference) concerning an individual's obsessions and compulsions. The Y-BOCS has an internal consistency of .91 (Goodman et al., 1989). 2.3.2. The Obsessive± Compulsive Inventory (OCI; Foa, Kozak, Salkovskis, Coles, & Amir, 1998) This is a 42-item self-report measure of OCD, which includes seven subscales: washing, checking, doubting, obsessing, hoarding, ordering, and neutralizing. It yields a total score based on the distress caused by obsessions and compulsions. The distress total scale has excellent internal consistency for OCs (a=.92) and NACs (a=.95). It also discriminates reliably between individuals with and without OCD. 2.3.3. The Social Phobia and Anxiety Inventory (SPAI; Turner et al., 1989) This is a 45-item standard self-report measure. Each item is rated on a 1± 7 Likert scale according to how frequently it is experienced. Several studies have demonstrated that the SPAI has good psychometric properties (Turner et al., 1989). 2.4. Procedure The OCs were given the OCRs along with the other measures at their initial evaluation. The NACs and GSPs were asked to complete the OCRS, as well as other standard measures of psychopathology, either in the Center for the Treatment and Study of Anxiety or at home. 3. Results Means and standard deviations were calculated separately for the OCs, GSPs, and NACs for each of the three OCRS subscales and for each of the three situation types. Mixed analyses of variance (ANOVAs) were performed with one between subjects variable (groups: OCs, GSPs, and NACs) and one within subjects variable (situation type: low risk, OC relevant, and high risk). Significant interactions were followed by analyses of simple main effects to determine the source of the interaction. These analyses were performed for each subscale: urge, distress, and responsibility. 3.1. Urge to rectify the situation The means and standard deviations for the urge subscale are depicted in Fig. 1. The analysis of Urge revealed significant main effects of situation type [ F(2,84) = 233.68, P < .001] and of group [ F(2,42) = 5.03, P < .01] that were
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
265
Fig. 1. Ratings of urge to rectify the situation.
modified by an interaction of Group Situation type [ F(4,84) = 3.09, P < .02]; therefore, we performed simple main effects analyses. Simple main effects of situation type revealed that groups differed on urges to rectify OC situations [ F(2,44) = 4.02, P < .03] and low-risk situations [ F(2,44) = 8.96, P < .001], but not on urges to fix high-risk situations [ F(2,44) = 2.31, P =.11]. Follow-up t tests revealed that OCs reported more urges to fix OC situations [t(28) = 2.4, P < .02] and low-risk situations [t(28) = 3.47, P < .02] than did NACs. Similarly, OCs had significantly more urges in OC situations [t(28) = 2.35, P < .03] and low-risk situations [t(28) = 3.19, P < .01] than did GSPs. GSPs and NACs did not differ in their urges to rectify the situation in either OC [t(28) = 0.5, P =.96] or low-risk situations [t(28) = 0.70, P =.49]. Simple main effects of group revealed that OCs responded differentially to situation types [ F(2,28) = 53.09, P < .001]. Paired t tests indicated that the OCs had more urges in the high-risk situations than in the LR situations [t(14) = 10.28, P < .001] and OC situations [t(14) = 4.02, P < .001]. Furthermore, OCs had more urges in OC situations than in the LR situations [t(14) = 6.75, P < .001]. GSPs also responded differentially to situation types [ F(2,28) = 78.21, P < .001], reporting more urges in HR situations than in LR [t(14) = 10.67, P < .001] and OC situations [t(14) = 8.32, P < .001], and more urges in OC situations than in LR situations [t(14) = 5.69, P < .001].
266
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
NACs also responded differentially to situation types [ F(2,28) = 110.67, P < .001]. Paired t tests revealed that the NACs had more urges in HR situations than in LR situations [t(14) = 14.84, P < .001] and OC situations [t(14) = 7.62, P < .001]. Furthermore, NACs had more urges in OC than LR situations [t(14) = 7.33, P < .001]. 3.2. Distress upon leaving the situation unrectified The means and standard deviations for the distress subscale are depicted in Fig. 2. The analysis of distress revealed significant main effects of situation type [ F(2,84) = 274.55, P < .001] and group [ F(2,42) = 7.65, P < .001] that were modified by a significant interaction of Group Situation type [ F(4,84) = 3.07, P < .02]. In order to prove this interaction, we performed simple main effects analyses. Simple main effects of situation type revealed that groups did not differ on distress in high-risk situations [ F(2,44) = 1.45, P =.25]. However, groups responded differently to OC situations [ F(2,44) = 10.26, P < .001] and low-risk situation [ F(2,44) = 9.76, P < .001]. Follow-up t tests revealed that OCs were more distressed than NACs on OC situations [t(28) = 4.43, P < .001] and on the LR situations [t(28) = 3.76, P < .001]. Similarly, OCs scored higher than GSPs on
Fig. 2. Ratings of distress upon leaving the situation unrectified.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
267
OC situations [t(28) = 2.74, P < .01] and on the LR situations [t(28) = 3.20, P < .01]. GSPs and NACs did not differ in their rating of OC situations [t(28) = 1.54, P =.13] or LR situations [t(28) = 0.84, P =.41]. Simple main effects of group revealed that OCs responded differentially to situation types [ F(2,28) = 88.10, P < .001]. Paired t tests indicated that OCs rated the HR situations as more distressing than LR situations [t(14) = 14.60, P < .001] and OC situations [t(14) = 5.98, P < .001]. Furthermore, OCs rated OC situations as more distressing than LR situations [t(14) = 7.32, P < .001]. GSPs also responded differentially to situation types [ F(2,28) = 91.69, P < .001]. Paired t tests indicated that GSPs rated HR situations as more distressing than LR situations [t(14) = 11.47, P < .001] and OC situations [t(14) = 9.10, P < .001]. Furthermore, GSPs rated OC situations as more distressing than LR situations [t(14) = 4.70, P < .001]. The NACs also responded differentially to situation types [ F(2,28) = 99.47, P < .001]. Paired t tests revealed that NACs rated HR situations as more distressing than LR situations [t(14) = 10.69, P < .001] and OC situations [t(14) = 9.75, P < .001]. Furthermore, NACs rated OC situations as more distressing than LR situations [t(14) = 4.35, P < .001]. 3.3. Responsibility if the unrectified situation were to result in harm The means and standard deviations on the responsibility subscale are depicted in Fig. 3. The analysis of responsibility revealed significant main effects of situation type [ F(2,84) = 191.17, P < .001] and group [ F(2,42) = 8.42, P < .001]. Because there was a trend toward an interaction of Group Situation type [ F(4,84) = 2.23, P =.07], we performed simple main effects analyses. Simple main effects of situation type revealed that groups did not differ on responsibility in high-risk situations [ F(2, 44) = 2.11, P =.13]. However, groups responded differently to OC situations [ F(2,44) = 9.97, P < .001] and LR situations [ F(2,44) = 10.69, P < .001]. Follow-up t tests revealed that OCs scored higher than NACs on OC situations [t(28) = 4.46, P < .001] and LR situations [t(28) = 4.00, P < .001]. Similarly, OCs scored higher than GSPs on OC situations [t(28) = 2.29, P < .05] and LR situations [t(28) = 2.77, P < .01]. GSPs scored higher than NACs on OC situations [t(28) = 2.17, P < .05] and LR situations [t(28) = 2.01, P < .05]. Simple main effects of group revealed that OCs responded differentially to situation types [ F(2,28) = 42.61, P < .001]. Paired t tests indicated that the OCs had higher responsibility ratings in the HR situations than in the LR situations [t(14) = 10.53, P < .001] and OC situations [t(14) = 5.62, P < .001]. Furthermore, OCs had higher responsibility ratings in OC situations than in LR situations [t(14) = 8.25, P < .001]. GSPs also responded differentially to situation types [ F(2,28) = 78.21, P < .001]. Paired t tests indicated that the GSPs had higher responsibility ratings
268
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
Fig. 3. Ratings of responsibility if the unrectified situation were to result in harm.
