- Email: [email protected]
Assessing excessive reassurance seeking in the anxiety disorders
Journal of Anxiety Disorders 25 (2011) 911–917
Contents lists available at ScienceDirect
Journal of Anxiety Disorders
Assessing excessive reassurance seeking in the anxiety disorders Neil A. Rector a,b,∗ , Katy Kamkar b,c , Stephanie E. Cassin a,b , Lindsay E. Ayearst d , Judith M. Laposa b,c a
Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada University of Toronto, Ontario, Canada c Centre for Addiction and Mental Health, Toronto, Ontario, Canada d York University, Toronto, Ontario, Canada b
a r t i c l e
i n f o
Article history: Received 15 November 2010 Received in revised form 9 May 2011 Accepted 9 May 2011 Keywords: Reassurance seeking Safety behaviors Anxiety Anxiety disorders Measurement CBT
a b s t r a c t Reassurance seeking has long been hypothesized to be a key factor in the maintenance of anxiety within contemporary cognitive-behavioral approaches to the conceptualization and treatment of anxiety disorders. However, empirical studies have lagged due to the absence of a reliable and valid measure of reassurance seeking. The present study sought to develop and examine the psychometric properties of a theoretically derived measure of reassurance seeking in treatment-seeking participants with DSM-IV-TR (American Psychiatric Association, 2000) social phobia (n = 116), generalized anxiety disorder (n = 75), panic disorder with or without agoraphobia (n = 50), and obsessive compulsive disorder (n = 42). Participants (N = 283) completed the Reassurance Seeking Scale (RSS), Depression Anxiety Stress Scale, Beck Anxiety Inventory, and Beck Depression Inventory-II. An exploratory factor analysis resulted in a coherent three factor solution reflecting the need to seek excessive reassurance regarding: (1) uncertainty about decisions, (2) attachment and the security of relationships, and (3) perceived general threat and anxiety. The RSS was found to possess good internal consistency and was moderately correlated with measures of anxiety, stress, and depression. The psychometric properties of the RSS appear promising for the promotion of programmatic research on reassurance seeking and its treatment in the anxiety disorders. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction Excessive reassurance seeking (ERS) has long been described as an important mechanism in managing psychological distress. Reassurance seeking has been shown to immediately reduce anxiety, but this temporary reduction in anxiety is typically followed by a paradoxical increase in anxiety and an urge to seek additional reassurance, leading to increased frequency of reassurance seeking over time (Abramowitz, Schwartz, & Whiteside, 2002; Salkovskis & Warwick, 1986). Excessive reassurance seeking also appears to interfere with habituation to anxiety and contributes to the maintenance of threat and an underestimation of the person’s ability to cope on their own with anxiety (Lohr, Olatunji, & Sawchuk, 2007; Parrish & Radomsky, 2010). Several definitions of ERS have been proposed, ranging from generic behavioral definitions such as “direct verbal requests for the repetitive provision of old information” (Salkovskis, 1985), to definitions that include the motivational factors thought to underlie ERS in specific psychological disorders.
∗ Corresponding author at: Sunnybrook Research Institute, Department of Psychiatry, Sunnybrook Health Sciences Centre, 2075 Bayview Ave., Toronto, Ontario, M4N 3M5, Canada. Tel.: +1 416 480 6100x2233; fax: +1 416 480 5345. E-mail address: [email protected] (N.A. Rector). 0887-6185/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.janxdis.2011.05.003
Salkovskis (1985) described ERS in the context of obsessive compulsive disorder (OCD) as an attempt to “put things right” and avert the possibility of being blamed by self or others for something that one may be responsible for by diffusing responsibility for adverse events. Rachman (2002) notes that despite appearing as requests for information, requests for reassurance are most typically attempts to find safety from harm. Thus, requests for reassurance might temporarily reduce psychological distress by reducing the perceived threat and/or increasing perceived coping resources by diffusing responsibility for adverse events so that one does not have to cope with the threat alone. Reassurance sought by individuals with OCD has been conceptualized as a compulsive checking behavior (Parrish & Radomsky, 2010; Rachman, 2002) and a form of neutralization behavior (Salkovskis, 1985). Reassurance seeking, neutralization, and compulsive checking share common features, and all can be conceptualized as attempts to reduce the probability of an adverse event, the effects of the event, or one’s responsibility for the event (Rachman, 2002; Salkovskis, 1996). For example, individuals with OCD might ask for reassurance that they will not become contaminated or spread contaminants to others after coming into contact with bodily fluids or dirty substances. They might also ask others to ensure that doors are locked and appliances are turned off in order to prevent harm to self or others, or as a way
912
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917
of diffusing responsibility in the event that a catastrophic event occurs. In the first empirical study to examine the content, triggers, and functions of ERS, individuals with OCD were compared to those with depression and healthy controls using a semi-structured interview (Parrish & Radomsky, 2010). Individuals with OCD reported that they most frequently sought reassurance regarding potential general threats (e.g., “Are you sure the stove is off”?), social threats (e.g., “Are you sure you’re not mad at me?”), and perceived performance/competence (e.g., “Would you tell me if I made the wrong choice?”). At the onset of ERS episodes, individuals with OCD reported elevated levels of anxiety and perceived threat. The most commonly endorsed triggers for ERS episodes were anxiety and doubts regarding potential threats, and to a lesser extent, perceived social threats. The primary functions of ERS were to reduce anxiety and to prevent harm. Consistent with observations made by Rachman (2002), it was concluded that ERS is functionally equivalent to checking behavior in the context of OCD. Although preliminary empirical studies examining ERS have focused on OCD, ERS has been hypothesized to play an important role in the maintenance of anxiety across the anxiety disorders, including social phobia (Heerey & Kring, 2007), generalized anxiety disorder (Woody & Rachman, 1994), panic disorder (Onur, Alkin, & Tural, 2007), and health anxiety (Abramowitz & Moore, 2007; Taylor & Asmundson, 2004). In this broader context, ERS has been defined as “the repeated solicitation of safety-related information from others about a threatening object, situation or interpersonal characteristic despite having already received this information” (Parrish & Radomsky, 2010, p. 211) with the purpose of “seeking to restore a sense of confidence or to reduce anxiety or apprehension” (Simpson & Weiner, 1989). In the anxiety literature, ERS has been included among a variety of “safety signals” (Lohr et al., 2007), “safety behaviors” (Abramowitz & Moore, 2007), and “anxiety neutralizing behaviors” (Parrish & Radomsky, 2006) used by anxious individuals to promote a sense of security. Safety behaviors have been defined as “behaviors which are performed in order to prevent or minimize a feared catastrophe” (Salkovskis, 1991), whereas safety signals are cues that signal either the offset of an aversive event or the absence of onset of an aversive event, and thus, assure safety from threat and allow for reduced vigilance (Lohr et al., 2007). Safety signals may be inanimate objects (e.g., hospital), but they are frequently other people (Carter, Hollon, Carson, & Shelton, 1995). Due to their fear of evaluation, individuals with social phobia often seek reassurance from others to ensure their anxiety symptoms were not apparent to others, they appeared competent during a presentation, or they came across okay at a social gathering. A recent psychometric study of safety behaviors in social phobia found that the item “Ask others about your performance” was one of the highest loading (.61) items on a factor reflecting ‘active’ safety behaviors, defined as actions performed by an individual in an attempt to present well in social situations (Cuming et al., 2009). A study examining social interactions in dyads consisting of two non-socially anxious individuals or one socially anxious and one non-socially anxious individual found that the social interactions with socially anxious individuals were characterized by more frequent reassurance seeking and giving (Heerey & Kring, 2007). Further, the reassurance seeking of socially anxious individuals was negatively correlated with their partners’ ratings of positive affect and perceptions of interaction quality, highlighting the interpersonal consequences of ERS. Individuals with GAD make frequent contact with, and repeatedly seek reassurance from, family, friends, professionals, and authorities (Woody & Rachman, 1994). The diagnostic criteria proposed for the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) includes “repeatedly seeking reassurance due to worries” as a behavior associated with
GAD (www.dsm5.org). Individuals with GAD might call romantic partners or friends to ensure their relationships are secure, family members to ensure they are safe, and doctors to ensure they are in good health. Individuals with GAD might also consult with several people before making decisions due to their intolerance of uncertainty (Dugas, Gagnon, Ladouceur, & Freeston, 1998). Individuals with panic disorder and agoraphobia experience anxiety when in places or situations in which escape might be difficult or help might not be available in the event of a panic attack. Bowlby (1973) suggested that agoraphobia can be understood as a condition of “anxious attachment” related to fear and apprehension over the availability and responsiveness of key attachment figures. Notably, anxiety is reduced substantially when the individual is accompanied by a “safe” reassuring person (Carter et al., 1995). Individuals with panic disorder who experience predominant respiratory symptoms (i.e., shortness of breath, choking/smothering sensations, fear of dying, chest pain/discomfort, and tingling/numbness) have been found to engage in significantly more reassurance seeking behavior compared to those without prominent respiratory symptoms (Onur et al., 2007). Those with predominant respiratory symptoms might rely on comfort provided by the safe person as a way of coping with a distressing sense of insecurity and disability (Onur et al., 2007). Taken together, empirical studies of reassurance seeking in OCD (Parrish & Radomsky, 2006, 2010) and cognitive-behavioral modeling of the anxiety disorders, suggest that ERS is a common problem that is motivated by a variety of concerns, including perceived threats to the safety of self/others, doubts about personal competence/abilities such as decision making, and social threats to interpersonal relationships. Beyond the conceptual development of ERS in the anxiety disorders, empirical research conducted with dysphoric and depressed participants has contributed to our understanding of the motivational factors underlying ERS as it pertains to different clinical disorders. In the context of depression research, ERS has been defined as “the relatively stable tendency to excessively and persistently seek assurance from others that one is lovable and worthy, regardless of whether such assurance has already been provided” (Burns, Brown, Plant, Sachs-Ericsson, & Joiner, 2006, p. 136). Seeking of reassurance to reduce threats associated with potential loss and abandonment in depression is consistent with cognitive accounts of depression that focus on the role of loss and issues pertaining to worthiness as a motivation source for ERS (Beck, 1967, 1976). The psychometric operationalization of ERS in depression research has occurred exclusively with The Depressive Interpersonal Relationships Inventory—Reassurance Seeking subscale which consists of four items that assess an individual’s tendency to excessively seek reassurance that others truly care about him or her (Joiner & Metalsky, 2001). For example, respondents are asked, “Do you frequently seek reassurance from the people you feel close to as to whether they really care about you?” and “Do the people you feel close to sometimes become irritated with you for seeking reassurance from them about whether they really care about you?” (Joiner & Metalsky, 2001, p. 372). A recent meta-analysis (Starr & Davila, 2008) reported that ERS is associated with both concurrent depression and interpersonal rejection, but concluded, “Perhaps the most striking finding of this meta-analysis is the near complete lack of methodological diversity across studies. Research has virtually always relied on a single, four item self-report measure of excessive reassurance seeking. Although this scale has shown strong psychometric properties, the exclusivity of its use makes it impossible to explore whether other methods of assessing excessive reassurance seeking yield similar results (p. 773).” Given accumulating evidence from the anxiety literature that a wide variety of perceived threats can trigger ERS, the Reassurance Seeking subscale (Joiner & Metalsky, 2001) may have inadequate
