Journal of Anxiety Disorders 24 (2010) 189195

Contents lists available at ScienceDirect

Journal of Anxiety Disorders

Its not just the judgementsIts that I dont know: Intolerance of uncertainty as a predictor of social anxiety
R. Nicholas Carleton, Kelsey C. Collimore, Gordon J.G. Asmundson *
Department of Psychology and the Anxiety and Illness Behaviours Laboratory, University of Regina, Regina, Saskatchewan S4S 0A2, Canada

A R T I C L E I N F O

A B S T R A C T

Article history: Received 1 June 2009 Received in revised form 8 October 2009 Accepted 19 October 2009 Keywords: Intolerance of uncertainty Social anxiety Social phobia Generalized anxiety disorder

Interest in the role of intolerance of uncertainty (IU) the tendency for a person to consider the possibility of a negative event occurring as unacceptable and threatening irrespective of the probability of its occurrence in anxiety disorders has been increasing in recent research. IU has been implicated as an important construct associated with generalized anxiety disorder (GAD); however, a growing body of research suggests that levels of IU are also high among individuals with other anxiety disorders. Despite the increasing interest, few studies have examined the relationship between IU and social anxiety (SA). The purpose of the present investigation was to further investigate the relationship between IU and SA. Participants included 286 community members (71% women) from Canada who completed measures of IU, SA, anxiety sensitivity, and fear of negative evaluation (FNE). Regression analyses revealed that the inhibitory anxiety dimension of IU, the fear of socially observable anxiety symptoms dimension of anxiety sensitivity, and the FNE were consistently signicant predictors of SA symptoms. Unexpectedly, IU and FNE were often comparable predictors of SA variance. Moreover, participants with SA symptoms consistent with SAD exhibited levels of IU comparable to those reported by participants with worry symptoms consistent with GAD. Comprehensive ndings, implications, and directions for future research are discussed. 2009 Elsevier Ltd. All rights reserved.

1. Introduction Social anxiety (SA) refers to anxiety or apprehension experienced in interpersonal or performance situations (Watson & Friend, 1969). Individuals with high SA fear being negatively evaluated by others (Rapee & Heimberg, 1997; Stein, Jang, & Livesley, 1999), making a bad impression, or acting in a way that might be embarrassing (Antony & Swinson, 2000). There is also evidence that SA may result from fearing positive evaluation, suggesting concern related to evaluation in general (Weeks, Heimberg, & Rodebaugh, 2008; Weeks, Heimberg, Rodebaugh, & Norton, 2008). Researchers have shown that social anxiety disorder (SAD) is related to and exacerbated by fears that other people can detect symptoms of SA (e.g., blushing; Rector, Szacun-Shimizu, & Leybman, 2007). Such fears are conceptualized within the anxiety sensitivity (AS; Peterson & Reiss, 1992) construct, which denotes the propensity to appraise anxiety-related somatic sensations, cognitive changes, and social consequences based on expectations of harmful consequences (Reiss & McNally, 1985; Taylor, 1999). Substantial research has demonstrated a direct relationship

* Corresponding author. Tel.: +1 306 347 2415; fax: +1 306 337 3275. E-mail address: gordon.asmundson@uregina.ca (Gordon J.G. Asmundson). 0887-6185/$ see front matter 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.janxdis.2009.10.007

between the social subscale of the Anxiety Sensitivity Index (ASI; Peterson & Reiss, 1992), and both SA and SAD (Anderson and Hope, 2009; Asmundson & Stein, 1994; Ball, Otto, Pollack, Uccello, & Rosenbaum, 1995) as well as indirect relationships with the ASI somatic and cognitive subscales (Carleton et al., 2009). The AS construct has been associated with intolerance of uncertainty (IU) the tendency for a person to consider the possibility of a negative event occurring as unacceptable and threatening irrespective of the probability of its occurrence (Carleton, Sharpe, & Asmundson, 2007) and may be causally dependent on IU (Carleton, Sharpe, et al., 2007). For persons who are intolerant of uncertainty, engaging in situations with uncertain outcomes may induce and perpetuate a heightened level of anxiety (Dugas, Gosselin, & Landouceur, 2001). People with high IU are more likely to interpret ambiguous information as threatening (Heydayati, Dugas, Buhr, & Francis, 2003), therein exacerbating their physiological arousal (Greco & Roger, 2001, 2003) which serves to facilitate self-perpetuating cycles of fear (Barlow, 2002). IU has been a useful construct in theory and research associated with generalized anxiety disorder (GAD) and obsessive compulsive disorder (Dugas et al., 2001; Holaway, Heimberg, & Coles, 2006); however, it is now also drawing attention from researchers investigating panic disorder (Dugas, Marchand, & Ladouceur, 2005; Simmons, Matthews, Paulus, & Stein, 2008). In contrast, there is a relative paucity of research directly investigating the relationship