in the HR situations than in the LR situations [t(14) = 10.08, P < .001] and OC situations [t(14) = 8.34, P < .001]. Furthermore, GSPs had higher responsibility ratings in OC situations than in LR situations [t(14) = 4.64, P < .001]. The NACs also responded differentially to situation types [ F(2,28) = 59.73, P < .001]. Paired t tests revealed that the NACs had higher responsibility ratings in the HR situations than in the LR situations [t(14) = 8.05, P < .001] and OC situations [t(14) = 7.74, P < .001]. Furthermore, NACs had higher responsibility ratings in OC situations than in LR situations [t(14) = 3.76, P < .01]. 3.4. Correlational analysis In order to examine the relationship between the OCRS subscales (i.e., urge, distress, and responsibility) and the severity of OCD symptoms, we correlated the total score of the OCI for OCs with their scores on each of the OCRS subscales on OC-relevant situations. The OCI frequency score was significantly correlated with all three OCRS subscales for the OCs: urge (r =.56, P < .03), distress (r =.58, P < .03), and responsibility (r =.65, P < .01). No significant correlations emerged for the NACs: urge (r =.17, P =.54), distress (r =.19, P =.49), and responsibility (r =.15, P =.59), nor for the GSPs:
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
269
urge (r =.05, P =.86), distress (r =.25, P =.36), and responsibility (r =.28, P =.31). To examine the relationship between severity of generalized social phobia and the OCRS subscales, we correlated the total score of the SPAI for GSPs with their scores on each of the OCRS subscales on OC-relevant situations. The SPAI severity score was not significantly correlated with any of the three OCRS subscales for the GSPs: urge (r = .19, P =.53), distress (r =.06, P =.84), and responsibility (r =.01, P =.98). 4. Discussion The results of the present study provide partial support for the hypothesis that an inflated perception of personal responsibility underlies the pathology of OCD (e.g., Rachman, 1993; Salkovskis, 1985). Consistent with our inflated responsibility hypothesis, OCs exhibited greater responsibility about low-risk (e.g., ``You see a piece of string on the ground'') and OC-relevant situations (e.g., ``You see a sharp wire in a parking lot'') than did NACs and GSPs. However, GSPs also expressed a greater sense of responsibility than did nonanxious individuals on OC but not on the low-risk situations. Interestingly, the groups did not differ in their ratings of responsibility for the high-risk situations. This suggests that the tendency for inflated responsibility varies with the content of the situation. The finding that OCs differed from NACs more on the low-risk, but not on the highrisk situations is convergent with Lopatka and Rachman's (1995) results that low, but not the HR condition affected urge to check and discomfort. The findings that socially anxious individuals also show inflated responsibility question the unique status of inflated responsibility in OCD. Perhaps an elevated sense of responsibility is common to individuals with anxiety disorders, with OCs being on the high end of the continuum. Moreover, our results suggest that OCs, as compared to NACs and GSPs, show not only inflated responsibility but also higher degree of urge to rectify low-risk and OC-relevant situations and distress upon leaving these situations unrectified. GSPs, on the other hand, did not differ from controls on either of these two measures. Thus, urges to rectify potentially harmful situations and distress upon not rectifying these situations seem to be more unique features of OCD than does inflated responsibility. The correlational analyses support this interpretation. As would be expected, in the OCD group but not in the NAC or GSP groups, the degree of perceived responsibility was highly correlated with OCD severity. However, in OCs, the degree of urge to rectify situations and the distress level upon failure to rectify these situations were also highly correlated with OCD severity. These results are consistent with Salkovskis et al.'s (1995) hypothesis that OCs lack the omission bias. While the present study demonstrates a strong empirical support for the hypothesis that inflated responsibility is an important factor in OCD, several
270
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
questions remain unanswered. First, all the situations included in the OCRS imply the potential harm to others, and thus the results can only be generalized to these types of situations. However, as discussed earlier, Salkovskis (1989) suggested that OCs are also characterized by inflated responsibility for harm to oneself. The present results do not examine this hypothesis. Second, all the situations in the current study imply the potential of a negative outcome. Do OCs have inflated feelings of responsibility for positive outcomes? We will address these issues in future studies.
Appendix Name/ID #___
OCRS This form describes situations that you may encounter in real life. Imagine what you would feel (not what you would do) if you were in the situation. Use the scale below to choose the number that best describes how you would FEEL in each situation and write the number on the line to the left of EACH question. Remember, we are interested in what you would feel, not what you would or could actually do. Not at all
Somewhat
0
1
2
Quite a bit 3
4
5
Very much so 6
7
EXAMPLE: You see smoke coming from a house nearby while you are standing near a fire alarm. 6 I would have an urge to pull the fire alarm although I might not actually do so. 2 I would feel distressed if I did not pull the fire alarm. 4 I would feel it was my fault if something bad happened because I did not pull the fire alarm. (1) You see a large trash can in the middle of the street as you drive past a school but stop. ___I would have an urge to remove the trash can from the middle of the street although I might not actually do so. ___I would feel distressed if I left the trash can in the middle of the street. ___I would feel it was my fault if something bad happened because I left the trash can in the middle of the street.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
271
(2) You see a child about to run into a busy street after his ball. ___I would have an urge to stop the child from running into the street although I might not actually do so. ___I would feel distressed if I left the child to run into the street. ___I would feel it was my fault if I left the child to run into the street and something bad happened. (3) You see that the window in a classroom is slightly open. ___I would have an urge to close the window although I might not actually do so. ___I would feel distressed if I left the window. ___I would feel it was my fault if I left the window and something bad happened. (4) You see a shopping cart with a sharp rusty metal edge. ___I would have an urge to remove the sharp rusty metal edge although I might not actually do so. ___I would feel distressed if I left the sharp rusty metal edge. ___I would feel it was my fault if I left the sharp rusty metal edge and something bad happened. (5) You see some stuffing coming out of a prize stuffed animal at a carnival. ___I would have an urge to fix the stuffed animal although I might not actually do so. ___I would feel distressed if I left the stuffed animal. ___I would feel it was my fault if I left the stuffing coming out of the stuffed animal and something bad happened. (6) You see an unattended toddler pick up scissors. ___I would have an urge to take the scissors from the toddler although I might not actually do so. ___I would feel distressed if I left the scissors with the toddler. ___I would feel it was my fault if I left the scissors with the toddler and something bad happened. (7) You see that an electrical wire to a desk light in a public library has a crack in it. ___I would have an urge to remove the cracked electrical wire although I might not actually do so. ___I would feel distressed if I left the cracked electrical wire. ___I would feel it was my fault if something bad happened because I left the cracked electrical wire. (8) You see a person sitting alone in a diner is choking. ___I would have an urge to try to dislodge the food although I might not actually do so. ___I would feel distressed if I left the person choking. ___I would feel it was my fault if something bad happened because I left the person choking.