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917
construct validity in the broader assessment of ERS due to its exclusive focus on perceived threats of social loss or rejection. Parrish and Radomsky (2010) compared participants with depression versus OCD and found that the primary type of perceived threat that triggered ERS differed between clinical groups, such that individuals with depression most frequently sought reassurance regarding perceived social threats whereas those with OCD most frequently sought reassurance regarding perceived general threats. The qualitative assessment of ERS beyond the sole use of the Reassurance Seeking subscale demonstrated the significance of ERS to anxiety disorders, moving beyond previous research that only used the Reassurance Seeking subscale and found that ERS was unrelated cross-sectionally and prospectively to the development of anxiety symptoms (Joiner & Metalsky, 2001; Joiner & Schmidt, 1998). Interestingly, Parrish and Radomsky (2010) found that although the source of threat differed between the depressed and OCD groups, the primary motivation for ERS was similar in both diagnostic groups: ERS episodes were precipitated by elevated anxiety and threat estimations, and anxiety reduction was cited most frequently as the primary function of ERS, thus highlighting the importance of the ERS construct in anxiety. 1.1. Study rationale Empirical and clinical anecdotal evidence converge in suggesting that excessive reassurance is commonly sought by individuals with anxiety disorders in response to a variety of perceived threats, which in turn, contributes to the long-term maintenance of anxiety and threat. The existing measures of reassurance seeking were not designed to assess diverse triggers for seeking reassurance in the anxiety disorders, and instead, assess reassurance only with a limited set of items pertaining to perceived social threats in the context of depression (Joiner & Metalsky, 2001). There have been recent attempts to measure reassurance seeking behaviors with a single (Muse, McManus, Hackman, Williams, & Williams, 2010; “I seek reassurance about my health”) or limited set of behavioral items in response to physical health threats (Speckens, Spinhoven, vanHemert, & Bolk, 2000) and worry and uncertainty (Comer et al., 2009; Gosselin et al., 2008). However, the content and construct validity of these scales as a measure of the motivations for seeking reassurance and their clinical usefulness to assess ERS in the broader anxiety disorder spectrum may be somewhat limited. Given accumulating evidence that ERS is a common mechanism across the anxiety disorders, the aim of the current study was to develop a comprehensive measure of ERS tapping a variety of reasons for seeking reassurance that would be transdiagnostic across the anxiety disorders, and potentially a variety of clinical and non-clinical populations. 2. Method 2.1. Participants Participants (N = 283) were continuous referrals to a large university-based anxiety disorders assessment and treatment clinic. The sample was comprised of individuals meeting DSM-IVTR (American Psychiatric Association, 2000) criteria for primary social phobia (n = 116), generalized anxiety disorder (n = 75), panic disorder with or without agoraphobia (n = 50), and obsessive compulsive disorder (n = 42). All diagnoses were based on the Structured Clinical Interview for Axis 1 Disorders (SCID-1/P version 2.0). Participants had a mean age of 35.21 years (SD = 10.68). The sample was 55.8% female, 58.7% single, predominantly Caucasian (83.7%), and well-educated (89.4% had at least some college/university education). The majority of the sample (61.5%) was either working or
913
attending school full-time. At the time of assessment, 11% of the sample had a secondary current MDD diagnosis.
2.2. Measures Beck Anxiety Inventory (BAI; Beck & Steer, 1993) – The BAI is a 21-item self-report measure of anxiety symptoms. Respondents are asked to report how much they have been bothered by several symptoms during the previous week on a Likert scale ranging from 0 (“not at all”) to 3 (“severely, I could barely stand it”). The BAI demonstrated adequate internal consistency (˛ = .92) in the current study. Beck Depression Inventory (BDI-II; Beck, Steer, & Brown, 1996) – The BDI-II is a 21-item self-report measure that assesses the severity of depression symptoms. Respondents are asked to select one of four statements, on a 0–3 scale, which best describes how they have been feeling over the past 2 weeks. The BDI-II demonstrated adequate internal consistency (˛ = .92) in the current study. Depression Anxiety Stress Scale (DASS; Lovibond & Lovibond, 1995) – The DASS is a 42-item self-report measure that assesses the severity of depression, anxiety, and stress during the previous week. Respondents are asked to rate the extent to which each item applied to them on a Likert scale ranging from 0 (“did not apply to me at all”) to 3 (“applied to me very much, or most of the time”). All three DASS subscales demonstrated adequate internal consistency (˛ = .90 to .96) in the current study, as did the total score (˛ = .96). Reassurance Seeking Scale (RSS) – The RSS is a self-report measure that was designed to assess reassurance seeking in a variety of situations. The item content for the RSS was generated based on the previous description of putative anxiety-related triggers for excessive reassurance seeking, and as culled from clinical notes from patients with anxiety disorders participating in psychological treatment. Respondents are asked to rate the frequency with which they seek reassurance on a Likert scale ranging from 1 (“not at all”) to 5 (“extremely”). The item content for the RSS was generated based on the previous description of the mechanisms associated with ERS in the anxiety disorders and as culled from clinical notes from patients participating in cognitive-behavioral therapy for their anxiety condition. The details of the scale are described further in Section 3.
2.3. Procedure This study received ethical approval from the institutional research ethics board. All participants completed the questionnaire packet at the time of their initial diagnostic assessment.
2.4. Statistical analysis We subjected the item pool to a principal axis factor analysis with an oblique rotation using Mplus 4.0 (Muthén & Muthén, 2006). Mplus was chosen over SPSS because of its ability to handle categorical data. As noted by Flora, Finkel, and Foshee (2003), when item responses are given using a Likert-type scale format, the observed responses are ordered, categorical manifestations of a continuous, psychological process of judgment about item content. Research has demonstrated extensively that exploratory factor analytic techniques that rely on normal-theory estimation using Pearson product–moment relations are not robust to such categorization of continuous variables (e.g., Bollen, 1989; Nunnally & Bernstein, 1994; West, Finch, & Curran, 1995). Mplus uses the matrix of polychoric correlations rather than the matrix of Pearson product–moment correlations for the analysis as this is considered the more appropriate choice for data that uses Likert scale ratings (Flora et al., 2003).
914
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917
3. Results
Table 1 Promax-rotated factor loadings.
3.1. Exploratory factor analysis
Item content
Using a sample of 283 participants, with all 40 items as variables, the analysis proceeded by first calculating the matrix of polychoric correlations among all possible pairs of items. We examined the polychoric correlation matrix to identify item pairs that were highly correlated and therefore possibly locally dependent. Results suggested the elimination of three items from the scale. Next the EFA was conducted with the polychoric correlations using ULS estimation. To assess number of factors underlying the set of test items, we used parallel analysis (Horn, 1965). Parallel analysis (O’Connor, 2000) for the 95th percentile in 1000 random data sets suggested a three-factor solution. However, research has suggested that parallel analysis is less reliable when using the polychoric correlation matrix. As such, we also examined a scree plot of the eigenvalues resulting from the exploratory factor analysis. The scree plot clearly showed that a large first factor might be adequate to explain the relationship among these items. However, the plot also suggested that a three-factor solution might underlie the data. Because we expected the factors to be correlated, we applied an oblique rotation, the Promax method, to the three-factor solution to aid interpretation. The one factor model demonstrated poor fit to the data based on the root mean square residual (RMSR = .09). Although there are no clear guidelines with respect to evaluation of fit indices, fit was considered good if the RMSR ≤ .05. The fit for the three-factor model was better (RMSR = .05). Through several iterations of retaining and deleting items on the basis of their polychoric inter-item correlations, factor loadings, communality estimates, and contribution to coefficient alpha, the total number of items was reduced from 40 to 30. Factor loadings for each of the 30 items are reported in Table 1. The final three-factor model of ERS demonstrated a good fit to the data (RMSR = .048). The first factor consisted of 13 items pertaining to seeking reassurance around making decisions and the factor was labeled: RSS-Decision-Making. For example, items include excessive reassurance seeking “Prior to making a decision,” “When you have to choose among alternative options,” or “When you doubt your decision.” The second factor consisted of eight items measuring reassurance seeking pertaining to social affiliation and rejection and the factor was subsequently labeled: RSS-Social Attachment. Examples of items include seeking excessive reassurance “To whether you are loved or cared,” “To get support from others” and “To get approval from others.” The third factor included nine items assessing seeking reassurance to reduce perceived general threat and coping with anxiety and was subsequently labeled, RSS-General Threat. Examples of items include seeking reassurance, “To prevent the occurrence of a catastrophic event,” “To whether something bad is going to happen to you” and “To feel more relaxed.” Correlations among the three factors were strong and positive (r’s ranging from .59 to .64). The three-factor based scales were then scored and reliability estimates were calculated. Alphas for the three scales ranged from .88 to .93, suggesting excellent internal consistency based on a minimal acceptable or lower bound standard of .70 to .79. Because coefficient alpha is influenced by the number of items on a scale, the average inter-item correlation was also examined to assess the internal consistency of the scales. Researchers have previously recommended that for assessing broad constructs, a minimum inter-item correlation of .15 to .20 is desired, whereas for a reliable and valid measure of a narrower construct, a higher inter-item correlation will be necessary (e.g., .40 to .45) (Clark & Watson, 1995). Average inter-item correlations for the three factor based scales were consistent with recommendations for assessment of narrower constructs, ranging from .45 to .52. A summary of the properties for
Prior to making a decision? When you have to choose among alternative options? When you doubt your decision? To whether you have considered all the possible details prior to making a decision? When you have to do something on your own? Before initiating or doing things? When you think you have made the wrong decision? To gain more certainty about a situation or something that is uncertain? Prior to making a change in some areas of your life (e.g., career, academic, relationships)? When you have a lot of responsibility about something? Before exploring something new? To avoid feeling responsible for the outcome of decisions in major areas of your life? To decrease your sense of personal responsibility? To whether you are loved or cared for? When you are not getting “enough attention”? To whether you are a lovable/caring person? To get support from others? To get approval from others? To feel close to others? To whether you have received a negative evaluation? To whether others are upset with you? To whether something bad is going to happen to you? To make sure you are okay? To prevent the occurrence of a catastrophic event? When you think a negative event is likely to occur? To whether you are safe? To feel more relaxed? To feel better inside? To turn off your anxiety feelings? To gain more peace and serenity within yourself?