190

R.N. Carleton et al. / Journal of Anxiety Disorders 24 (2010) 189195

between IU and SA, including SAD. Researchers have found that IU can predict changes in reported levels of SA (Riskind, Tzur, Williams, Mann, & Shahar, 2007); however, the ability to tolerate the uncertainty associated with social situations may be a critical element in determining SA. For example, in persons with SAD, uncertainty is often associated with SA before a social encounter (e.g., catastrophizing about possible occurrences), during the social encounter (e.g., catastrophizing about ambiguous stimuli), and/or after the social encounter (e.g., catastrophizing about possible consequences; Antony & Rowa, 2008). The IU and SA relationship has also been directly demonstrated as independent of the established relationships between SA, FNE, and AS using data from a sample of Netherlands participants (Boelen & Reijntjes, 2009). In that study, IU accounted for variance in SA symptoms beyond negative affect, FNE, and AS. While novel and highly coherent, Boelen and Reijntjess study included (1) only one measure of SA, (2) used the Intolerance of Uncertainty Scale aume, Letarte, Dugas, & Ladouceur, 1994), (IUS; Freeston, Rhe which can be psychometrically unstable (Carleton, Norton, & Asmundson, 2007), (3) did not include a measure of positive affectivity, and (4) used only the ASI total score. The primary goal of the current study was to replicate the previous nding of a relationship between IU and SA (Boelen & Reijntjes, 2009) with data from a North American community sample. A secondary goal was to extend those ndings to include measures of the various facets of SA (i.e., social interaction and performance anxiety, social distress and avoidance), negative and positive affect, as well as psychometrically stable measures of IU and AS. More specically, the dimensions of IU and AS were not assessed in the previous study despite evidence that the dimensions of each construct function in importantly distinct ways (Collimore, McCabe, Carleton, & Asmundson, 2008; Gosselin et al., 2008; Taylor, 1999). The nal goal of this study was to compare levels of IU across participants reporting symptoms congruent with SAD (without co-occurring GAD), relative to GAD (without comorbid SAD), or both (comorbid SAD and GAD), thereby paralleling previous comparative analyses of GAD and panic disorder with agoraphobia (Dugas et al., 2005). 2. Methods 2.1. Participants Participants included community members (n = 286) from Canada [82 men, 1854 years (Mage = 29.9; SD = 10.8) and 204 women, 1855 years (Mage = 29.8 SD = 10.7)], who completed several self-report measures as part of a larger investigation that was approved by the University Research Ethics Board. Participants were solicited with web-based advertising to participate in research exploring fear. Web-based data collection has been demonstrated to be a valid approach for questionnaire-based research in North America that is comparable to other data collection methods (Gosling, Vazire, Srivastava, & John, 2004) and has been used in related investigations of fear constructs (Carleton & Asmundson, 2009; Carleton, Norton, et al., 2007). Most participants (70%) reported having at least some postsecondary education, being employed or working at home (37% full-time, 19% part-time, and 7% as homemakers) and being either part or full time students (36%). Most participants identied their ethnicity as Caucasian (88%), First Nations (2%), or Asian (2%). Approximately half (52%) reported being single and 35% reported being married. 2.2. Measures Anxiety Sensitivity Index-3 (ASI-3; Taylor et al., 2007). The ASI-3 is an 18-item self-report measure assessing the tendency to fear