272
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
(9) You see a paralyzed person's wheelchair is stuck in a grate. ___I would have an urge to help the paralyzed person get the wheelchair out of the grate although I might not actually do so. ___I would feel distressed if I left the paralyzed person's wheelchair stuck in the grate. ___I would feel it was my fault if something bad happened because I left the paralyzed person's wheelchair stuck in the grate. (10) You see a splinter of wood on a ceiling beam in a restaurant. ___I would have an urge to remove the splinter of wood although I might not actually do so. ___I would feel distressed if I left the splinter of wood. ___I would feel if was my fault if something bad happened because I left the splinter of wood. (11) You see a hubcap on a bike trail. ___I would have an urge to remove the hubcap from the bike trail although I might not actually do so. ___I would feel distressed if I left the hubcap on the bike trail. ___I would feel it was my fault if I left the hubcap on the bike trail and something bad happened. (12) You see a small rock on the sidewalk. ___I would have an urge to remove the rock although I might not actually do so. ___I would feel distressed if I left the rock. ___I would feel it was my fault if I left the rock and something bad happened. (13) You see a person who has just been in a car accident is unable to open the car door. ___I would have an urge to open the door for the person although I might not actually do so. ___I would feel distressed if I left the door unopened. ___I would feel it was my fault if I left the door unopened and something bad happened. (14) You see that a sewage drain is clogged up with leaves. ___I would have an urge to remove the leaves although I might not actually do so. ___I would feel distressed if I left the leaves. ___I would feel it was my fault if I left the leaves and something bad happened. (15) You see a person in a subway have an epileptic seizure. ___I would have an urge to help the person although I might not actually do so. ___I would feel distressed if I left the person having a seizure on the ground. ___I would feel it was my fault if something bad happened because I left the person having a seizure. (16) You see a leg on a chair in a library is loose. ___I would have an urge to put the chair in the corner although I might not actually do so.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
273
___I would feel distressed if I left the chair with a loose leg. ___I would feel it was my fault if something bad happened because I left the chair with the loose leg. (17) You see a box of stale crackers in an office cabinet. ___I would have an urge to throw away the crackers although I might not actually do so. ___I would feel distressed if I left the crackers. ___I would feel it was my fault if something bad happened because I left the crackers. (18) You see a pin on the floor of a dressing room in a department store. ___I would have an urge to remove the pin from the dressing room floor although I might not actually do so. ___I would feel distressed if I left the pin on the dressing room floor. ___I would feel it was my fault if left the pin on the dressing room floor and something bad happened. (19) You see a person faint while you are shopping in a supermarket. ___I would have an urge to help the person who fainted although I might not actually do so. ___I would feel distressed if I left the person who fainted in the supermarket. ___I would feel it was my fault if I left the person who fainted in the supermarket and something bad happened. (20) You see a piece of string on the ground. ___I would have an urge to remove the piece of string although I might not actually do so. ___I would feel distressed if I left the piece of string on the ground. ___I would feel it was my fault if something bad happened because I left the piece of string on the ground. (21) You see that a road sign on a highway is a little bent. ___I would have an urge to flatten out the road sign although I might not actually do so. ___I would feel distressed if I left the road sign. ___I would feel it was my fault if I left the road sign and something bad happened. (22) You see some nails on a road. ___I would have an urge to remove the nails from the road although I might not actually do so. ___I would feel distressed if I left the nails on the road. ___I would feel it was my fault if I left the nails on the road and something bad happened. (23) You see a sharp wire in a parking lot. ___I would have an urge to remove the sharp wire from the parking lot although I might not actually do so. ___I would feel distressed if I left the sharp wire in the parking lot.
274
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
___I would feel it was my fault if I left the sharp wire in the parking lot and something bad happened. (24) You see an elderly person next to you is struggling to get out of their seat on a bus. ___I would have an urge to help the elderly person stand up although I might not actually do so. ___I would feel distressed if I left the elderly person struggling. ___I would feel it was my fault if something bad happened because I left the elderly person struggling. (25) You see the muffler on the car next to you is loose. ___I would have an urge to fix the muffler although I might not actually do so. ___I would feel distressed if I left the muffler. ___I would feel it was my fault if I left the muffler and something bad happened. (26) You see a banana peel on the steps in a baseball stadium. ___I would have an urge to remove the banana peel from the steps although I might not actually do so. ___I would feel distressed if I left the banana peel on the steps. ___I would feel it was my fault if something bad happened because I left the banana peel on the steps. (27) You see the knob on a gas oven was left on in a church kitchen. ___I would have an urge to turn off the gas oven knob although I might not actually do so. ___I would feel distressed if I left the gas oven knob on. ___I would feel it was my fault if I left the gas oven knob on and something bad happened.
References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association. Baer, L., Brown-Beasley, M. W., Sorce, J., & Henriques, A. (1993). Computer-assisted administration of a structured interview for obsessive ± compulsive disorder. American Journal of Psychiatry, 150, 1737 ± 1738. Foa, E. B., Kozak, M. J., Salkovskis, P., Coles, M., & Amir, N. (1998). The validation of a new obsessive ± compulsive disorder scale: the Obsessive ± Compulsive Inventory. Psychological Assessment, 10 (3), 206 ± 214. Foa, E. B., Steketee, G., & Young, M. C. (1984). Agoraphobia: phenomenological aspects, associated characteristics, and theoretical considerations. Clinical Psychology Review, 4, 431 ± 457. Freeston, M. H., Ladouceur, R., Thibodeau, N., & Gagnon, F. (1992). Cognitive intrusions in a nonclinical population: II. Associations with depressive, anxious, and compulsive symptoms. Behaviour Research and Therapy, 30 (3), 263 ± 271. Goodman, W. K., Price, L. H., Rasmussen, S. A., Mazure, C., Fleischmann, R. L., Hill, C. L., Heninger, G. R., & Charney, D. S. (1989). The Yale-Brown Obsessive ± Compulsive Scale. Archives of General Psychiatry, 46, 1006 ± 1011.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
275
Hodgson, R. J., & Rachman, S. (1977). Obsessional ± compulsive complaints. Behaviour Research and Therapy, 15, 389 ± 395. Ladouceur, R., Rheaume, J., Freeston, M. H., Aublet, F., Jean, K., Lachance, S., Langlois, F., & De Pokomandy-Morin, K. (1995). Experimental manipulations of responsibility: an analogue test for models of obsessive ± compulsive disorder. Behaviour Research and Therapy, 33 (8), 937 ± 946. Lopatka, C., & Rachman, S. (1995). Perceived responsibility and compulsive checking: an experimental analysis. Behaviour Research and Therapy, 33 (6), 673 ± 684. Miller, R. (1986). Beyond ANOVA, basics of applied statistics. New York, NY: Wiley International. Rachman, S. (1993). Obsessions, responsibility and guilt. Behaviour Research and Therapy, 31 (2), 149 ± 154. Rachman, S., Thordarson, D. S., Shafran, R., & Woody, S. R. (1995). Perceived responsibility: structure and significance. Behaviour Research and Therapy, 33 (7), 779 ± 784. Salkovskis, P. M. (1985). Obsessional ± compulsive problems: a cognitive-behavioural analysis. Behaviour Research and Therapy, 23 (5), 571 ± 583. Salkovskis, P. M. (1989). Cognitive-behavioural factors and the persistence of intrusive thoughts in obsessional problems. Behaviour Research and Therapy, 27 (6), 677 ± 682. Salkovskis, P. M., Richards, H. C., & Forrester, E. (1995). The relationship between obsessional problems and intrusive thoughts. Behavioural and Cognitive Psychotherapy, 23, 281 ± 299. Shafran, R. (1997). The manipulation of responsibility in obsessive ± compulsive disorder. British Journal of Clinical Psychology, 36, 397 ± 407. Turner, S. M., Beidel, D. C., Dancu, C. V., & Stanley, M. A. (1989). An empirically derived inventory to measure social fears and anxiety: the Social Phobia and Anxiety Inventory. Psychological Assessment, 1, 35 ± 40.