DecisionMaking
Social Attachment
General Threat
.92 .91
−.02 −.21
−.07 .12
.84 .82
.06 −.09
−.06 .04
.75
−.01
.03
.71 .66
−.03 .09
.15 .10
.64
.12
.15
.60
.23
−.12
.55
.09
.21
.52 .50
.08 .17
.13 .06
.44
.23
.05
−.13 −.07
.89 .78
.01 −.03
−.06
.78
.07
.06 .25 −.13 .17
.72 .69 .65 .50
.00 −.16 .18 .11
.27 .03
.47 −.11
.05 .76
.07 −.13
−.06 −.02
.76 .77
.12
.03
.69
.02 .15 .01 .12 −.02
−.01 .03 .24 .15 .23
.66 .58 .56 .53 .52
Note. Factor loadings ≥.40 are in bold.
the 30-item scale with three-factor based subscales is presented in Table 2. 3.2. Scale descriptives and validity Scale descriptives for the measures used in the present study are presented in Table 3. To test the direction and magnitude of the association between RSS scores and measures of anxiety, stress, and depression, zero-order Pearson correlations were computed Table 2 Properties of the Reassurance Seeking Scale. Factor
Label
Total items
Range of loadings
Alpha
AIC
1 2 3
Decision-Making Social Attachment General Threat
13 8 9
.44 to .92 .47 to .89 .52 to .77
.93 .88 .88
.52 .48 .45
Note. AIC: average inter-item correlation. N = 283.
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917 Table 3 Scale descriptives. Scale Reassurance Seeking Scale Decision-Making Social Attachment General Threat Depression Anxiety Stress Scale Anxiety Stress Depression Beck Anxiety Inventory Beck Depression Inventory-II
M
SD
42.54 24.13 27.53
11.31 7.33 7.99
15.71 20.50 16.90 20.25 22.40
9.72 9.26 11.84 12.22 12.20
(Table 4). Scores on all three factors of the RSS were found to be significantly and positively correlated with scores on the DASS, BAI, and BDI-II. As an exploratory test of the convergent and divergent validity of the RSS, the correlations among the anxiety measures and the three RSS subscales were compared, as were the correlations between the depression measures and the three RSS subscales. Following the procedures set out by Meng, Rosenthal, and Rubin (1992), the correlation between the BAI and the RSS-General Threat subscale was significantly greater than between the BAI and the RSS-Decision-Making subscale (z = 2.95, p < .01) and between the DASS-Anxiety and the RSS-Social Attachment subscale (z = 3.95, p < .01). Similarly, the correlation between the DASS-Anxiety subscale and the RSS-General Threat subscale was significantly greater than between the DASS-Anxiety subscale and the RSS-DecisionMaking subscale (z = 2.00, p < .01) and between the DASS-Anxiety and the RSS-Social Attachment subscale (z = 3.24, p < .01). Finally, the correlation between the DASS-Anxiety subscale and the RSSGeneral Threat subscale was significantly greater than between the DASS-Depression subscale and the RSS-General Threat subscale (z = 2.90, p < .01). In contrast to the anxiety measures, the correlation between the depression measures (DASS-Depression and BDI) and the three RSS subscales were not significantly different. 4. Discussion The present study sought to develop and examine psychometric properties of a measure of reassurance seeking that assessed common triggers for patients with primary DSM-IV anxiety disorders. Items were theoretically derived and an exploratory factor analysis resulted in a three factor solution reflecting excessive reassurance seeking around indecisiveness and making decisions, social affiliation and fear of loss and rejection, and perceived general threat and the ability to cope with anxiety. The new measure, the Reassurance Seeking Scale, was found to possess good internal consistency,
915
and was positively and moderately correlated with measures of anxiety, stress, and depression. There was also some evidence of convergent validity for the new measure, with excessive reassurance pertaining to perceived general threat and anxiety being more strongly associated with anxiety symptom severity than depression symptom severity. White and Barlow (2002) describe safety behaviors broadly as actions that promote feelings of security that can be derived from the external and internal environment. Reassurance seeking can be conceptualized as a particular type of safety behavior aimed at reducing threat and restoring a sense of confidence and security through the repeated solicitation of safety-related information from others (Parrish & Radomsky, 2010; Woody & Rachman, 1994). Reassurance seeking as a safety behavior may emerge from both the overestimation of threat and danger in conjunction with an underestimation of the person’s perceived ability to cope with the threat and danger. The three RSS factors that resulted from the exploratory factor analysis closely reflect this operationalization of reassurance seeking in the anxiety disorders. Patients report engaging in excessive reassurance seeking when faced with indecision, doubt about the security of their relationships, and in anticipation of general threat. They also map on to the only study examining the content and function of reassurance seeking reported by clinical samples of individuals with OCD. The clinical group in the study by Parrish and Radomsky (2010) reported seeking reassurance regarding perceived threats to safety, perceived social threats (such as loss or rejection), and concerns about personal performance/competence (such as decision making). Results of this study yielding a factor characterizing reassurance seeking when feeling uncertain about decisions is consistent with the empirical literature supporting a link between anxiety disorders and indecisiveness. Individuals with DSM-IV diagnosed anxiety disorders score higher than non-clinical participants on facets of perfectionism related to concern over mistakes (e.g., “I should be upset if I make a mistake”) and doubts about actions (e.g., “I usually have doubts about the simple everyday things I do”) (Antony, Purdon, Huta, & Swinson, 1998; Purdon, Antony, & Swinson, 1999). The doubts about actions facet reflects the tendency to be uncertain about the correctness of one’s decisions or behaviors. These perceived threats may lead individuals with anxiety disorders to seek reassurance (i.e., safety) to ensure they do not make mistakes or poor decisions, particularly if they overestimate the consequences of making such an error. A recent study with nonclinical participants (Wu & Wei, 2008) found significant positive correlations between elevated scores on concerns over mistakes and doubts about actions and ERS. The second factor measuring social attachment and threat overlaps conceptually with the previous psychometric operationalization of ERS in depression research, which has occurred
Table 4 Correlation matrix. RSS Decisions RSS Decisions Attachment Threat DASS Anxiety Stress Depression BAI BDI
DASS Attach
Threat
.64 .63
– .59
–
.33 .42 .32 .28 .45
.31 .46 .40 .22 .40
Anxiety
Stress
Depress
– .66 .53 .83 .59
– .60 .51 .66
– .49 .78
BAI
BDI
– .61
–
–
.48 .43 .33 .42 .37
Note. RSS: Reassurance Seeking Scale (subscales: Decision-Making (Decisions), Social Attachment (Attach), General Threat (Threat)); DASS: Depression Anxiety Stress Scale (subscales: Depression, Anxiety, Stress); BAI: Beck Anxiety Inventory; BDI: Beck Depression Inventory-II. All correlations are significant, p < .001.
916
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917
exclusively with The Depressive Interpersonal Relationships Inventory—Reassurance Seeking subscale. This subscale consists of four items that assess an individual’s tendency to excessively seek reassurance that others truly care about him or her (Joiner & Metalsky, 2001). For example, respondents are asked, “Do you frequently seek reassurance from the people you feel close to as to whether they really care about you?” and “Do the people you feel close to sometimes become irritated with you for seeking reassurance from them about whether they really care about you?” (Joiner & Metalsky, 2001, p. 372). In the previous study by Parrish and Radomsky (2010), depressed patients were more likely to report ERS following attachment/evaluation triggers than were patients with OCD, but the latter group still reported attachment/evaluation triggers as a common reason for ERS. One possibility is that attachment/evaluation triggers increase the likelihood of ERS in those experiencing general negative affectivity, depression or anxiety, opposed to the specific form of negative affectivity (i.e., depression versus anxiety). A second possibility is that the heightened ERS following attachment/evaluation triggers in patients with a primary DSM-IV anxiety disorder, such as the current study, is due to the presence of secondary comorbid depression. However, in the current study only 7% of the sample had a secondary MDD diagnosis (or 11% if those in partial remission are included). Also, the correlation between the RSS-Social Attachment subscale and the BDI-II was significant but only accounted for 16% of the variance. Finally, there was uniformity in the pattern of associations between the BDI-II and all three RSS subscales. A third possibility is that ERS occurs in mood and anxiety disorders to the extent that there is a shared underlying vulnerability. It has been noted by others that the four-item ERS scale developed by Joiner and Metalsky contains items that are virtually identical to ones that measure anxious attachment (Shaver, Schachner, & Mikulincer, 2005). Research has demonstrated that anxious attachment is associated with depression (e.g., Bifulco, Moran, Ball, & Bernazzani, 2002) largely through the presence of self-criticism and dysfunctional beliefs (Shaver et al., 2005), factors that are not specific to depression. As such, it has been argued that ERS is more centrally related to anxious attachment. In a direct test of this hypothesis, Shaver et al. found a strong association between measures of anxious attachment and ERS, and found that a significant association between ERS and depression was attenuated to nonsignificance when anxious attachment was controlled, even though the association between anxious attachment and ERS remained when controlling for depression. Other research with non-clinical samples has similarly found a strong association between anxious attachment and ERS (Mikulincer, Gilliath, Sapir-Lavid, & Yaakobi, 2003). In summary, previous findings demonstrating an association between ERS pertaining to worries about rejection and abandonment and depression, and the results of the current study pointing to seeking reassurance around attachment/evaluation concerns, suggests possibility of a common shared vulnerability relating to anxious attachment and associated negative interpersonal beliefs as one common trigger for ERS in depression and anxiety disorders. One area for future investigation is the examination of distinction and overlap in the three RSS factors in depressed and anxious samples along with other putative measures of relationship and cognitive vulnerability. Rachman (2002) has hypothesized that requests for reassurance are most typically attempts to find safety from harm – to reduce psychological distress by reducing the perceived threat and/or increasing perceived coping resources. The third factor of the RSS (Threat/Anxiety) appears to be largely tapping the seeking of reassurance to find safety from looming general threat and to reduce feelings of anxiety. In the previous study by Parrish and Radomsky (2010), the primary reason why individuals with OCD reported they sought reassurance was to cope with elevated
anxiety and general threats, and this was more frequently a trigger for ERS in the OCD sample than it was for participants with major depression. Further, we found that measures of anxious arousal (i.e., BAI; DASS-A) were more strongly associated with the RSS-General Threat/Anxiety dimension than with the other two RSS dimensions, thus providing evidence for the specificity of this factor to anxious threat and somatic discomfort. These results are in keeping with cognitive-behavioral conceptualizations that emphasize the role of exaggerated threat appraisal in leading to the somatic and behavioral aspects of the anxiety cycle (e.g., Beck & Clark, 1997; Beck, Emery, & Greenberg, 1985; Ouimet, Gawronski, & Dozois, 2009). The findings of the current study regarding ERS in the anxiety disorders provide direction for clinicians to conceptualize both the general concerns that motivate such behavior in combination with a functional assessment of the disorder-specific triggers for ERS behaviors. Cognitive-behavioral treatment manuals with proven efficacy for the treatment of anxiety disorders often include explicit guidelines for the reduction of safety behaviors and reassurance – seeking in particular, although future research is required to test the short and long-term effects of reducing ERS directly. 4.1. Limitations At this preliminary stage, psychometric properties of the RSS appear promising. Although the sample size in the current study was considered sufficient for exploratory factor analysis (Floyd & Widaman, 1995; Gorsuch, 1983), some authors advocate for a subject to item ratio of 10:1 (Nunnally & Bernstein, 1994). Further validation studies with larger samples, including other anxiety disorders not sampled in the current study such as PTSD, are required to ensure the construct validity and factor stability of the scale. Retest-reliability studies are also required to assess the temporal stability of the RSS. One important direction for future research would be to use item response theory to test the relations between items and the construct of reassurance seeking, and to eliminate redundant or biased items (Embretson & Reise, 2000). Finally, treatment studies will determine the sensitivity of the measure in response to clinical gains. The RSS could be a useful clinical tool to assess changes in reassurance seeking over the course of treatment, and to examine whether reduction in reassurance seeking mediates reduction in anxiety. Acknowledgements The authors would like to thank Sam Hanig and Vincent Man for their editorial assistance in the preparation of this manuscript. References Abramowitz, J. S., & Moore, E. L. (2007). An experimental analysis of hypochondriasis. Behavior Research and Therapy, 45, 413–424. Abramowitz, J. S., Schwartz, S. A., & Whiteside, S. P. (2002). A contemporary conceptual model of hypochondriasis. Mayo Clinic Proceedings, 77, 1323–1330. Antony, M. M., Purdon, C. L., Huta, V., & Swinson, R. P. (1998). Dimensions of perfectionism across the anxiety disorders. Behavior Research and Therapy, 36, 1143–1154. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association. Beck, A. T. (1967). Depression: clinical, experimental, and theoretical aspects. New York: Harper and Row. Republished as: Beck, A. T. (1970). Depression: causes and treatment. Philadelphia: University of Pennsylvania Press Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: Meridian. Beck, A. T., & Clark, D. A. (1997). An information processing model of anxiety: automatic and strategic processes. Behavior Research and Therapy, 35, 49–58. Beck, A. T., & Steer, R. A. (1993). Manual for the beck anxiety inventory. San Antonio, TX: Psychological Corporation. Beck, A. T., Emery, G., & Greenberg, R. I. (1985). Anxiety disorders and phobias: a cognitive perspective. New York, NY: Basic Books.