symptoms of anxiety based on the belief that they may have harmful consequences (e.g., It scares me when I blush in front of people). Items are rated on a 5-point Likert scale ranging from 0 (very little) to 4 (very much). Factor analyses supports a robust 3factor structure corresponding to the three dimensions of AS (fear of somatic sensations, somatic; fear of cognitive dyscontrol, cognitive, and fear of socially observable symptoms of anxiety, social; Taylor, Koch, Woody, & McLean, 1996; Zinbarg, Barlow, & Brown, 1997). The ASI-3 has demonstrated improved internal consistency and factorial validity relative to the original ASI (Peterson & Reiss, 1992). The ASI-3 has also demonstrated evidence for convergent, discriminant, and criterion validity (Taylor et al., 2007). Measures of AS demonstrate unique incremental validity beyond trait anxiety (Rapee & Medoro, 1994) and trait-level negative affectivity/ neuroticism (Zvolensky, Kotov, Antipova, Leen-Feldner, & Schmidt, 2005). The internal consistency was acceptable for the total score (a = .92), the somatic subscale score (a = .86), the cognitive subscale score (a = .89), and the social subscale score (a = .84). The average inter-item correlation was .40. Brief Fear of Negative Evaluation scale, version 2 (BFNE-II; Carleton, Collimore, & Asmundson, 2007; Carleton, McCreary, Norton, & Asmundson, 2006). The BFNE-II is a 12-item revised version of the Brief Fear of Negative Evaluation scale (BFNE; Leary, 1983) used for measuring fears of negative evaluation (e.g., I am afraid that others will not approve of me). The BFNE has been correlated with the Social Avoidance and Distress Scale (SADS; Watson & Friend, 1969); however, it more accurately depicts FNE than it does SA (Miller, 1995). Revisions to the BFNE were made in accordance with previously suggested changes (Taylor, 1993) to remove a methodological issue stemming from four reverseworded items by revising those items to be straightforwardly worded (Carleton, Collimore, et al., 2007; Carleton et al., 2006; Weeks et al., 2005). Items are rated on a 5-point Likert scale ranging from 0 (not at all characteristic of me) to 4 (extremely characteristic of me). The BFNE-II demonstrates excellent internal consistency, correlates highly with the BFNE, and factor analyses have supported a unitary solution (Carleton, Collimore, et al., 2007; Carleton et al., 2006). The internal consistency was acceptable (a = .98) and the average inter-item correlation was .76. Generalized Anxiety Disorder Assessment (GAD-7; Spitzer, Kroenke, Williams, & Lowe, 2006). The GAD-7 is a 7-item selfreport measure designed to assess GAD symptoms (e.g., Feeling nervous, anxious or on edge). Items are rated on a 4-point Likert scale ranging from 0 (not at all) to 3 (nearly every day). Use of the total 7-item score typically provides sufcient sensitivity and specicity for discerning clinical and nonclinical samples (i.e., a cut-off score of 10 with reports of symptoms interfering with daily activities can be used to distinguish persons reporting clinically signicant distress; Spitzer et al., 2006; Swinson, 2006). The internal consistency was acceptable (a = .89) and the average inter-item correlation was .54. Intolerance of Uncertainty Scale, Short Form (IUS-12; Carleton, Norton, et al., 2007). The IUS-12 is a 12-item short-form of the original 27-item Intolerance of Uncertainty Scale (Freeston et al., 1994) that measures reactions to uncertainty, ambiguous situations, and the future (e.g., Unforeseen events upset me greatly). Items are scored on a 5-point Likert scale ranging from 1 (not at all characteristic of me) to 5 (entirely characteristic of me). The IUS-12 has a strong correlation with the original scale, r = .96, and has been shown to have two factors, including prospective anxiety (7 items; e.g., I cant stand being taken by surprise) and inhibitory anxiety (5 items; e.g., When its time to act, uncertainty paralyses me), both with identically high internal consistencies, a = .85 (Carleton, Norton, et al., 2007). The internal consistency was acceptable for the total score (a = .92), the prospective anxiety

R.N. Carleton et al. / Journal of Anxiety Disorders 24 (2010) 189195

191

subscale score (a = .87) and the inhibitory anxiety subscale score (a = .90). The average inter-item correlation was .49. Social Avoidance and Distress Scale (SADS; Watson & Friend, 1969). The SADS is a 28-item self-report measure assessing the tendency to avoid or be distressed by social situations (e.g., I try to avoid situations which force me to be very sociable). The original scale presented respondents with true/false questions; however, in line with previous research (High & Caplan, 2009) we used a 5point Likert scale ranging from 1 (not at all characteristic of me) to 5 (entirely characteristic of me). The factor structure is generally considered to be unitary (Watson & Friend, 1969). The internal consistency was acceptable for the total score (a = .96) and the average inter-item correlation was .45. Social Interaction Phobia Scale (SIPS; Carleton et al., 2009). The SIPS is a 14-item self-report measure designed to assess symptoms specic to SAD (e.g., I have difculty talking with other people). Each item is measured on a 5-point Likert scale, ranging from 0 (not at all) to 4 (extremely). Respondents indicate how much each item bothered them during the past week. The items were derived as a subset of items from the Social Interaction Anxiety and Social Phobia Scales (Mattick & Clarke, 1998). The SIPS is designed to measure three symptom dimensions (i.e., social interaction anxiety, fear of overt evaluation, fear of attracting attention); however, use of the total score provides optimal sensitivity and specicity for discerning clinical and nonclinical samples (i.e., a cut-off score of 21 can typically be used to distinguish persons reporting clinically signicant social distress). The internal consistency was acceptable for the total score (a = .96) and the average inter-item correlation was .63. Positive and Negative Affect Schedule-Expanded Form (PANAS-X; Watson & Clark, 1994; Watson, Clark, & Harkness, 1994). The PANAS-X is a 60-item measure which assesses the extent to which respondents have experienced each of 60 words and phrases that describe different feelings and emotions within the past few weeks (e.g., cheerful or nervous). There is a strong convergence between trait and state indices of affect when using the PANAS-X, suggesting that the PANAS-X reects trait levels of affect (Watson & Clark, 1994). Each item is measured on a 5-point Likert scale, ranging from 0 (very slightly/not at all) to 4 (extremely). There is support for the construct validity of the PANAS-X scales, particularly the convergence of the negative affect and positive affect scales with measures of neuroticism and extraversion, respectively (for a review, see Watson, 2000). The internal consistency was acceptable for both the negative affect scale (a = .91) and the positive affect scale (a = .89). The average scale inter-item correlations were .50 for negative affect and .45 for the positive affect. 2.3. Analyses Response differences between women and men on each subscale were assessed with t-tests. Pearson correlations were calculated for each of the measured variables to provide indications of directions for subsequent regression analyses. Hierarchical regression analyses were performed with each of the SADS total score and the SIPS total score as dependent variables. Boelen and Reijntjes (2009) placed IU in the nal step of their hierarchical regression to evaluate whether it would account for any additional variance beyond neuroticism/negative affectivity, FNE, and AS (total score), and their results supported IU as accounting for signicant variance beyond these measures; however, IU, AS, and FNE have been posited to represent fundamental underlying cognitive constructs (Reiss & McNally, 1985; Taylor, 1999), with IU as a potentially fundamental cognitive construct underlying all anxiety disorders (Carleton, Sharpe, et al., 2007). In contrast, negative affect and positive affect have been