Inflated perception of responsibility for harm in obsessive±compulsive disorder Edna B. Foaa,*, Nader Amirb, Karen V.A. Bogertc, Chris Molnarc, Amy Przeworskic a
Department of Psychiatry, University of Pennsylvania, Philadelphia, PA b Department of Psychology, University of Georgia, Athens, GA, USA c Department of Psychiatry, Allegheny University of the Health Sciences, USA Received 9 April 1999; received in revised form 23 May 1999; accepted 21 April 2000
Abstract The present study examined the proposition that inflated responsibility is implicated in obsessive ± compulsive disorder (OCD). Compared to non-anxious control participants (NACs), and an anxious control group with generalized social phobia (GSPs), we predicted that individuals with OCD (OCs) would exhibit a greater urge to rectify situations involving potential risk, would report more distress upon leaving such situations unrectified, and would feel more personal responsibility if the unrectified situations resulted in harm. Fifteen OCs, 15 NACs, and 15 GSPs completed the Obsessive ± Compulsive Responsibility Scale (OCRS), which included low-risk, OC-relevant, and high-risk situations. Compared to NACs and GSPs, OCs reported more urges, distress, and responsibility in low-risk and OC-relevant situations; no group differences were detected on high-risk situations. GSPs and NACs differed only in their responsibility in OCrelevant situations, with GSPs reporting higher responsibility than controls. Our results suggest that compared to OCs, NACs, and GSPs can better differentiate between situations that merit concern and ones that do not. D 2001 Elsevier Science Inc. All rights reserved. Keywords: Obsessive-compulsive disorder; Cognitive distortions; Inflated responsibility
* Corresponding author. Department of Psychiatry, University of Pennsylvania, 3535 Market Street, Philadelphia, PA 19104, USA. Tel no.: 215-746-3327; fax: 215-746-3311. Email address: [email protected] (E.B. Foa). 0887-6185/01/$ ± see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 0 8 8 7 - 6 1 8 5 ( 0 1 ) 0 0 0 6 2 - 7
260
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
1. Introduction Clinical observations suggest that individuals with obsessive ± compulsive disorder (OCs) suffer from excessive personal responsibility for harm to others. Foa, Steketee, and Young (1984) noted that individuals with agoraphobia generally fear harm that might come to themselves, whereas OCs are often preoccupied with harm that may come to others. Such clinical observations led Salkovskis (1985, 1989) to propose that an exaggerated sense of responsibility for negative thoughts (e.g., ``I might stab my child'') and anticipated negative events (``I will run over a pedestrian if I'm not careful'') underlie the psychopathology of OCD. Similarly, Rachman (1993) suggested that many patients with obsessive ± compulsive disorder (OCD), especially those with checking compulsions, experience urges to ritualize only in those circumstances in which they assume personal responsibility for safety. Specifically, Salkovskis (1985) argued that OCs' distressing thoughts are typically related to feelings of responsibility for harm to others. In order to reconcile the discrepancy between his proposition and the observation that some OCs are not concerned with others' safety, Salkovskis extended the customary use of the concept of responsibility to include feelings of liability for one's own harm. In a later exposition of this model, Salkovskis, Richards, and Forrester (1995) further emphasize the central role of the responsibility-neutralizing link in the etiology of OCD. They suggested that obsessions of OCs are not only characterized by a heightened sense of responsibility, but are also distinguished by the belief that not acting in a potentially harmful situation is akin to directly causing the harm. To explore the relationship between responsibility and neutralizing, Freeston, Ladouceur, Thibodeau, and Gagnon (1992) administered the Cognitive Intrusions Questionnaire to college students. Factor analysis of responses on the questionnaire yielded five factors, only one of which was significantly related to questions of compulsive activity. This factor, the valence factor, represented the person's evaluation of the intrusion as a function of responsibility and guilt. The highest loading item on this factor was ``To what extent would you feel responsible if the thought content were to happen?,'' therefore, the authors interpreted the results as demonstrating a link between responsibility and compulsive behavior. Rachman, Thordarson, Shafran, and Woody (1995) examined the relationship between responsibility and OCD symptoms in a sample of college students. To this end, they devised a questionnaire that measured four aspects of responsibility: responsibility for harm, responsibility in social contexts, a positive outlook towards responsibility, and thought±action fusion. Only the thought± action fusion aspect correlated with OC symptom severity as measured by the Maudsley Obsessive ± Compulsive Inventory (MOCI; Hodgson & Rachman, 1977). Consequently, the authors concluded that inflated responsibility is not a general characteristic of OCD, but, rather, may be specific to personally relevant situations.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
261
To explore the role of responsibility in the performance of compulsions, Ladouceur et al. (1995) carried out two experiments in which they manipulated degree of responsibility in college students. In the first experiment, participants were asked to determine whether they had heard a sound before. They were allowed to listen to the sound as many times as they wanted (i.e., check) before responding. Although the manipulation was successful in producing more feelings of responsibility in the high-responsibility (HR) group, participants in this group did not exhibit more checking behavior than participants in the lowresponsibility (LR) group. The authors attributed the negative results to the absence of a relationship between the responsibility manipulation and the performed task. Therefore, in a second experiment, participants in the HR group were instructed to sort different kinds of pills seemingly intended for public consumption. Participants in the LR group were told that the pill-sorting task was designed to inform the experimenter about color perception. Participants in the HR condition exhibited more preoccupation with making errors and were more anxious during the task compared to those in the LR condition. However, no group differences emerged on self-report of doubting, urges to check, and time required to complete the task. At best, the studies described above demonstrate a limited relationship between responsibility and OCD-like symptoms. Moreover, it is important to note that these studies were conducted with nonpatient students, and therefore the significance of the findings for our understanding of OCD is unclear. To date, only two studies have utilized individuals diagnosed with OCD to study the role of responsibility (Lopatka & Rachman, 1995; Shafran, 1997). Both studies examined the hypotheses that inflating responsibility would increase urge to neutralize, the degree of discomfort experienced, and the estimated probabilities of bad outcomes. Deflating responsibility was hypothesized to decrease urge, discomfort, and estimated probabilities. In the study of Lopatka and Rachman (1995), participants were instructed to perform a task at home that usually evoked an urge to check under ``highresponsibility'' instructions (HRI), ``low-responsibility'' instructions (LRI), and ``controlled'' instructions (CI). Responsibility was manipulated using contracts in which the experimenter assumed complete responsibility for potential negative events following the task (LRI) or the participant assumed complete responsibility for such events (HRI). Although the HRI condition produced a higher degree of responsibility compared with the CI condition, it did not produce stronger urged to check, higher discomfort, or higher estimated probability of harm. In contrast, the LRI reduced the subjects' urges to check, discomfort, and estimated probability of harm. Thus, the hypothesis that responsibility plays a role in OCD was partially supported. In a later study with a sample of OCs with varied phenomenology, Shafran (1997) manipulated degree of responsibility by varying the presence/absence of the experimenter during a task. The manipulation was successful in increasing perceived responsibility for threat, urge to neutralize, discomfort, and estimated