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917 Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for the beck depression inventory (2nd ed.). Boston: Harcourt, Brace, and Company. Bifulco, A., Moran, P., Ball, C., & Bernazzani, O. (2002). Adult attachment style I: is relationship to clinical depression. Social Psychiatry and Psychiatric Epidemiology, 37, 50–59. Bollen, K. A. (1989). Structural equations with latent variables. Wiley series in probability and mathematical statistics. New York: Wiley. Bowlby, J. (1973). Attachment and loss (volume II): separation: anxiety and anger. New York: Basic Books. Burns, A. B., Brown, J. S., Plant, E. A., Sachs-Ericsson, N., & Joiner, T. E., Jr. (2006). On the specific depressotypic nature of excessive reassurance-seeking. Personality and Individual Differences, 40, 135–145. Carter, M. M., Hollon, S. D., Carson, R., & Shelton, R. C. (1995). Effects of a safe person on induced distress following a biological challenge in panic disorder with agoraphobia. Journal of Abnormal Psychology, 104, 156–163. Clark, L. A., & Watson, D. (1995). Constructing validity: basic issues in objective scale development. Psychological Assessment, 7, 309–319. Comer, J. S., Roy, A. K., Furr, J. M., Gotimer, K., Beidas, R. S., Dugas, M. J., et al. (2009). The Intolerance of Uncertainty Scale for children: a psychometric evaluation. Psychological Assessment, 21, 402–411. Cuming, S., Rapee, R. M., Kemp, N., Abbott, M. J., Peters, L., & Gaston, J. E. (2009). A self-report measure of subtle avoidance and safety behaviors relevant to social anxiety: development and psychometric properties. Journal of Anxiety Disorders, 23, 879–883. Dugas, M., Gagnon, F., Ladouceur, R., & Freeston, M. (1998). Generalized anxiety disorder – a preliminary test of a conceptual model. Behavior Research and Therapy, 36, 215–226. Embretson, S. E., & Reise, S. P. (2000). Item response theory for psychologists. Mahwah, NJ: Lawrence Erlbaum. Flora, D. B., Finkel, E. J., & Foshee, V. A. (2003). Higher order factor structure of a self-control test: evidence from confirmatory factor analysis with polychoric correlations. Educational and Psychological Measurement, 63, 112–127. Floyd, F. J., & Widaman, K. F. (1995). Factor analysis in the development and refinement of clinical assessment instruments. Psychological Assessment, 7, 286– 299. Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Gosselin, P., Ladouceur, R., Evers, A., Laverdiere, A., Routhier, S., & Tremblay-Pickard, M. (2008). Evaluation of intolerance of uncertainty: development and validation of a new self-report measure. Journal of Anxiety Disorders, 22, 1427–1439. Heerey, E. A., & Kring, A. M. (2007). Interpersonal consequences of social anxiety. Journal of Abnormal Psychology, 116, 125–134. Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179–185. Joiner, T. E., Jr., & Metalsky, G. I. (2001). Excessive reassurance seeking: delineating a risk factor involved in the development of depressive symptoms. Psychological Science, 12, 371–378. Joiner, T. E., Jr., & Schmidt, N. B. (1998). Excessive reassurance-seeking predicts depressive but not anxious reactions to acute stress. Journal of Abnormal Psychology, 107, 533–537. Lohr, J. M., Olatunji, B. O., & Sawchuk, C. N. (2007). A functional analysis of danger and safety signals in anxiety disorders. Clinical Psychology Review, 27, 114–126. Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the Depression Anxiety Stress Scales (2nd ed.). Sydney: Psychology Foundation. Meng, X. L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated correlation coefficients. Psychological Bulletin, 111, 172–175. Mikulincer, M., Gilliath, O., Sapir-Lavid, Y., & Yaakobi, E. (2003). Attachment theory and concern for others’ welfare: evidence that activation of the sense of secure base promotes endorsement of self-transcendence values. Basic and Applied Social Psychology, 25, 299–312.
917
Muse, K., McManus, F., Hackman, A., Williams, M., & Williams, M. (2010). Intrusive imagery in sever health anxiety: prevalence, nature and links with memories and maintenance cycles. Behavior Research and Therapy, 48, 792–798. Muthén, L. K., & Muthén, B. O. (2006). Mplus user’s guide (4th ed.). Los Angeles: Muthén & Muthén. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill. O’Connor, B. P. (2000). SPSS and SAS programs for determining the number of components using parallel analysis and Velicer’s MAP test. Behavior Research Methods, Instrumentation, and Computers, 32, 396–402. Onur, E., Alkin, T., & Tural, U. (2007). Panic disorder symptoms: further clinical differences. Depression and Anxiety, 24, 479–486. Ouimet, A. J., Gawronski, B., & Dozois, D. J. A. (2009). Cognitive vulnerability to anxiety: a review and an integrative model. Clinical Psychology Review, 29, 459– 470. Parrish, C. L., & Radomsky, A. S. (2006). An experimental analysis of responsibility and reassurance: relationships with compulsive checking. International Journal of Behavioral Consultation and Therapy, 2, 174–191. Parrish, C. L., & Radomsky, A. S. (2010). Why do people seek reassurance and check repeatedly? An investigation of factors involved in compulsive behavior in OCD and depression. Journal of Anxiety Disorders, 24, 211–222. Purdon, C., Antony, M. M., & Swinson, R. P. (1999). Psychometric properties of the Frost Multidimensional Perfectionism Scale in a clinical anxiety disorders sample. Journal of Clinical Psychology, 55, 1271–1286. Rachman, S. (2002). A cognitive theory of compulsive checking. Behavior Research and Therapy, 40, 624–639. Salkovskis, P. M. (1985). Obsessional-compulsive problems: a cognitive-behavioral analysis. Behavior Research and Therapy, 23, 571–583. Salkovskis, P. M. (1991). The importance of behavior in the maintenance of anxiety and panic: a cognitive account. Behavioral Psychotherapy, 19, 6–19. Salkovskis, P. M. (1996). The cognitive approach to anxiety: threat beliefs, safety seeking behavior and the special case of health anxiety and obsessions. In: P. M. Salkovskis (Ed.), Frontiers of cognitive therapy (pp. 48–74). New York, NY: Guilford Press. Salkovskis, P. M., & Warwick, H. M. C. (1986). Morbid preoccupations, health anxiety, and reassurance: a cognitive behavioral approach to hypochondriasis. Behavior Research and Therapy, 24, 597–602. Shaver, P. R., Schachner, D. A., & Mikulincer, M. (2005). Attachment style, excessive reassurance seeking, relationship process, and depression. Personality and Social Psychology Bulletin, 31, 343–359. Simpson, J., & Weiner, E. (Eds.). (1989). Oxford english dictionary. United Kingdom: Oxford University Press. Speckens, A. E. M., Spinhoven, P., van Hemert, A. M., & Bolk, J. H. (2000). The Reassurance Questionnaire (RQ): psychometric properties of a self-report questionnaire. Psychological Medicine, 30, 841–847. Starr, L. R., & Davila, J. (2008). Excessive reassurance seeking, depression, and interpersonal rejection: a meta-analytic review. Clinical Psychology Review, 117, 762–775. Taylor, S., & Asmundson, G. J. G. (2004). Treating health anxiety: a cognitive behavioral approach. New York: Guilford Press. West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models with nonnormal variables. In: R. H. Hoyle (Ed.), Structural equation modeling: concepts, issues, and applications (pp. 56–75). Thousand Oaks, CA: Sage. White, K. S., & Barlow, D. H. (2002). Panic disorder and agoraphobia. In: D. H. Barlow (Ed.), Anxiety and its disorders: the nature and treatment of anxiety and panic. 2 (pp. 328–379). New York: Guilford Press. Woody, S., & Rachman, S. (1994). Generalized anxiety disorder as an unsuccessful search for safety. Clinical Psychology Review, 14, 743–753. Wu, T. F., & Wei, M. (2008). Perfectionism and negative mood: the mediating roles of validation from others versus self. Journal of Counseling Psychology, 55, 276–288.