posited as overarching emotional traits and states (Watson, 2000; Watson & Clark, 1994; Watson et al., 1994). Two series of regression analyses were performed. First, the IUS total score was entered as block 1 of the independent variables, with the ASI-3 total score entered as block 2, the BFNE-II entered as block 3, and the negative affect and positive affect subscales of the PANAS-X entered as block 4. Second, the IUS-12 subscales were entered as block 1 of the independent variables, with the ASI-3 subscales entered as block 2, the BFNE-II entered as block 3, and the negative affect and positive affect subscales of the PANAS-X entered as block 4. In this fashion, the regression model tested the variance accounted for by each construct overall and each of the construct dimensions in a linear fashion congruent with existing theory. To be thorough, model results from the analyses wherein the independent variables were entered in reverse order are also presented in parentheses to delineate the variance IU predicts beyond negative affect, positive affect, FNE, and the ASI-3 subscales. The last analyses required that participants from the sample be selected based on whether they reported symptoms consistent with a probable diagnosis of SAD or GAD, or both. Participants were selected using very conservative cutoff scores as recommended for the SIPS (i.e., a cutoff score of 30; Carleton et al., 2009) and the GAD-7 (i.e., a cut-off score of 15; Spitzer et al., 2006). This resulted in four groups of participants who reported (1) SA symptoms consistent with people diagnosed with SAD (without clinically signicant GAD symptoms), (2) GAD symptoms consistent with people diagnosed with GAD (without clinically signicant SA symptoms), (3) SA and GAD symptoms consistent with people diagnosed with both SAD and GAD, and (4) SA and GAD symptom levels lower than would be consistent with people diagnosed with SAD or GAD. Thereafter, IU total and subscale scores were compared across the four groups using analysis of variance post hoc comparisons to determine whether (ANOVA) and Scheffe differences existed in IU levels between persons with clinically signicant SAD, GAD, both disorders, or neither disorder. 3. Results 3.1. Descriptive statistics Descriptive statistics for each dependent variable are presented in Table 1. None of the indices of univariate skewness and kurtosis in the clinical sample were sufciently out of range to preclude the planned analyses (Curran, West, & Finch, 1996; Tabachnick & Fidell, 2001). Men and women were comparable on most subscales (ps > .05); however, men reported higher scores on the PANAS-X positive affect scale, t(284) = 2.65, p < .01, r2 = .02, while women reported higher scores on the GAD-7, t(284) = 2.14, p < .05, r2 = .02. In both cases the effect sizes were quite small (Cohen, 1988); accordingly, men and women were not analyzed separately. The correlation analyses suggested small to large relationships between each of the variables (Table 1). 3.2. Regression analyses The regression results indicated a robust relationship between the IUS-12 total score and the SIPS total score, independent of all other independent variables and similar to the precedent results from Boelen and Reijntjes (Table 2). The regression results also indicated a robust relationship between the inhibitory anxiety subscale of the IUS-12 and the SIPS total score, independent of all other independent variables and comparable with the ASI-3 social subscale and the BFNE-II (Table 2). There were no indications of problems with multicollinearity (i.e., all tolerances > .30 and all variance ination factors < 4.00), problems

192 Table 1 Descriptive statistics.

R.N. Carleton et al. / Journal of Anxiety Disorders 24 (2010) 189195

Correlations M (SD) 1. IUS-12 total score 2. IUS-12 prospective anxiety subscale 3. IUS-12 inhibitory anxiety subscale 4. ASI-3 total score 5. ASI-3 somatic subscale 6. ASI-3 cognitive subscale 7. ASI-3 social subscale 8. BFNE-II 9. PANAS-X negative affect 10. PANAS-X positive affect 11. SIPS total score 12. SADS total score 30.69 (11.07) 19.45 (6.58) S (.14) .50 .25 K (.29) .46 .71 1 1.00 .94 2 3 4 5 6 7 8 9 10 11

11.24 19.14 5.39 4.72 9.03 23.77 23.60 27.77 18.00 52.84

(5.32) (14.38) (5.33) (5.41) (6.20) (14.98) (8.93) (8.22) (14.56) (24.76)