262
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
probability of threat. However, responsibility for thoughts and control over the threat were not affected. It is important to note that in both the Lopatka and Rachman's (1995) and the Shafron's studies (1997) perceived responsibility and estimated probability of threat were similarly affected by the responsibility manipulation. Therefore, it is unclear whether responsibility or estimated probabilities was the crucial variable in determining the urge to check and its associated discomfort. The study described in this paper aims to explore further the relationship between OCD and inflated personal responsibility for harm in a treatment-seeking population of OCs with checking compulsions by investigating whether inflated responsibility is a general characteristic of OCD or is specific to OCD concerns. We hypothesized that compared to nonanxious controls (NACs) and anxious controls with generalized social phobia (GSPs), OCs would report: (1) greater urge to rectify situations involving potential risk, (2) more emotional distress upon leaving such situations unrectified, and (3) more personal responsibility for harm if the unrectified situation later resulted in a negative outcome. In order to examine these hypotheses and the degree of specificity of these characteristics, we developed the Obsessive ± Compulsive Responsibility Scale (OCRS), which includes 27 situations varying in degree of risk and OC relevance. 2. Method 2.1. Participants Participants were 15 OCs with primary checking compulsions (OCs), 15 GSPs, and 15 NACs.1 OCs and GSPs were individuals seeking treatment for their disorder at the Center for the Treatment and Study of Anxiety. All OCs and GSPs were diagnosed with their respective disorder according to DSM-IV criteria (American Psychiatric Association, 1994). OCs had a score of 16 or higher on the Yale-Brown Obsessive± Compulsive Scale (Y-BOCS; Goodman et al., 1989); M = 24.5, S.D. = 7.4. GSPs mean SPAI (Turner, Beidel, Dancu, & Stanley, 1989) score of M = 132.7, S.D. = 16.4. NACs were students (undergraduate, continuing education, and graduate levels) and hospital staff at a large, urban university. Individuals who volunteered to participate received monetary compensation. NACs did not meet any DSM-IV Axis I disorder. The groups did not differ on age or years of education, but OCs scored significantly higher than NACs and GSPs on self1
A preliminary version of this study included individuals with the diagnosis of OCD with varied phenomenology (n = 24) and a group of NACs (n = 20). In the present version, we included only OCs who had primary checking rituals (n = 15), as the OCRS includes OC-relevant situations that trigger an urge to check only. We then dropped the last five participants in NAC group to minimize the effect of unequal variances across groups (Miller, 1986, p. 57).
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
263
Table 1 Sample characteristics Demographics % male Age Education OCI Distress
OCs (N = 15)
GSPs (N = 15)
NACs (N = 15)
53 male 34.1 years (S.D. = 8.4) 15.7 years (S.D. = 2.7) 70.9 (S.D. = 34.8)
53 male 35.5 years (S.D. = 10.0) 16.1 years (S.D. = 1.6) 22.5 (S.D. = 22.9)
73 male 33.7 years (S.D. = 12.2) 15.9 years (S.D. = 3.0) 6.1 (S.D. = 6.7)
Note: Y-BOCS: Yale-Brown Obsessive ± Compulsive Scale; OCI Distress: Obsessive ± Compulsive Inventory Distress total score.
report measures of OCD severity; the latter two of which did not significantly differ (Table 1). 2.2. Materials The initial version of the OCRS included 36 situations, 12 in each of 3 categories: low risk (LR), OC relevant, and high risk (HR). Based on pilot data from OCs and NACs, three items from each category were removed because they yielded responses that were inconsistent with those for the remaining items of the corresponding scales. The final version was comprised of 27 situations, 9 LR, 9 OC, and 9 HR. Examples of LR situations are: ``You see some stuffing coming out of a prize stuffed animal at a carnival'' and ``You see a piece of string on the ground.'' Examples of OC situations are: ``You see some nails on a road'' and ``You see a sharp wire in a parking lot.'' Examples of HR situations are: ``You see a person sitting alone in a diner is choking'' and ``You see a person faint in a supermarket.'' The instructions were as follows: ``This form describes situations that you may encounter in real life. Imagine what you would feel (not what you would do) if you were in the situation. . .'' Subjects were asked to rate, on three separate 0± 7 scales, the degree to which they would feel the urge to rectify the situation, the degree of distress they would feel upon leaving the situation unrectified, and the degree of personal responsibility they would feel if the unrectified situation later resulted in harm. The reliability of the OCRS was determined by examining the internal consistency of the three subscales in OC-relevant situations. Reliability was excellent: a's were .89 for the urge scale, .93 for the distress scale, and .94 for the responsibility scale. 2.3. Measures 2.3.1. The Y-BOCS (Baer, Brown-Beasley, Sorce, & Henriques, 1993; Goodman et al., 1989) This is a comprehensive interview measure of obsessive ± compulsive symptom severity (Goodman et al., 1989), which yields a total score based on various
264
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
aspects of psychopathology (e.g., distress and interference) concerning an individual's obsessions and compulsions. The Y-BOCS has an internal consistency of .91 (Goodman et al., 1989). 2.3.2. The Obsessive± Compulsive Inventory (OCI; Foa, Kozak, Salkovskis, Coles, & Amir, 1998) This is a 42-item self-report measure of OCD, which includes seven subscales: washing, checking, doubting, obsessing, hoarding, ordering, and neutralizing. It yields a total score based on the distress caused by obsessions and compulsions. The distress total scale has excellent internal consistency for OCs (a=.92) and NACs (a=.95). It also discriminates reliably between individuals with and without OCD. 2.3.3. The Social Phobia and Anxiety Inventory (SPAI; Turner et al., 1989) This is a 45-item standard self-report measure. Each item is rated on a 1± 7 Likert scale according to how frequently it is experienced. Several studies have demonstrated that the SPAI has good psychometric properties (Turner et al., 1989). 2.4. Procedure The OCs were given the OCRs along with the other measures at their initial evaluation. The NACs and GSPs were asked to complete the OCRS, as well as other standard measures of psychopathology, either in the Center for the Treatment and Study of Anxiety or at home. 3. Results Means and standard deviations were calculated separately for the OCs, GSPs, and NACs for each of the three OCRS subscales and for each of the three situation types. Mixed analyses of variance (ANOVAs) were performed with one between subjects variable (groups: OCs, GSPs, and NACs) and one within subjects variable (situation type: low risk, OC relevant, and high risk). Significant interactions were followed by analyses of simple main effects to determine the source of the interaction. These analyses were performed for each subscale: urge, distress, and responsibility. 3.1. Urge to rectify the situation The means and standard deviations for the urge subscale are depicted in Fig. 1. The analysis of Urge revealed significant main effects of situation type [ F(2,84) = 233.68, P < .001] and of group [ F(2,42) = 5.03, P < .01] that were
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
265
Fig. 1. Ratings of urge to rectify the situation.