Contents lists available at ScienceDirect
Journal of Anxiety Disorders
Assessing excessive reassurance seeking in the anxiety disorders Neil A. Rector a,b,∗ , Katy Kamkar b,c , Stephanie E. Cassin a,b , Lindsay E. Ayearst d , Judith M. Laposa b,c a
Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada University of Toronto, Ontario, Canada c Centre for Addiction and Mental Health, Toronto, Ontario, Canada d York University, Toronto, Ontario, Canada b
a r t i c l e
i n f o
Article history: Received 15 November 2010 Received in revised form 9 May 2011 Accepted 9 May 2011 Keywords: Reassurance seeking Safety behaviors Anxiety Anxiety disorders Measurement CBT
a b s t r a c t Reassurance seeking has long been hypothesized to be a key factor in the maintenance of anxiety within contemporary cognitive-behavioral approaches to the conceptualization and treatment of anxiety disorders. However, empirical studies have lagged due to the absence of a reliable and valid measure of reassurance seeking. The present study sought to develop and examine the psychometric properties of a theoretically derived measure of reassurance seeking in treatment-seeking participants with DSM-IV-TR (American Psychiatric Association, 2000) social phobia (n = 116), generalized anxiety disorder (n = 75), panic disorder with or without agoraphobia (n = 50), and obsessive compulsive disorder (n = 42). Participants (N = 283) completed the Reassurance Seeking Scale (RSS), Depression Anxiety Stress Scale, Beck Anxiety Inventory, and Beck Depression Inventory-II. An exploratory factor analysis resulted in a coherent three factor solution reflecting the need to seek excessive reassurance regarding: (1) uncertainty about decisions, (2) attachment and the security of relationships, and (3) perceived general threat and anxiety. The RSS was found to possess good internal consistency and was moderately correlated with measures of anxiety, stress, and depression. The psychometric properties of the RSS appear promising for the promotion of programmatic research on reassurance seeking and its treatment in the anxiety disorders. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction Excessive reassurance seeking (ERS) has long been described as an important mechanism in managing psychological distress. Reassurance seeking has been shown to immediately reduce anxiety, but this temporary reduction in anxiety is typically followed by a paradoxical increase in anxiety and an urge to seek additional reassurance, leading to increased frequency of reassurance seeking over time (Abramowitz, Schwartz, & Whiteside, 2002; Salkovskis & Warwick, 1986). Excessive reassurance seeking also appears to interfere with habituation to anxiety and contributes to the maintenance of threat and an underestimation of the person’s ability to cope on their own with anxiety (Lohr, Olatunji, & Sawchuk, 2007; Parrish & Radomsky, 2010). Several definitions of ERS have been proposed, ranging from generic behavioral definitions such as “direct verbal requests for the repetitive provision of old information” (Salkovskis, 1985), to definitions that include the motivational factors thought to underlie ERS in specific psychological disorders.
∗ Corresponding author at: Sunnybrook Research Institute, Department of Psychiatry, Sunnybrook Health Sciences Centre, 2075 Bayview Ave., Toronto, Ontario, M4N 3M5, Canada. Tel.: +1 416 480 6100x2233; fax: +1 416 480 5345. E-mail address: [email protected] (N.A. Rector). 0887-6185/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.janxdis.2011.05.003
Salkovskis (1985) described ERS in the context of obsessive compulsive disorder (OCD) as an attempt to “put things right” and avert the possibility of being blamed by self or others for something that one may be responsible for by diffusing responsibility for adverse events. Rachman (2002) notes that despite appearing as requests for information, requests for reassurance are most typically attempts to find safety from harm. Thus, requests for reassurance might temporarily reduce psychological distress by reducing the perceived threat and/or increasing perceived coping resources by diffusing responsibility for adverse events so that one does not have to cope with the threat alone. Reassurance sought by individuals with OCD has been conceptualized as a compulsive checking behavior (Parrish & Radomsky, 2010; Rachman, 2002) and a form of neutralization behavior (Salkovskis, 1985). Reassurance seeking, neutralization, and compulsive checking share common features, and all can be conceptualized as attempts to reduce the probability of an adverse event, the effects of the event, or one’s responsibility for the event (Rachman, 2002; Salkovskis, 1996). For example, individuals with OCD might ask for reassurance that they will not become contaminated or spread contaminants to others after coming into contact with bodily fluids or dirty substances. They might also ask others to ensure that doors are locked and appliances are turned off in order to prevent harm to self or others, or as a way
912
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917
of diffusing responsibility in the event that a catastrophic event occurs. In the first empirical study to examine the content, triggers, and functions of ERS, individuals with OCD were compared to those with depression and healthy controls using a semi-structured interview (Parrish & Radomsky, 2010). Individuals with OCD reported that they most frequently sought reassurance regarding potential general threats (e.g., “Are you sure the stove is off”?), social threats (e.g., “Are you sure you’re not mad at me?”), and perceived performance/competence (e.g., “Would you tell me if I made the wrong choice?”). At the onset of ERS episodes, individuals with OCD reported elevated levels of anxiety and perceived threat. The most commonly endorsed triggers for ERS episodes were anxiety and doubts regarding potential threats, and to a lesser extent, perceived social threats. The primary functions of ERS were to reduce anxiety and to prevent harm. Consistent with observations made by Rachman (2002), it was concluded that ERS is functionally equivalent to checking behavior in the context of OCD. Although preliminary empirical studies examining ERS have focused on OCD, ERS has been hypothesized to play an important role in the maintenance of anxiety across the anxiety disorders, including social phobia (Heerey & Kring, 2007), generalized anxiety disorder (Woody & Rachman, 1994), panic disorder (Onur, Alkin, & Tural, 2007), and health anxiety (Abramowitz & Moore, 2007; Taylor & Asmundson, 2004). In this broader context, ERS has been defined as “the repeated solicitation of safety-related information from others about a threatening object, situation or interpersonal characteristic despite having already received this information” (Parrish & Radomsky, 2010, p. 211) with the purpose of “seeking to restore a sense of confidence or to reduce anxiety or apprehension” (Simpson & Weiner, 1989). In the anxiety literature, ERS has been included among a variety of “safety signals” (Lohr et al., 2007), “safety behaviors” (Abramowitz & Moore, 2007), and “anxiety neutralizing behaviors” (Parrish & Radomsky, 2006) used by anxious individuals to promote a sense of security. Safety behaviors have been defined as “behaviors which are performed in order to prevent or minimize a feared catastrophe” (Salkovskis, 1991), whereas safety signals are cues that signal either the offset of an aversive event or the absence of onset of an aversive event, and thus, assure safety from threat and allow for reduced vigilance (Lohr et al., 2007). Safety signals may be inanimate objects (e.g., hospital), but they are frequently other people (Carter, Hollon, Carson, & Shelton, 1995). Due to their fear of evaluation, individuals with social phobia often seek reassurance from others to ensure their anxiety symptoms were not apparent to others, they appeared competent during a presentation, or they came across okay at a social gathering. A recent psychometric study of safety behaviors in social phobia found that the item “Ask others about your performance” was one of the highest loading (.61) items on a factor reflecting ‘active’ safety behaviors, defined as actions performed by an individual in an attempt to present well in social situations (Cuming et al., 2009). A study examining social interactions in dyads consisting of two non-socially anxious individuals or one socially anxious and one non-socially anxious individual found that the social interactions with socially anxious individuals were characterized by more frequent reassurance seeking and giving (Heerey & Kring, 2007). Further, the reassurance seeking of socially anxious individuals was negatively correlated with their partners’ ratings of positive affect and perceptions of interaction quality, highlighting the interpersonal consequences of ERS. Individuals with GAD make frequent contact with, and repeatedly seek reassurance from, family, friends, professionals, and authorities (Woody & Rachman, 1994). The diagnostic criteria proposed for the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) includes “repeatedly seeking reassurance due to worries” as a behavior associated with
GAD (www.dsm5.org). Individuals with GAD might call romantic partners or friends to ensure their relationships are secure, family members to ensure they are safe, and doctors to ensure they are in good health. Individuals with GAD might also consult with several people before making decisions due to their intolerance of uncertainty (Dugas, Gagnon, Ladouceur, & Freeston, 1998). Individuals with panic disorder and agoraphobia experience anxiety when in places or situations in which escape might be difficult or help might not be available in the event of a panic attack. Bowlby (1973) suggested that agoraphobia can be understood as a condition of “anxious attachment” related to fear and apprehension over the availability and responsiveness of key attachment figures. Notably, anxiety is reduced substantially when the individual is accompanied by a “safe” reassuring person (Carter et al., 1995). Individuals with panic disorder who experience predominant respiratory symptoms (i.e., shortness of breath, choking/smothering sensations, fear of dying, chest pain/discomfort, and tingling/numbness) have been found to engage in significantly more reassurance seeking behavior compared to those without prominent respiratory symptoms (Onur et al., 2007). Those with predominant respiratory symptoms might rely on comfort provided by the safe person as a way of coping with a distressing sense of insecurity and disability (Onur et al., 2007). Taken together, empirical studies of reassurance seeking in OCD (Parrish & Radomsky, 2006, 2010) and cognitive-behavioral modeling of the anxiety disorders, suggest that ERS is a common problem that is motivated by a variety of concerns, including perceived threats to the safety of self/others, doubts about personal competence/abilities such as decision making, and social threats to interpersonal relationships. Beyond the conceptual development of ERS in the anxiety disorders, empirical research conducted with dysphoric and depressed participants has contributed to our understanding of the motivational factors underlying ERS as it pertains to different clinical disorders. In the context of depression research, ERS has been defined as “the relatively stable tendency to excessively and persistently seek assurance from others that one is lovable and worthy, regardless of whether such assurance has already been provided” (Burns, Brown, Plant, Sachs-Ericsson, & Joiner, 2006, p. 136). Seeking of reassurance to reduce threats associated with potential loss and abandonment in depression is consistent with cognitive accounts of depression that focus on the role of loss and issues pertaining to worthiness as a motivation source for ERS (Beck, 1967, 1976). The psychometric operationalization of ERS in depression research has occurred exclusively with The Depressive Interpersonal Relationships Inventory—Reassurance Seeking subscale which consists of four items that assess an individual’s tendency to excessively seek reassurance that others truly care about him or her (Joiner & Metalsky, 2001). For example, respondents are asked, “Do you frequently seek reassurance from the people you feel close to as to whether they really care about you?” and “Do the people you feel close to sometimes become irritated with you for seeking reassurance from them about whether they really care about you?” (Joiner & Metalsky, 2001, p. 372). A recent meta-analysis (Starr & Davila, 2008) reported that ERS is associated with both concurrent depression and interpersonal rejection, but concluded, “Perhaps the most striking finding of this meta-analysis is the near complete lack of methodological diversity across studies. Research has virtually always relied on a single, four item self-report measure of excessive reassurance seeking. Although this scale has shown strong psychometric properties, the exclusivity of its use makes it impossible to explore whether other methods of assessing excessive reassurance seeking yield similar results (p. 773).” Given accumulating evidence from the anxiety literature that a wide variety of perceived threats can trigger ERS, the Reassurance Seeking subscale (Joiner & Metalsky, 2001) may have inadequate