.73 .94 1.04 1.30 .57 .03 .60 .14 .82 .14

.41 .26 .31 .87 .66 1.29 .32 .69 .17 .77

.91 .71 .50 .59 .69 .59 .60 .29 .69 .54

.73 .60 .43 .50 .59 .50 .52 .26 .59 .48

.72 .52 .61 .71 .61 .61 .29 .71 .54

.83 .86 .85 .65 .58 .26 .68 .44

.62 .53 .43 .41 .20 .45 .23

.59 .49 .50 .22 .54 .32

.71 .55 .23 .72 .54

.49 .28 .68 .49

.25 .54 .35

.28 .36

.78

NoteS: Skew (standard error); K: kurtosis (standard error). All correlations were statistically signicant (p < .01); IUS-12: intolerance of uncertainy-12; ASI-3: anxiety sensitivity index-3; BFNE-II: brief fear of negative evaluation scale-II; PANAS-X: positive and negative affect scale; SIPS: social interaction phobia scale; SADS: social avoidance and distress scale.

with outliers, problems with normality, or problems with homoscedasticity (Tabachnick & Fidell, 2001). The inhibitory anxiety subscale of the IUS-12 accounted for half of the variance in SIPS scores. Notably, neither negative affect nor positive affect were statistically signicant predictors of SIPS scores beyond IU, AS, and FNE. When the order of entry for the independent variables was reversed, negative affect and positive affect accounted for a third of the variance in SIPS scores (Table 2); however, IU, AS, and FNE each continued to contribute statistically signicant amounts to the variance accounted for. To further explore the association with IU, a subsequent regression was performed (details not shown) wherein the BFNE-II was entered as block 1, with the IUS12 subscales in block 2, the ASI-3 subscales in block 3, and negative and positive affect in block 4. Results indicated that the BFNE-II accounted for 47% of the variance in SIPS score, whereas in the previous analysis the inhibitory anxiety subscale of the IUS-12 accounted for 51% of the variance in the total SIPS score. The regression results also indicated a robust relationship between the IUS-12 total and the SADS total score, independent of all other independent variables and similar to the precedent results from Boelen and Reijntjes (Table 3). The regression results also
Table 2 Regression results for SIPS. Model step Constant Coefcient statistics

indicated a robust relationship between the inhibitory anxiety subscale of the IUS-12 and the SADS total score, independent of all other independent variables and comparable with the ASI-3 social subscale and the BFNE-II (Table 3). There were no indications of problems with multicollinearity (i.e., all tolerances > .30 and all variance ination factors < 4.00), problems with outliers, problems with normality, or problems with homoscedasticity (Tabachnick & Fidell, 2001). In this case, the inhibitory anxiety subscale of the IUS-12 accounted for a third of the variance in SADS scores, which was less than with the SIPS, but still the largest portion for the SADS. For the SADS, the absence of positive affect (but not negative affect) was a statistically signicant predictor of SADS scores beyond IU, AS, and FNE. When the order of entry for the independent variables was reversed, negative affect and positive affect accounted for a fth of the variance in SADS scores (Table 2); however, IU, AS, and FNE each continued to contribute statistically signicant amounts to the variance accounted for. To further explore the association with IU, a subsequent regression was performed (details not shown) wherein the BFNE-II was entered as block 1, with the IUS-12 subscales in block 2, the ASI-3 subscales in block 3, and negative and positive affect in block 4.

Correlations p <.01 <.01 <.01 <.01 .12 .39 r Part r

Model step statistics (IVs in reversed order)

b
Total scores 1 (4) 2 (3) 3 (2) 4 (1) IUS-12 total score ASI-3 total score BFNE-II PANAS-X negative affect PANAS-X positive affect .39 .21 .32 .12 .06

t 2.23 5.26 3.48 6.44 1.56 .86 1.56 1.28 3.93 4.10 .46 .98 4.79 1.25 1.05

DR2

DF

.69 .68 .68 .54 .28

.19 .13 .24 .06 .03

.48 (.04) .07 (.05) .06 (.21) <.01 (.32)

260.47 (27.68) 45.36 (36.42) 46.16 (124.51) 1.69 (65.67)

<.01 (<.01) <.01 (<.01) <.01 (<.01) .19 (<.01)

Subscale scores 1 (4) IUS-12 prospective anxiety subscale IUS-12 inhibitory anxiety subscale ASI-3 social subscale ASI-3 somatic subscale ASI-3 cognitive subscale BFNE-II PANAS-X negative affect PANAS-X positive affect .07 .26 .25 .02 .05 .26 .06 .04 .12 .20 <.01 <.01 .65 .33 <.01 .21 .30 .59 .71 .72 .45 .54 .68 .54 .28 .05 .14 .15 .02 .04 .17 .05 .04 .51 (.04) 149.70 (13.66) <.01 (<.01)

2 (3)

.09 (.08)

20.51 (18.19)

<.01 (<.01)

3 (2) 4 (1)

.03 (.21) <.01 (.32)

25.97 (124.51) 1.42 (65.67)

<.01 (<.01) .24 (<.01)

Note: IUS-12: intolerance of uncertainy-12; ASI-3: anxiety sensitivity index-3; BFNE-II: brief fear of negative evaluation scale-II; PANAS-X: positive and negative affect scale; SIPS: social interaction phobia scale.