modified by an interaction of Group Situation type [ F(4,84) = 3.09, P < .02]; therefore, we performed simple main effects analyses. Simple main effects of situation type revealed that groups differed on urges to rectify OC situations [ F(2,44) = 4.02, P < .03] and low-risk situations [ F(2,44) = 8.96, P < .001], but not on urges to fix high-risk situations [ F(2,44) = 2.31, P =.11]. Follow-up t tests revealed that OCs reported more urges to fix OC situations [t(28) = 2.4, P < .02] and low-risk situations [t(28) = 3.47, P < .02] than did NACs. Similarly, OCs had significantly more urges in OC situations [t(28) = 2.35, P < .03] and low-risk situations [t(28) = 3.19, P < .01] than did GSPs. GSPs and NACs did not differ in their urges to rectify the situation in either OC [t(28) = 0.5, P =.96] or low-risk situations [t(28) = 0.70, P =.49]. Simple main effects of group revealed that OCs responded differentially to situation types [ F(2,28) = 53.09, P < .001]. Paired t tests indicated that the OCs had more urges in the high-risk situations than in the LR situations [t(14) = 10.28, P < .001] and OC situations [t(14) = 4.02, P < .001]. Furthermore, OCs had more urges in OC situations than in the LR situations [t(14) = 6.75, P < .001]. GSPs also responded differentially to situation types [ F(2,28) = 78.21, P < .001], reporting more urges in HR situations than in LR [t(14) = 10.67, P < .001] and OC situations [t(14) = 8.32, P < .001], and more urges in OC situations than in LR situations [t(14) = 5.69, P < .001].
266
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
NACs also responded differentially to situation types [ F(2,28) = 110.67, P < .001]. Paired t tests revealed that the NACs had more urges in HR situations than in LR situations [t(14) = 14.84, P < .001] and OC situations [t(14) = 7.62, P < .001]. Furthermore, NACs had more urges in OC than LR situations [t(14) = 7.33, P < .001]. 3.2. Distress upon leaving the situation unrectified The means and standard deviations for the distress subscale are depicted in Fig. 2. The analysis of distress revealed significant main effects of situation type [ F(2,84) = 274.55, P < .001] and group [ F(2,42) = 7.65, P < .001] that were modified by a significant interaction of Group Situation type [ F(4,84) = 3.07, P < .02]. In order to prove this interaction, we performed simple main effects analyses. Simple main effects of situation type revealed that groups did not differ on distress in high-risk situations [ F(2,44) = 1.45, P =.25]. However, groups responded differently to OC situations [ F(2,44) = 10.26, P < .001] and low-risk situation [ F(2,44) = 9.76, P < .001]. Follow-up t tests revealed that OCs were more distressed than NACs on OC situations [t(28) = 4.43, P < .001] and on the LR situations [t(28) = 3.76, P < .001]. Similarly, OCs scored higher than GSPs on
Fig. 2. Ratings of distress upon leaving the situation unrectified.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
267
OC situations [t(28) = 2.74, P < .01] and on the LR situations [t(28) = 3.20, P < .01]. GSPs and NACs did not differ in their rating of OC situations [t(28) = 1.54, P =.13] or LR situations [t(28) = 0.84, P =.41]. Simple main effects of group revealed that OCs responded differentially to situation types [ F(2,28) = 88.10, P < .001]. Paired t tests indicated that OCs rated the HR situations as more distressing than LR situations [t(14) = 14.60, P < .001] and OC situations [t(14) = 5.98, P < .001]. Furthermore, OCs rated OC situations as more distressing than LR situations [t(14) = 7.32, P < .001]. GSPs also responded differentially to situation types [ F(2,28) = 91.69, P < .001]. Paired t tests indicated that GSPs rated HR situations as more distressing than LR situations [t(14) = 11.47, P < .001] and OC situations [t(14) = 9.10, P < .001]. Furthermore, GSPs rated OC situations as more distressing than LR situations [t(14) = 4.70, P < .001]. The NACs also responded differentially to situation types [ F(2,28) = 99.47, P < .001]. Paired t tests revealed that NACs rated HR situations as more distressing than LR situations [t(14) = 10.69, P < .001] and OC situations [t(14) = 9.75, P < .001]. Furthermore, NACs rated OC situations as more distressing than LR situations [t(14) = 4.35, P < .001]. 3.3. Responsibility if the unrectified situation were to result in harm The means and standard deviations on the responsibility subscale are depicted in Fig. 3. The analysis of responsibility revealed significant main effects of situation type [ F(2,84) = 191.17, P < .001] and group [ F(2,42) = 8.42, P < .001]. Because there was a trend toward an interaction of Group Situation type [ F(4,84) = 2.23, P =.07], we performed simple main effects analyses. Simple main effects of situation type revealed that groups did not differ on responsibility in high-risk situations [ F(2, 44) = 2.11, P =.13]. However, groups responded differently to OC situations [ F(2,44) = 9.97, P < .001] and LR situations [ F(2,44) = 10.69, P < .001]. Follow-up t tests revealed that OCs scored higher than NACs on OC situations [t(28) = 4.46, P < .001] and LR situations [t(28) = 4.00, P < .001]. Similarly, OCs scored higher than GSPs on OC situations [t(28) = 2.29, P < .05] and LR situations [t(28) = 2.77, P < .01]. GSPs scored higher than NACs on OC situations [t(28) = 2.17, P < .05] and LR situations [t(28) = 2.01, P < .05]. Simple main effects of group revealed that OCs responded differentially to situation types [ F(2,28) = 42.61, P < .001]. Paired t tests indicated that the OCs had higher responsibility ratings in the HR situations than in the LR situations [t(14) = 10.53, P < .001] and OC situations [t(14) = 5.62, P < .001]. Furthermore, OCs had higher responsibility ratings in OC situations than in LR situations [t(14) = 8.25, P < .001]. GSPs also responded differentially to situation types [ F(2,28) = 78.21, P < .001]. Paired t tests indicated that the GSPs had higher responsibility ratings
268
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
Fig. 3. Ratings of responsibility if the unrectified situation were to result in harm.