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917
construct validity in the broader assessment of ERS due to its exclusive focus on perceived threats of social loss or rejection. Parrish and Radomsky (2010) compared participants with depression versus OCD and found that the primary type of perceived threat that triggered ERS differed between clinical groups, such that individuals with depression most frequently sought reassurance regarding perceived social threats whereas those with OCD most frequently sought reassurance regarding perceived general threats. The qualitative assessment of ERS beyond the sole use of the Reassurance Seeking subscale demonstrated the significance of ERS to anxiety disorders, moving beyond previous research that only used the Reassurance Seeking subscale and found that ERS was unrelated cross-sectionally and prospectively to the development of anxiety symptoms (Joiner & Metalsky, 2001; Joiner & Schmidt, 1998). Interestingly, Parrish and Radomsky (2010) found that although the source of threat differed between the depressed and OCD groups, the primary motivation for ERS was similar in both diagnostic groups: ERS episodes were precipitated by elevated anxiety and threat estimations, and anxiety reduction was cited most frequently as the primary function of ERS, thus highlighting the importance of the ERS construct in anxiety. 1.1. Study rationale Empirical and clinical anecdotal evidence converge in suggesting that excessive reassurance is commonly sought by individuals with anxiety disorders in response to a variety of perceived threats, which in turn, contributes to the long-term maintenance of anxiety and threat. The existing measures of reassurance seeking were not designed to assess diverse triggers for seeking reassurance in the anxiety disorders, and instead, assess reassurance only with a limited set of items pertaining to perceived social threats in the context of depression (Joiner & Metalsky, 2001). There have been recent attempts to measure reassurance seeking behaviors with a single (Muse, McManus, Hackman, Williams, & Williams, 2010; “I seek reassurance about my health”) or limited set of behavioral items in response to physical health threats (Speckens, Spinhoven, vanHemert, & Bolk, 2000) and worry and uncertainty (Comer et al., 2009; Gosselin et al., 2008). However, the content and construct validity of these scales as a measure of the motivations for seeking reassurance and their clinical usefulness to assess ERS in the broader anxiety disorder spectrum may be somewhat limited. Given accumulating evidence that ERS is a common mechanism across the anxiety disorders, the aim of the current study was to develop a comprehensive measure of ERS tapping a variety of reasons for seeking reassurance that would be transdiagnostic across the anxiety disorders, and potentially a variety of clinical and non-clinical populations. 2. Method 2.1. Participants Participants (N = 283) were continuous referrals to a large university-based anxiety disorders assessment and treatment clinic. The sample was comprised of individuals meeting DSM-IVTR (American Psychiatric Association, 2000) criteria for primary social phobia (n = 116), generalized anxiety disorder (n = 75), panic disorder with or without agoraphobia (n = 50), and obsessive compulsive disorder (n = 42). All diagnoses were based on the Structured Clinical Interview for Axis 1 Disorders (SCID-1/P version 2.0). Participants had a mean age of 35.21 years (SD = 10.68). The sample was 55.8% female, 58.7% single, predominantly Caucasian (83.7%), and well-educated (89.4% had at least some college/university education). The majority of the sample (61.5%) was either working or
913
attending school full-time. At the time of assessment, 11% of the sample had a secondary current MDD diagnosis.
2.2. Measures Beck Anxiety Inventory (BAI; Beck & Steer, 1993) – The BAI is a 21-item self-report measure of anxiety symptoms. Respondents are asked to report how much they have been bothered by several symptoms during the previous week on a Likert scale ranging from 0 (“not at all”) to 3 (“severely, I could barely stand it”). The BAI demonstrated adequate internal consistency (˛ = .92) in the current study. Beck Depression Inventory (BDI-II; Beck, Steer, & Brown, 1996) – The BDI-II is a 21-item self-report measure that assesses the severity of depression symptoms. Respondents are asked to select one of four statements, on a 0–3 scale, which best describes how they have been feeling over the past 2 weeks. The BDI-II demonstrated adequate internal consistency (˛ = .92) in the current study. Depression Anxiety Stress Scale (DASS; Lovibond & Lovibond, 1995) – The DASS is a 42-item self-report measure that assesses the severity of depression, anxiety, and stress during the previous week. Respondents are asked to rate the extent to which each item applied to them on a Likert scale ranging from 0 (“did not apply to me at all”) to 3 (“applied to me very much, or most of the time”). All three DASS subscales demonstrated adequate internal consistency (˛ = .90 to .96) in the current study, as did the total score (˛ = .96). Reassurance Seeking Scale (RSS) – The RSS is a self-report measure that was designed to assess reassurance seeking in a variety of situations. The item content for the RSS was generated based on the previous description of putative anxiety-related triggers for excessive reassurance seeking, and as culled from clinical notes from patients with anxiety disorders participating in psychological treatment. Respondents are asked to rate the frequency with which they seek reassurance on a Likert scale ranging from 1 (“not at all”) to 5 (“extremely”). The item content for the RSS was generated based on the previous description of the mechanisms associated with ERS in the anxiety disorders and as culled from clinical notes from patients participating in cognitive-behavioral therapy for their anxiety condition. The details of the scale are described further in Section 3.
2.3. Procedure This study received ethical approval from the institutional research ethics board. All participants completed the questionnaire packet at the time of their initial diagnostic assessment.
2.4. Statistical analysis We subjected the item pool to a principal axis factor analysis with an oblique rotation using Mplus 4.0 (Muthén & Muthén, 2006). Mplus was chosen over SPSS because of its ability to handle categorical data. As noted by Flora, Finkel, and Foshee (2003), when item responses are given using a Likert-type scale format, the observed responses are ordered, categorical manifestations of a continuous, psychological process of judgment about item content. Research has demonstrated extensively that exploratory factor analytic techniques that rely on normal-theory estimation using Pearson product–moment relations are not robust to such categorization of continuous variables (e.g., Bollen, 1989; Nunnally & Bernstein, 1994; West, Finch, & Curran, 1995). Mplus uses the matrix of polychoric correlations rather than the matrix of Pearson product–moment correlations for the analysis as this is considered the more appropriate choice for data that uses Likert scale ratings (Flora et al., 2003).
914
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917
3. Results
Table 1 Promax-rotated factor loadings.
3.1. Exploratory factor analysis
Item content
Using a sample of 283 participants, with all 40 items as variables, the analysis proceeded by first calculating the matrix of polychoric correlations among all possible pairs of items. We examined the polychoric correlation matrix to identify item pairs that were highly correlated and therefore possibly locally dependent. Results suggested the elimination of three items from the scale. Next the EFA was conducted with the polychoric correlations using ULS estimation. To assess number of factors underlying the set of test items, we used parallel analysis (Horn, 1965). Parallel analysis (O’Connor, 2000) for the 95th percentile in 1000 random data sets suggested a three-factor solution. However, research has suggested that parallel analysis is less reliable when using the polychoric correlation matrix. As such, we also examined a scree plot of the eigenvalues resulting from the exploratory factor analysis. The scree plot clearly showed that a large first factor might be adequate to explain the relationship among these items. However, the plot also suggested that a three-factor solution might underlie the data. Because we expected the factors to be correlated, we applied an oblique rotation, the Promax method, to the three-factor solution to aid interpretation. The one factor model demonstrated poor fit to the data based on the root mean square residual (RMSR = .09). Although there are no clear guidelines with respect to evaluation of fit indices, fit was considered good if the RMSR ≤ .05. The fit for the three-factor model was better (RMSR = .05). Through several iterations of retaining and deleting items on the basis of their polychoric inter-item correlations, factor loadings, communality estimates, and contribution to coefficient alpha, the total number of items was reduced from 40 to 30. Factor loadings for each of the 30 items are reported in Table 1. The final three-factor model of ERS demonstrated a good fit to the data (RMSR = .048). The first factor consisted of 13 items pertaining to seeking reassurance around making decisions and the factor was labeled: RSS-Decision-Making. For example, items include excessive reassurance seeking “Prior to making a decision,” “When you have to choose among alternative options,” or “When you doubt your decision.” The second factor consisted of eight items measuring reassurance seeking pertaining to social affiliation and rejection and the factor was subsequently labeled: RSS-Social Attachment. Examples of items include seeking excessive reassurance “To whether you are loved or cared,” “To get support from others” and “To get approval from others.” The third factor included nine items assessing seeking reassurance to reduce perceived general threat and coping with anxiety and was subsequently labeled, RSS-General Threat. Examples of items include seeking reassurance, “To prevent the occurrence of a catastrophic event,” “To whether something bad is going to happen to you” and “To feel more relaxed.” Correlations among the three factors were strong and positive (r’s ranging from .59 to .64). The three-factor based scales were then scored and reliability estimates were calculated. Alphas for the three scales ranged from .88 to .93, suggesting excellent internal consistency based on a minimal acceptable or lower bound standard of .70 to .79. Because coefficient alpha is influenced by the number of items on a scale, the average inter-item correlation was also examined to assess the internal consistency of the scales. Researchers have previously recommended that for assessing broad constructs, a minimum inter-item correlation of .15 to .20 is desired, whereas for a reliable and valid measure of a narrower construct, a higher inter-item correlation will be necessary (e.g., .40 to .45) (Clark & Watson, 1995). Average inter-item correlations for the three factor based scales were consistent with recommendations for assessment of narrower constructs, ranging from .45 to .52. A summary of the properties for
Prior to making a decision? When you have to choose among alternative options? When you doubt your decision? To whether you have considered all the possible details prior to making a decision? When you have to do something on your own? Before initiating or doing things? When you think you have made the wrong decision? To gain more certainty about a situation or something that is uncertain? Prior to making a change in some areas of your life (e.g., career, academic, relationships)? When you have a lot of responsibility about something? Before exploring something new? To avoid feeling responsible for the outcome of decisions in major areas of your life? To decrease your sense of personal responsibility? To whether you are loved or cared for? When you are not getting “enough attention”? To whether you are a lovable/caring person? To get support from others? To get approval from others? To feel close to others? To whether you have received a negative evaluation? To whether others are upset with you? To whether something bad is going to happen to you? To make sure you are okay? To prevent the occurrence of a catastrophic event? When you think a negative event is likely to occur? To whether you are safe? To feel more relaxed? To feel better inside? To turn off your anxiety feelings? To gain more peace and serenity within yourself?