R.N. Carleton et al. / Journal of Anxiety Disorders 24 (2010) 189195 Table 3 Regression results for SADS. Model step Constant Coefcient statistics Correlations p 5.54 5.05 .18 3.65 .47 3.90 <.01 <.01 .86 <.01 .64 <.01 r Part r Model step statistics (IVs in reversed order)

193

b
Total scores 1 (4) 2 (3) 3 (2) 4 (1) IUS-12 total score ASI-3 total score BFNE-II PANAS-X negative affect PANAS-X positive affect .82 .02 .39 .08 .59

DR2

DF

.54 .44 .49 .35 .36

.24 .01 .17 .02 .19

.29 (.06) .01 (.01) .04 (.10) .03 (.20)

117.49 (25.53) 2.29 (4.36) 15.93 (40.44) 7.62 (35.80)

<.01 (<.01) .13 (<.05) <.01 (<.01) .19 (<.01)

Subscale scores 1 (4) IUS-12 prospective anxiety subscale IUS-12 inhibitory anxiety subscale ASI-3 social subscale ASI-3 somatic subscale ASI-3 cognitive subscale BFNE-II PANAS-X negative affect PANAS-X positive affect .13 .24 .29 .15 .06 .13 .04 .21 6.17 1.88 2.86 3.71 2.39 .87 1.86 .74 4.25 <.01 .06 <.01 <.01 .02 .38 .06 .46 <.01 .48 .54 .54 .23 .32 .49 .35 .36 .09 .13 .17 .11 .04 .09 .03 .20 .30 (.04) 61.65 (10.47) <.01 (<.01)

2 (3)

.06 (.07)

8.84 (9.99)

<.01 (<.01)

3 (2) 4 (1)

.01 (.10) .04 (.20)

5.19 (40.44) 9.15 (35.80)

.02 (<.01) <.01 (<.01)

Note: IUS-12: intolerance of uncertainy-12; ASI-3: anxiety sensitivity index-3; BFNE-II: brief fear of negative evaluation scale-II; PANAS-X: positive and negative affect scale; SADS: social avoidance and distress scale.

Results indicated that the BFNE-II accounted for 24% of the variance in SADS score, whereas in the previous analysis the inhibitory anxiety subscale of the IUS-12 accounted for 30% of the variance (see Table 3). 3.3. Group comparisons The results of the ANOVA indicated statistically signicant differences based on symptom groups for the IUS-12 total score, F(3, 282) = 61.46, p < .001, eta2 = .40, the IUS-12 prospective anxiety subscale score, F(3, 282) = 34.81, p < .001, eta2 = .27, and the IUS-12 inhibitory anxiety subscale score F(3, 282) = 75.55, p < .001, eta2 = .45. The substantially different sample sizes within each of the symptoms groups, while not prohibitive for ANOVA, does make meeting the assumption of homogeneity of variance more important (Tabachnick & Fidell, 2001); however, even if violated, a correction can be made using appropriately discriminating post hoc tests and an increasingly stringent alpha (Judd, McClelland, & Culhane, 1995; Tabachnick & Fidell, 2001). The assumption of homogeneity was met for the IUS-12 total score (p > .10) and prospective anxiety subscale (p > .10). In contrast, there was a slight variance for the IUS-12 inhibitory anxiety subscale (p = .03). Nevertheless, use of Scheffe post hoc tests and the large effect sizes suggest that the differences are likely to be robust (Table 4). Overall, people with symptoms consistent with both GAD and SAD reported higher IU levels than people with symptoms consistent with either SAD or GAD alone (who reported comparable IU levels), who in turn, reported higher IU levels than people who reported symptoms that were consistent with neither SAD nor GAD.

4. Discussion The current study had three goals. The rst was to replicate the previously demonstrated relationship between IU and SA (Boelen & Reijntjes, 2009) with data from a North American community sample. The second goal was to extend those ndings to include measures of social interaction and performance anxiety, social avoidance and distress, negative and positive affect, as well as psychometrically stable measures and subscales of IU and AS. The third and nal goal paralleled previous research comparing IU levels across GAD and panic disorder with agoraphobia samples (Dugas et al., 2005) by comparing levels of IU across participants reporting symptoms consistent with probable diagnoses of SAD, GAD, both disorders, or neither disorder. The results of the correlation analyses demonstrated signicant interrelationships between all of the variables of interest and in theoretically congruent directions (e.g., positive affect was negatively correlated with SA, whereas FNE and SA were positively correlated). These results are in accordance with growing research indicating a relationship between SA, FNE (Weeks et al., 2005), AS (Orsillo, Lilienfeld, & Heimberg, 1994), negative affect/ neuroticism (Barlow, 2002), and IU (Boelen & Reijntjes, 2009). The current and more specic analysis demonstrated that within AS and IU, the subscales are differentially associated with SA, whether measured as a function of social interaction and performance anxiety or social avoidance and distress. The rst series of regression results supported a robust relationship between IU and SA as measured by the SIPS and the SADS, independent of all other variables and similar to the precedent