in the HR situations than in the LR situations [t(14) = 10.08, P < .001] and OC situations [t(14) = 8.34, P < .001]. Furthermore, GSPs had higher responsibility ratings in OC situations than in LR situations [t(14) = 4.64, P < .001]. The NACs also responded differentially to situation types [ F(2,28) = 59.73, P < .001]. Paired t tests revealed that the NACs had higher responsibility ratings in the HR situations than in the LR situations [t(14) = 8.05, P < .001] and OC situations [t(14) = 7.74, P < .001]. Furthermore, NACs had higher responsibility ratings in OC situations than in LR situations [t(14) = 3.76, P < .01]. 3.4. Correlational analysis In order to examine the relationship between the OCRS subscales (i.e., urge, distress, and responsibility) and the severity of OCD symptoms, we correlated the total score of the OCI for OCs with their scores on each of the OCRS subscales on OC-relevant situations. The OCI frequency score was significantly correlated with all three OCRS subscales for the OCs: urge (r =.56, P < .03), distress (r =.58, P < .03), and responsibility (r =.65, P < .01). No significant correlations emerged for the NACs: urge (r =.17, P =.54), distress (r =.19, P =.49), and responsibility (r =.15, P =.59), nor for the GSPs:
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
269
urge (r =.05, P =.86), distress (r =.25, P =.36), and responsibility (r =.28, P =.31). To examine the relationship between severity of generalized social phobia and the OCRS subscales, we correlated the total score of the SPAI for GSPs with their scores on each of the OCRS subscales on OC-relevant situations. The SPAI severity score was not significantly correlated with any of the three OCRS subscales for the GSPs: urge (r = .19, P =.53), distress (r =.06, P =.84), and responsibility (r =.01, P =.98). 4. Discussion The results of the present study provide partial support for the hypothesis that an inflated perception of personal responsibility underlies the pathology of OCD (e.g., Rachman, 1993; Salkovskis, 1985). Consistent with our inflated responsibility hypothesis, OCs exhibited greater responsibility about low-risk (e.g., ``You see a piece of string on the ground'') and OC-relevant situations (e.g., ``You see a sharp wire in a parking lot'') than did NACs and GSPs. However, GSPs also expressed a greater sense of responsibility than did nonanxious individuals on OC but not on the low-risk situations. Interestingly, the groups did not differ in their ratings of responsibility for the high-risk situations. This suggests that the tendency for inflated responsibility varies with the content of the situation. The finding that OCs differed from NACs more on the low-risk, but not on the highrisk situations is convergent with Lopatka and Rachman's (1995) results that low, but not the HR condition affected urge to check and discomfort. The findings that socially anxious individuals also show inflated responsibility question the unique status of inflated responsibility in OCD. Perhaps an elevated sense of responsibility is common to individuals with anxiety disorders, with OCs being on the high end of the continuum. Moreover, our results suggest that OCs, as compared to NACs and GSPs, show not only inflated responsibility but also higher degree of urge to rectify low-risk and OC-relevant situations and distress upon leaving these situations unrectified. GSPs, on the other hand, did not differ from controls on either of these two measures. Thus, urges to rectify potentially harmful situations and distress upon not rectifying these situations seem to be more unique features of OCD than does inflated responsibility. The correlational analyses support this interpretation. As would be expected, in the OCD group but not in the NAC or GSP groups, the degree of perceived responsibility was highly correlated with OCD severity. However, in OCs, the degree of urge to rectify situations and the distress level upon failure to rectify these situations were also highly correlated with OCD severity. These results are consistent with Salkovskis et al.'s (1995) hypothesis that OCs lack the omission bias. While the present study demonstrates a strong empirical support for the hypothesis that inflated responsibility is an important factor in OCD, several
270
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
questions remain unanswered. First, all the situations included in the OCRS imply the potential harm to others, and thus the results can only be generalized to these types of situations. However, as discussed earlier, Salkovskis (1989) suggested that OCs are also characterized by inflated responsibility for harm to oneself. The present results do not examine this hypothesis. Second, all the situations in the current study imply the potential of a negative outcome. Do OCs have inflated feelings of responsibility for positive outcomes? We will address these issues in future studies.
Appendix Name/ID #___
OCRS This form describes situations that you may encounter in real life. Imagine what you would feel (not what you would do) if you were in the situation. Use the scale below to choose the number that best describes how you would FEEL in each situation and write the number on the line to the left of EACH question. Remember, we are interested in what you would feel, not what you would or could actually do. Not at all
Somewhat
0
1
2
Quite a bit 3
4
5
Very much so 6
7
EXAMPLE: You see smoke coming from a house nearby while you are standing near a fire alarm. 6 I would have an urge to pull the fire alarm although I might not actually do so. 2 I would feel distressed if I did not pull the fire alarm. 4 I would feel it was my fault if something bad happened because I did not pull the fire alarm. (1) You see a large trash can in the middle of the street as you drive past a school but stop. ___I would have an urge to remove the trash can from the middle of the street although I might not actually do so. ___I would feel distressed if I left the trash can in the middle of the street. ___I would feel it was my fault if something bad happened because I left the trash can in the middle of the street.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
271
(2) You see a child about to run into a busy street after his ball. ___I would have an urge to stop the child from running into the street although I might not actually do so. ___I would feel distressed if I left the child to run into the street. ___I would feel it was my fault if I left the child to run into the street and something bad happened. (3) You see that the window in a classroom is slightly open. ___I would have an urge to close the window although I might not actually do so. ___I would feel distressed if I left the window. ___I would feel it was my fault if I left the window and something bad happened. (4) You see a shopping cart with a sharp rusty metal edge. ___I would have an urge to remove the sharp rusty metal edge although I might not actually do so. ___I would feel distressed if I left the sharp rusty metal edge. ___I would feel it was my fault if I left the sharp rusty metal edge and something bad happened. (5) You see some stuffing coming out of a prize stuffed animal at a carnival. ___I would have an urge to fix the stuffed animal although I might not actually do so. ___I would feel distressed if I left the stuffed animal. ___I would feel it was my fault if I left the stuffing coming out of the stuffed animal and something bad happened. (6) You see an unattended toddler pick up scissors. ___I would have an urge to take the scissors from the toddler although I might not actually do so. ___I would feel distressed if I left the scissors with the toddler. ___I would feel it was my fault if I left the scissors with the toddler and something bad happened. (7) You see that an electrical wire to a desk light in a public library has a crack in it. ___I would have an urge to remove the cracked electrical wire although I might not actually do so. ___I would feel distressed if I left the cracked electrical wire. ___I would feel it was my fault if something bad happened because I left the cracked electrical wire. (8) You see a person sitting alone in a diner is choking. ___I would have an urge to try to dislodge the food although I might not actually do so. ___I would feel distressed if I left the person choking. ___I would feel it was my fault if something bad happened because I left the person choking.