DecisionMaking
Social Attachment
General Threat
.92 .91
−.02 −.21
−.07 .12
.84 .82
.06 −.09
−.06 .04
.75
−.01
.03
.71 .66
−.03 .09
.15 .10
.64
.12
.15
.60
.23
−.12
.55
.09
.21
.52 .50
.08 .17
.13 .06
.44
.23
.05
−.13 −.07
.89 .78
.01 −.03
−.06
.78
.07
.06 .25 −.13 .17
.72 .69 .65 .50
.00 −.16 .18 .11
.27 .03
.47 −.11
.05 .76
.07 −.13
−.06 −.02
.76 .77
.12
.03
.69
.02 .15 .01 .12 −.02
−.01 .03 .24 .15 .23
.66 .58 .56 .53 .52
Note. Factor loadings ≥.40 are in bold.
the 30-item scale with three-factor based subscales is presented in Table 2. 3.2. Scale descriptives and validity Scale descriptives for the measures used in the present study are presented in Table 3. To test the direction and magnitude of the association between RSS scores and measures of anxiety, stress, and depression, zero-order Pearson correlations were computed Table 2 Properties of the Reassurance Seeking Scale. Factor
Label
Total items
Range of loadings
Alpha
AIC
1 2 3
Decision-Making Social Attachment General Threat
13 8 9
.44 to .92 .47 to .89 .52 to .77
.93 .88 .88
.52 .48 .45
Note. AIC: average inter-item correlation. N = 283.
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917 Table 3 Scale descriptives. Scale Reassurance Seeking Scale Decision-Making Social Attachment General Threat Depression Anxiety Stress Scale Anxiety Stress Depression Beck Anxiety Inventory Beck Depression Inventory-II
M
SD
42.54 24.13 27.53
11.31 7.33 7.99
15.71 20.50 16.90 20.25 22.40
9.72 9.26 11.84 12.22 12.20
(Table 4). Scores on all three factors of the RSS were found to be significantly and positively correlated with scores on the DASS, BAI, and BDI-II. As an exploratory test of the convergent and divergent validity of the RSS, the correlations among the anxiety measures and the three RSS subscales were compared, as were the correlations between the depression measures and the three RSS subscales. Following the procedures set out by Meng, Rosenthal, and Rubin (1992), the correlation between the BAI and the RSS-General Threat subscale was significantly greater than between the BAI and the RSS-Decision-Making subscale (z = 2.95, p < .01) and between the DASS-Anxiety and the RSS-Social Attachment subscale (z = 3.95, p < .01). Similarly, the correlation between the DASS-Anxiety subscale and the RSS-General Threat subscale was significantly greater than between the DASS-Anxiety subscale and the RSS-DecisionMaking subscale (z = 2.00, p < .01) and between the DASS-Anxiety and the RSS-Social Attachment subscale (z = 3.24, p < .01). Finally, the correlation between the DASS-Anxiety subscale and the RSSGeneral Threat subscale was significantly greater than between the DASS-Depression subscale and the RSS-General Threat subscale (z = 2.90, p < .01). In contrast to the anxiety measures, the correlation between the depression measures (DASS-Depression and BDI) and the three RSS subscales were not significantly different. 4. Discussion The present study sought to develop and examine psychometric properties of a measure of reassurance seeking that assessed common triggers for patients with primary DSM-IV anxiety disorders. Items were theoretically derived and an exploratory factor analysis resulted in a three factor solution reflecting excessive reassurance seeking around indecisiveness and making decisions, social affiliation and fear of loss and rejection, and perceived general threat and the ability to cope with anxiety. The new measure, the Reassurance Seeking Scale, was found to possess good internal consistency,
915
and was positively and moderately correlated with measures of anxiety, stress, and depression. There was also some evidence of convergent validity for the new measure, with excessive reassurance pertaining to perceived general threat and anxiety being more strongly associated with anxiety symptom severity than depression symptom severity. White and Barlow (2002) describe safety behaviors broadly as actions that promote feelings of security that can be derived from the external and internal environment. Reassurance seeking can be conceptualized as a particular type of safety behavior aimed at reducing threat and restoring a sense of confidence and security through the repeated solicitation of safety-related information from others (Parrish & Radomsky, 2010; Woody & Rachman, 1994). Reassurance seeking as a safety behavior may emerge from both the overestimation of threat and danger in conjunction with an underestimation of the person’s perceived ability to cope with the threat and danger. The three RSS factors that resulted from the exploratory factor analysis closely reflect this operationalization of reassurance seeking in the anxiety disorders. Patients report engaging in excessive reassurance seeking when faced with indecision, doubt about the security of their relationships, and in anticipation of general threat. They also map on to the only study examining the content and function of reassurance seeking reported by clinical samples of individuals with OCD. The clinical group in the study by Parrish and Radomsky (2010) reported seeking reassurance regarding perceived threats to safety, perceived social threats (such as loss or rejection), and concerns about personal performance/competence (such as decision making). Results of this study yielding a factor characterizing reassurance seeking when feeling uncertain about decisions is consistent with the empirical literature supporting a link between anxiety disorders and indecisiveness. Individuals with DSM-IV diagnosed anxiety disorders score higher than non-clinical participants on facets of perfectionism related to concern over mistakes (e.g., “I should be upset if I make a mistake”) and doubts about actions (e.g., “I usually have doubts about the simple everyday things I do”) (Antony, Purdon, Huta, & Swinson, 1998; Purdon, Antony, & Swinson, 1999). The doubts about actions facet reflects the tendency to be uncertain about the correctness of one’s decisions or behaviors. These perceived threats may lead individuals with anxiety disorders to seek reassurance (i.e., safety) to ensure they do not make mistakes or poor decisions, particularly if they overestimate the consequences of making such an error. A recent study with nonclinical participants (Wu & Wei, 2008) found significant positive correlations between elevated scores on concerns over mistakes and doubts about actions and ERS. The second factor measuring social attachment and threat overlaps conceptually with the previous psychometric operationalization of ERS in depression research, which has occurred
Table 4 Correlation matrix. RSS Decisions RSS Decisions Attachment Threat DASS Anxiety Stress Depression BAI BDI
DASS Attach
Threat
.64 .63
– .59
–
.33 .42 .32 .28 .45
.31 .46 .40 .22 .40
Anxiety
Stress
Depress
– .66 .53 .83 .59
– .60 .51 .66
– .49 .78
BAI
BDI
– .61
–
–
.48 .43 .33 .42 .37
Note. RSS: Reassurance Seeking Scale (subscales: Decision-Making (Decisions), Social Attachment (Attach), General Threat (Threat)); DASS: Depression Anxiety Stress Scale (subscales: Depression, Anxiety, Stress); BAI: Beck Anxiety Inventory; BDI: Beck Depression Inventory-II. All correlations are significant, p < .001.
916
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917
exclusively with The Depressive Interpersonal Relationships Inventory—Reassurance Seeking subscale. This subscale consists of four items that assess an individual’s tendency to excessively seek reassurance that others truly care about him or her (Joiner & Metalsky, 2001). For example, respondents are asked, “Do you frequently seek reassurance from the people you feel close to as to whether they really care about you?” and “Do the people you feel close to sometimes become irritated with you for seeking reassurance from them about whether they really care about you?” (Joiner & Metalsky, 2001, p. 372). In the previous study by Parrish and Radomsky (2010), depressed patients were more likely to report ERS following attachment/evaluation triggers than were patients with OCD, but the latter group still reported attachment/evaluation triggers as a common reason for ERS. One possibility is that attachment/evaluation triggers increase the likelihood of ERS in those experiencing general negative affectivity, depression or anxiety, opposed to the specific form of negative affectivity (i.e., depression versus anxiety). A second possibility is that the heightened ERS following attachment/evaluation triggers in patients with a primary DSM-IV anxiety disorder, such as the current study, is due to the presence of secondary comorbid depression. However, in the current study only 7% of the sample had a secondary MDD diagnosis (or 11% if those in partial remission are included). Also, the correlation between the RSS-Social Attachment subscale and the BDI-II was significant but only accounted for 16% of the variance. Finally, there was uniformity in the pattern of associations between the BDI-II and all three RSS subscales. A third possibility is that ERS occurs in mood and anxiety disorders to the extent that there is a shared underlying vulnerability. It has been noted by others that the four-item ERS scale developed by Joiner and Metalsky contains items that are virtually identical to ones that measure anxious attachment (Shaver, Schachner, & Mikulincer, 2005). Research has demonstrated that anxious attachment is associated with depression (e.g., Bifulco, Moran, Ball, & Bernazzani, 2002) largely through the presence of self-criticism and dysfunctional beliefs (Shaver et al., 2005), factors that are not specific to depression. As such, it has been argued that ERS is more centrally related to anxious attachment. In a direct test of this hypothesis, Shaver et al. found a strong association between measures of anxious attachment and ERS, and found that a significant association between ERS and depression was attenuated to nonsignificance when anxious attachment was controlled, even though the association between anxious attachment and ERS remained when controlling for depression. Other research with non-clinical samples has similarly found a strong association between anxious attachment and ERS (Mikulincer, Gilliath, Sapir-Lavid, & Yaakobi, 2003). In summary, previous findings demonstrating an association between ERS pertaining to worries about rejection and abandonment and depression, and the results of the current study pointing to seeking reassurance around attachment/evaluation concerns, suggests possibility of a common shared vulnerability relating to anxious attachment and associated negative interpersonal beliefs as one common trigger for ERS in depression and anxiety disorders. One area for future investigation is the examination of distinction and overlap in the three RSS factors in depressed and anxious samples along with other putative measures of relationship and cognitive vulnerability. Rachman (2002) has hypothesized that requests for reassurance are most typically attempts to find safety from harm – to reduce psychological distress by reducing the perceived threat and/or increasing perceived coping resources. The third factor of the RSS (Threat/Anxiety) appears to be largely tapping the seeking of reassurance to find safety from looming general threat and to reduce feelings of anxiety. In the previous study by Parrish and Radomsky (2010), the primary reason why individuals with OCD reported they sought reassurance was to cope with elevated
anxiety and general threats, and this was more frequently a trigger for ERS in the OCD sample than it was for participants with major depression. Further, we found that measures of anxious arousal (i.e., BAI; DASS-A) were more strongly associated with the RSS-General Threat/Anxiety dimension than with the other two RSS dimensions, thus providing evidence for the specificity of this factor to anxious threat and somatic discomfort. These results are in keeping with cognitive-behavioral conceptualizations that emphasize the role of exaggerated threat appraisal in leading to the somatic and behavioral aspects of the anxiety cycle (e.g., Beck & Clark, 1997; Beck, Emery, & Greenberg, 1985; Ouimet, Gawronski, & Dozois, 2009). The findings of the current study regarding ERS in the anxiety disorders provide direction for clinicians to conceptualize both the general concerns that motivate such behavior in combination with a functional assessment of the disorder-specific triggers for ERS behaviors. Cognitive-behavioral treatment manuals with proven efficacy for the treatment of anxiety disorders often include explicit guidelines for the reduction of safety behaviors and reassurance – seeking in particular, although future research is required to test the short and long-term effects of reducing ERS directly. 4.1. Limitations At this preliminary stage, psychometric properties of the RSS appear promising. Although the sample size in the current study was considered sufficient for exploratory factor analysis (Floyd & Widaman, 1995; Gorsuch, 1983), some authors advocate for a subject to item ratio of 10:1 (Nunnally & Bernstein, 1994). Further validation studies with larger samples, including other anxiety disorders not sampled in the current study such as PTSD, are required to ensure the construct validity and factor stability of the scale. Retest-reliability studies are also required to assess the temporal stability of the RSS. One important direction for future research would be to use item response theory to test the relations between items and the construct of reassurance seeking, and to eliminate redundant or biased items (Embretson & Reise, 2000). Finally, treatment studies will determine the sensitivity of the measure in response to clinical gains. The RSS could be a useful clinical tool to assess changes in reassurance seeking over the course of treatment, and to examine whether reduction in reassurance seeking mediates reduction in anxiety. Acknowledgements The authors would like to thank Sam Hanig and Vincent Man for their editorial assistance in the preparation of this manuscript. References Abramowitz, J. S., & Moore, E. L. (2007). An experimental analysis of hypochondriasis. Behavior Research and Therapy, 45, 413–424. Abramowitz, J. S., Schwartz, S. A., & Whiteside, S. P. (2002). A contemporary conceptual model of hypochondriasis. Mayo Clinic Proceedings, 77, 1323–1330. Antony, M. M., Purdon, C. L., Huta, V., & Swinson, R. P. (1998). Dimensions of perfectionism across the anxiety disorders. Behavior Research and Therapy, 36, 1143–1154. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association. Beck, A. T. (1967). Depression: clinical, experimental, and theoretical aspects. New York: Harper and Row. Republished as: Beck, A. T. (1970). Depression: causes and treatment. Philadelphia: University of Pennsylvania Press Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: Meridian. Beck, A. T., & Clark, D. A. (1997). An information processing model of anxiety: automatic and strategic processes. Behavior Research and Therapy, 35, 49–58. Beck, A. T., & Steer, R. A. (1993). Manual for the beck anxiety inventory. San Antonio, TX: Psychological Corporation. Beck, A. T., Emery, G., & Greenberg, R. I. (1985). Anxiety disorders and phobias: a cognitive perspective. New York, NY: Basic Books.