Table 4 post hoc comparisons. ANOVA Scheffe IUS-12 total score N Neither SAD GAD Both 206 39 17 24 M (SD) 26.62a (8.32) 37.97b (10.23) 38.76b (9.83) 48.04c (7.82) Prospective anxiety subscale M (SD) 17.50a (5.60) 22.67b (6.25) 23.18b (5.77) 28.38c (5.01) Inhibitory anxiety subscale M (SD) 9.13a (3.74) 15.31b (4.82) 15.59b (4.93) 19.67c (3.74)

post hoc comparisons (p < .05); SAD: social anxiety disorder; Note: Means with different subscripts in the same column are statistically signicantly different with Scheffe GAD: generalized anxiety disorder.

194

R.N. Carleton et al. / Journal of Anxiety Disorders 24 (2010) 189195

results from Boelen and Reijntjes (2009). The second series of hierarchical regression analyses predicting either social interaction and performance anxiety, or social avoidance and distress, provided separate quantications of the differential associations demonstrated in the correlation analysis. IU inhibitory anxiety accounted for more than half (51%) of the variance in social interaction and performance anxiety, with the ASI-3 social subscale and the BFNE-II accounting for statistically signicant, but smaller portions of variance (i.e., 9% and 3%, respectively). In this case, negative affect and positive affect were not statistically signicant predictors of social interaction and performance anxiety, together accounting for less than 1% of the variance. When the order was reversed and negative affect and positive affect were entered rst, the constructs accounted for a statistically signicant amount of the variance (32%), which was much less than IU accounted for when it was entered rst. Moreover, in the reverse order analysis IU continued to account for a statistically signicant amount of the variance (4%) above and beyond negative affect, positive affect, FNE, and the AS fear of socially observable anxiety symptoms. IU inhibitory anxiety accounted for a third (30%) of the variance in social avoidance and distress; however, this was notably less than the variance accounted for in social interaction and performance anxiety. The ASI-3 social subscale and BFNE-II again accounted for statistically signicant, but smaller, portions of variance (6% and 1%, respectively). In this case reduced positive affect but not negative affect accounted for an additional statistically signicant portion of variance (4%). This suggests that the absence of positive affect (Watson, 2000; Watson & Clark, 1994; Watson et al., 1994) may be related to social avoidance and distress. When the order was reversed and negative affect and positive affect were entered rst, the constructs accounted for a statistically signicant amount of the variance (20%) that was less than IU accounted for when it was entered rst. Moreover, IU again accounted for a statistically signicant amount of the variance (4%) above and beyond negative affect, positive affect, FNE, and the AS fear of socially observable anxiety symptoms. Overall, the results of the regression analyses support a robust, independent relationship between SA and inhibitory anxiety associated with IU. There was also evidence of a robust relationship between SA and the fear of socially observable anxiety symptoms, as well as SA and fears of being negatively evaluated. There was no evidence of a relationship between SA and worrying about future uncertainty (i.e., prospective anxiety), fears of somatic sensations, or fears of cognitive dyscontrol. In contrast to previous research (Boelen & Reijntjes, 2009), the relationship between negative affect and SA was equivocal at best because, although SA may result in negative affect or the absence of positive affect, neither negative affect nor positive affect predicted SA beyond cognitive constructs such as IU, or ASI-3 social, or FNE in this sample. The results suggest that the initially posited, tested, and theoretically congruent (Carleton, Sharpe, et al., 2007; Taylor, 1999; Watson, 2000) hierarchical ordering with IU as underlying and negative affect and positive affect as overarching constructs better reects the social interaction and performance anxiety data from this sample; however, the directional inuence of negative and positive affect on social avoidance and distress remains unknown. Positive affect may serve a protective function against SA; alternatively, lower fears of the uncertainty of social situations, socially observable anxiety symptoms, and fears of negative evaluation may facilitate the development of positive affect. The comparisons across the symptom groups suggest that there are differences in IU levels between persons with a probable diagnosis of SAD, GAD, neither, or both disorders. People reporting SAD and GAD symptoms well below levels reported by people meeting diagnostic criteria for either reported IU levels statistically signicantly lower than persons reporting symptoms congruent