272
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
(9) You see a paralyzed person's wheelchair is stuck in a grate. ___I would have an urge to help the paralyzed person get the wheelchair out of the grate although I might not actually do so. ___I would feel distressed if I left the paralyzed person's wheelchair stuck in the grate. ___I would feel it was my fault if something bad happened because I left the paralyzed person's wheelchair stuck in the grate. (10) You see a splinter of wood on a ceiling beam in a restaurant. ___I would have an urge to remove the splinter of wood although I might not actually do so. ___I would feel distressed if I left the splinter of wood. ___I would feel if was my fault if something bad happened because I left the splinter of wood. (11) You see a hubcap on a bike trail. ___I would have an urge to remove the hubcap from the bike trail although I might not actually do so. ___I would feel distressed if I left the hubcap on the bike trail. ___I would feel it was my fault if I left the hubcap on the bike trail and something bad happened. (12) You see a small rock on the sidewalk. ___I would have an urge to remove the rock although I might not actually do so. ___I would feel distressed if I left the rock. ___I would feel it was my fault if I left the rock and something bad happened. (13) You see a person who has just been in a car accident is unable to open the car door. ___I would have an urge to open the door for the person although I might not actually do so. ___I would feel distressed if I left the door unopened. ___I would feel it was my fault if I left the door unopened and something bad happened. (14) You see that a sewage drain is clogged up with leaves. ___I would have an urge to remove the leaves although I might not actually do so. ___I would feel distressed if I left the leaves. ___I would feel it was my fault if I left the leaves and something bad happened. (15) You see a person in a subway have an epileptic seizure. ___I would have an urge to help the person although I might not actually do so. ___I would feel distressed if I left the person having a seizure on the ground. ___I would feel it was my fault if something bad happened because I left the person having a seizure. (16) You see a leg on a chair in a library is loose. ___I would have an urge to put the chair in the corner although I might not actually do so.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
273
___I would feel distressed if I left the chair with a loose leg. ___I would feel it was my fault if something bad happened because I left the chair with the loose leg. (17) You see a box of stale crackers in an office cabinet. ___I would have an urge to throw away the crackers although I might not actually do so. ___I would feel distressed if I left the crackers. ___I would feel it was my fault if something bad happened because I left the crackers. (18) You see a pin on the floor of a dressing room in a department store. ___I would have an urge to remove the pin from the dressing room floor although I might not actually do so. ___I would feel distressed if I left the pin on the dressing room floor. ___I would feel it was my fault if left the pin on the dressing room floor and something bad happened. (19) You see a person faint while you are shopping in a supermarket. ___I would have an urge to help the person who fainted although I might not actually do so. ___I would feel distressed if I left the person who fainted in the supermarket. ___I would feel it was my fault if I left the person who fainted in the supermarket and something bad happened. (20) You see a piece of string on the ground. ___I would have an urge to remove the piece of string although I might not actually do so. ___I would feel distressed if I left the piece of string on the ground. ___I would feel it was my fault if something bad happened because I left the piece of string on the ground. (21) You see that a road sign on a highway is a little bent. ___I would have an urge to flatten out the road sign although I might not actually do so. ___I would feel distressed if I left the road sign. ___I would feel it was my fault if I left the road sign and something bad happened. (22) You see some nails on a road. ___I would have an urge to remove the nails from the road although I might not actually do so. ___I would feel distressed if I left the nails on the road. ___I would feel it was my fault if I left the nails on the road and something bad happened. (23) You see a sharp wire in a parking lot. ___I would have an urge to remove the sharp wire from the parking lot although I might not actually do so. ___I would feel distressed if I left the sharp wire in the parking lot.
274
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
___I would feel it was my fault if I left the sharp wire in the parking lot and something bad happened. (24) You see an elderly person next to you is struggling to get out of their seat on a bus. ___I would have an urge to help the elderly person stand up although I might not actually do so. ___I would feel distressed if I left the elderly person struggling. ___I would feel it was my fault if something bad happened because I left the elderly person struggling. (25) You see the muffler on the car next to you is loose. ___I would have an urge to fix the muffler although I might not actually do so. ___I would feel distressed if I left the muffler. ___I would feel it was my fault if I left the muffler and something bad happened. (26) You see a banana peel on the steps in a baseball stadium. ___I would have an urge to remove the banana peel from the steps although I might not actually do so. ___I would feel distressed if I left the banana peel on the steps. ___I would feel it was my fault if something bad happened because I left the banana peel on the steps. (27) You see the knob on a gas oven was left on in a church kitchen. ___I would have an urge to turn off the gas oven knob although I might not actually do so. ___I would feel distressed if I left the gas oven knob on. ___I would feel it was my fault if I left the gas oven knob on and something bad happened.
References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association. Baer, L., Brown-Beasley, M. W., Sorce, J., & Henriques, A. (1993). Computer-assisted administration of a structured interview for obsessive ± compulsive disorder. American Journal of Psychiatry, 150, 1737 ± 1738. Foa, E. B., Kozak, M. J., Salkovskis, P., Coles, M., & Amir, N. (1998). The validation of a new obsessive ± compulsive disorder scale: the Obsessive ± Compulsive Inventory. Psychological Assessment, 10 (3), 206 ± 214. Foa, E. B., Steketee, G., & Young, M. C. (1984). Agoraphobia: phenomenological aspects, associated characteristics, and theoretical considerations. Clinical Psychology Review, 4, 431 ± 457. Freeston, M. H., Ladouceur, R., Thibodeau, N., & Gagnon, F. (1992). Cognitive intrusions in a nonclinical population: II. Associations with depressive, anxious, and compulsive symptoms. Behaviour Research and Therapy, 30 (3), 263 ± 271. Goodman, W. K., Price, L. H., Rasmussen, S. A., Mazure, C., Fleischmann, R. L., Hill, C. L., Heninger, G. R., & Charney, D. S. (1989). The Yale-Brown Obsessive ± Compulsive Scale. Archives of General Psychiatry, 46, 1006 ± 1011.
E.B. Foa et al. / Anxiety Disorders 15 (2001) 259±275
275
Hodgson, R. J., & Rachman, S. (1977). Obsessional ± compulsive complaints. Behaviour Research and Therapy, 15, 389 ± 395. Ladouceur, R., Rheaume, J., Freeston, M. H., Aublet, F., Jean, K., Lachance, S., Langlois, F., & De Pokomandy-Morin, K. (1995). Experimental manipulations of responsibility: an analogue test for models of obsessive ± compulsive disorder. Behaviour Research and Therapy, 33 (8), 937 ± 946. Lopatka, C., & Rachman, S. (1995). Perceived responsibility and compulsive checking: an experimental analysis. Behaviour Research and Therapy, 33 (6), 673 ± 684. Miller, R. (1986). Beyond ANOVA, basics of applied statistics. New York, NY: Wiley International. Rachman, S. (1993). Obsessions, responsibility and guilt. Behaviour Research and Therapy, 31 (2), 149 ± 154. Rachman, S., Thordarson, D. S., Shafran, R., & Woody, S. R. (1995). Perceived responsibility: structure and significance. Behaviour Research and Therapy, 33 (7), 779 ± 784. Salkovskis, P. M. (1985). Obsessional ± compulsive problems: a cognitive-behavioural analysis. Behaviour Research and Therapy, 23 (5), 571 ± 583. Salkovskis, P. M. (1989). Cognitive-behavioural factors and the persistence of intrusive thoughts in obsessional problems. Behaviour Research and Therapy, 27 (6), 677 ± 682. Salkovskis, P. M., Richards, H. C., & Forrester, E. (1995). The relationship between obsessional problems and intrusive thoughts. Behavioural and Cognitive Psychotherapy, 23, 281 ± 299. Shafran, R. (1997). The manipulation of responsibility in obsessive ± compulsive disorder. British Journal of Clinical Psychology, 36, 397 ± 407. Turner, S. M., Beidel, D. C., Dancu, C. V., & Stanley, M. A. (1989). An empirically derived inventory to measure social fears and anxiety: the Social Phobia and Anxiety Inventory. Psychological Assessment, 1, 35 ± 40.