N.A. Rector et al. / Journal of Anxiety Disorders 25 (2011) 911–917 Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for the beck depression inventory (2nd ed.). Boston: Harcourt, Brace, and Company. Bifulco, A., Moran, P., Ball, C., & Bernazzani, O. (2002). Adult attachment style I: is relationship to clinical depression. Social Psychiatry and Psychiatric Epidemiology, 37, 50–59. Bollen, K. A. (1989). Structural equations with latent variables. Wiley series in probability and mathematical statistics. New York: Wiley. Bowlby, J. (1973). Attachment and loss (volume II): separation: anxiety and anger. New York: Basic Books. Burns, A. B., Brown, J. S., Plant, E. A., Sachs-Ericsson, N., & Joiner, T. E., Jr. (2006). On the specific depressotypic nature of excessive reassurance-seeking. Personality and Individual Differences, 40, 135–145. Carter, M. M., Hollon, S. D., Carson, R., & Shelton, R. C. (1995). Effects of a safe person on induced distress following a biological challenge in panic disorder with agoraphobia. Journal of Abnormal Psychology, 104, 156–163. Clark, L. A., & Watson, D. (1995). Constructing validity: basic issues in objective scale development. Psychological Assessment, 7, 309–319. Comer, J. S., Roy, A. K., Furr, J. M., Gotimer, K., Beidas, R. S., Dugas, M. J., et al. (2009). The Intolerance of Uncertainty Scale for children: a psychometric evaluation. Psychological Assessment, 21, 402–411. Cuming, S., Rapee, R. M., Kemp, N., Abbott, M. J., Peters, L., & Gaston, J. E. (2009). A self-report measure of subtle avoidance and safety behaviors relevant to social anxiety: development and psychometric properties. Journal of Anxiety Disorders, 23, 879–883. Dugas, M., Gagnon, F., Ladouceur, R., & Freeston, M. (1998). Generalized anxiety disorder – a preliminary test of a conceptual model. Behavior Research and Therapy, 36, 215–226. Embretson, S. E., & Reise, S. P. (2000). Item response theory for psychologists. Mahwah, NJ: Lawrence Erlbaum. Flora, D. B., Finkel, E. J., & Foshee, V. A. (2003). Higher order factor structure of a self-control test: evidence from confirmatory factor analysis with polychoric correlations. Educational and Psychological Measurement, 63, 112–127. Floyd, F. J., & Widaman, K. F. (1995). Factor analysis in the development and refinement of clinical assessment instruments. Psychological Assessment, 7, 286– 299. Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Gosselin, P., Ladouceur, R., Evers, A., Laverdiere, A., Routhier, S., & Tremblay-Pickard, M. (2008). Evaluation of intolerance of uncertainty: development and validation of a new self-report measure. Journal of Anxiety Disorders, 22, 1427–1439. Heerey, E. A., & Kring, A. M. (2007). Interpersonal consequences of social anxiety. Journal of Abnormal Psychology, 116, 125–134. Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179–185. Joiner, T. E., Jr., & Metalsky, G. I. (2001). Excessive reassurance seeking: delineating a risk factor involved in the development of depressive symptoms. Psychological Science, 12, 371–378. Joiner, T. E., Jr., & Schmidt, N. B. (1998). Excessive reassurance-seeking predicts depressive but not anxious reactions to acute stress. Journal of Abnormal Psychology, 107, 533–537. Lohr, J. M., Olatunji, B. O., & Sawchuk, C. N. (2007). A functional analysis of danger and safety signals in anxiety disorders. Clinical Psychology Review, 27, 114–126. Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the Depression Anxiety Stress Scales (2nd ed.). Sydney: Psychology Foundation. Meng, X. L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated correlation coefficients. Psychological Bulletin, 111, 172–175. Mikulincer, M., Gilliath, O., Sapir-Lavid, Y., & Yaakobi, E. (2003). Attachment theory and concern for others’ welfare: evidence that activation of the sense of secure base promotes endorsement of self-transcendence values. Basic and Applied Social Psychology, 25, 299–312.
917
Muse, K., McManus, F., Hackman, A., Williams, M., & Williams, M. (2010). Intrusive imagery in sever health anxiety: prevalence, nature and links with memories and maintenance cycles. Behavior Research and Therapy, 48, 792–798. Muthén, L. K., & Muthén, B. O. (2006). Mplus user’s guide (4th ed.). Los Angeles: Muthén & Muthén. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill. O’Connor, B. P. (2000). SPSS and SAS programs for determining the number of components using parallel analysis and Velicer’s MAP test. Behavior Research Methods, Instrumentation, and Computers, 32, 396–402. Onur, E., Alkin, T., & Tural, U. (2007). Panic disorder symptoms: further clinical differences. Depression and Anxiety, 24, 479–486. Ouimet, A. J., Gawronski, B., & Dozois, D. J. A. (2009). Cognitive vulnerability to anxiety: a review and an integrative model. Clinical Psychology Review, 29, 459– 470. Parrish, C. L., & Radomsky, A. S. (2006). An experimental analysis of responsibility and reassurance: relationships with compulsive checking. International Journal of Behavioral Consultation and Therapy, 2, 174–191. Parrish, C. L., & Radomsky, A. S. (2010). Why do people seek reassurance and check repeatedly? An investigation of factors involved in compulsive behavior in OCD and depression. Journal of Anxiety Disorders, 24, 211–222. Purdon, C., Antony, M. M., & Swinson, R. P. (1999). Psychometric properties of the Frost Multidimensional Perfectionism Scale in a clinical anxiety disorders sample. Journal of Clinical Psychology, 55, 1271–1286. Rachman, S. (2002). A cognitive theory of compulsive checking. Behavior Research and Therapy, 40, 624–639. Salkovskis, P. M. (1985). Obsessional-compulsive problems: a cognitive-behavioral analysis. Behavior Research and Therapy, 23, 571–583. Salkovskis, P. M. (1991). The importance of behavior in the maintenance of anxiety and panic: a cognitive account. Behavioral Psychotherapy, 19, 6–19. Salkovskis, P. M. (1996). The cognitive approach to anxiety: threat beliefs, safety seeking behavior and the special case of health anxiety and obsessions. In: P. M. Salkovskis (Ed.), Frontiers of cognitive therapy (pp. 48–74). New York, NY: Guilford Press. Salkovskis, P. M., & Warwick, H. M. C. (1986). Morbid preoccupations, health anxiety, and reassurance: a cognitive behavioral approach to hypochondriasis. Behavior Research and Therapy, 24, 597–602. Shaver, P. R., Schachner, D. A., & Mikulincer, M. (2005). Attachment style, excessive reassurance seeking, relationship process, and depression. Personality and Social Psychology Bulletin, 31, 343–359. Simpson, J., & Weiner, E. (Eds.). (1989). Oxford english dictionary. United Kingdom: Oxford University Press. Speckens, A. E. M., Spinhoven, P., van Hemert, A. M., & Bolk, J. H. (2000). The Reassurance Questionnaire (RQ): psychometric properties of a self-report questionnaire. Psychological Medicine, 30, 841–847. Starr, L. R., & Davila, J. (2008). Excessive reassurance seeking, depression, and interpersonal rejection: a meta-analytic review. Clinical Psychology Review, 117, 762–775. Taylor, S., & Asmundson, G. J. G. (2004). Treating health anxiety: a cognitive behavioral approach. New York: Guilford Press. West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models with nonnormal variables. In: R. H. Hoyle (Ed.), Structural equation modeling: concepts, issues, and applications (pp. 56–75). Thousand Oaks, CA: Sage. White, K. S., & Barlow, D. H. (2002). Panic disorder and agoraphobia. In: D. H. Barlow (Ed.), Anxiety and its disorders: the nature and treatment of anxiety and panic. 2 (pp. 328–379). New York: Guilford Press. Woody, S., & Rachman, S. (1994). Generalized anxiety disorder as an unsuccessful search for safety. Clinical Psychology Review, 14, 743–753. Wu, T. F., & Wei, M. (2008). Perfectionism and negative mood: the mediating roles of validation from others versus self. Journal of Counseling Psychology, 55, 276–288.