with diagnoses of either or both SAD and GAD. People reporting symptoms comparable to those reported by people diagnosed with either SAD or GAD reported comparable levels of IU. This suggests that differences wherein people with GAD reported higher IU levels than people with panic disorder with agoraphobia (Dugas et al., 2005) may not be pervasive across all anxiety disorders. People reporting symptoms consistent with both SAD and GAD reported statistically signicantly higher levels of IU than all other groups. These results contribute to a growing body of literature suggesting that IU may be a fundamental component of several anxiety disorders, and further evidence that IU may vary across anxiety disorders, being more pertinent for some anxiety disorders relative to others. There are several limitations to the current study that provide directions for future research. First, there was no diagnostic information available for the current sample. There may be important differences associated with people who have been formally diagnosed with SAD or GAD and their responses to the variables measured in the current study. Future researchers should explore these variables across diagnosed clinical samples, particularly persons with SAD. Moreover, future research should consider including a measure of IU as a treatment outcome measure among persons treated for SAD to determine whether reductions in SA symptoms correspond with reductions in IU levels. Alternatively, researchers might explore whether targeted reductions in IU levels result in reductions in SA symptoms, even in the absence of treatments targeting SA. Second, the structure of the current regression analyses suggests a linear hierarchical relationship between IU, AS, FNE, negative and positive affect, and SA that remains untested with longitudinal data. Although current theory supports cognitive constructs as determinant bases for affective constructs (Barlow, 2002), future researchers might attempt to complete prospective studies to better determine causal relationships, even if those relationships are reciprocal. Third, the Social Phobia Inventory (SPIN; Connor et al., 2000) used previously (Boelen & Reijntjes, 2009) was not available in the current data. Accordingly, it is possible that neuroticism is a signicant component of a dimension of SA measured by the SPIN but not the SIPS or the SADS. It is also possible that negative affect and neuroticism are sufciently differentor that the measure of neuroticism used in the previous study (i.e., the Shortened Eysenck Personality Questionnaire Neuroticism subscale; Eysenck, Eysenck, & Barrett, 1985) and the PANAS-X are sufciently different that the relationship was not replicable with the current data. Based on previous research with the PANAS-X (Watson, 2000), this possibility is unlikely; however, future researchers might test this possibility before ruling out a robust independent relationship between negative affect, neuroticism, and SA. The results of this study support a robust relationship between IU and SA that is independent of AS, FNE, negative affect, and positive affect. Indeed, the inability to tolerate the uncertainty associated with social situations may be a critical element in the development and maintenance of SAD. Although yet untested, treatments which focus on increasing tolerance for the uncertainty inherent in social situations may provide help in relieving SAD symptoms. Acknowledgements K.C. Collimore is supported by a Canadian Institutes of Health Research doctoral grant (FRN # 85321). References
Anderson, E. R., & Hope, D. A. (2009). The relationship among social phobia, objective and perceived physiological reactivity, and anxiety sensitivity in an adolescent population. Journal of Anxiety Disorders, 23, 1826.

R.N. Carleton et al. / Journal of Anxiety Disorders 24 (2010) 189195 Antony, M. M, & Rowa, K. (2008). Social anxiety disorder. Ashland, OH: Hogrefe & Huber Publishers. Antony, M. M., & Swinson, R. P. (2000). The shyness and social anxiety workbook: proven techniques for overcoming your fears. Oakland, CA: New Harbinger Publications. Asmundson, G. J. G., & Stein, M. B. (1994). Selective processing of social threat in patients with generalized social phobia: evaluation using a dot-probe paradigm. Journal of Anxiety Disorders, 8, 107117. Ball, S. G., Otto, M. W., Pollack, M. H., Uccello, R., & Rosenbaum, J. F. (1995). Differentiating social phobia and panic disorder: a test of core beliefs. Cognitive Therapy and Research, 19, 473481. Barlow, D. H. (2002). Anxiety and its disorders (2nd ed.). New York, NY: Guilford Press. Boelen, P. A., & Reijntjes, A. (2009). Intolerance of uncertainty and social anxiety. Journal of Anxiety Disorders, 23, 130135. Carleton, R. N., & Asmundson, G. J. G. (2009). The multidimensionality of fear of pain: construct independence for the fear of pain questionnaire-short form and the pain anxiety symptoms scale-20. The Journal of Pain, 10, 2937. Carleton, R. N., Collimore, K. C., & Asmundson, G. J. G. (2007). Social anxiety and fear of negative evaluation: construct validity of the BFNE-II. Journal of Anxiety Disorders, 21, 131141. Carleton, R. N., Collimore, K. C., Asmundson, G. J., McCabe, R. E., Rowa, K., & Antony, M. M. (2009). Rening and validating the social interaction anxiety scale and the social phobia scale. Depression and Anxiety, 26, E71E81. Carleton, R. N., McCreary, D. R., Norton, P. J., & Asmundson, G. J. G. (2006). Brief fear of negative evaluation scale-revised. Depression and Anxiety, 23, 297303. Carleton, R. N., Norton, M. A., & Asmundson, G. J. G. (2007b). Fearing the unknown: a short version of the intolerance of uncertainty scale. Journal of Anxiety Disorders, 21, 105117. Carleton, R. N., Sharpe, D., & Asmundson, G. J. G. (2007c). Anxiety sensitivity and intolerance of uncertainty: requisites of the fundamental fears? Behaviour Research and Therapy, 45, 23072316. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Mawah, NJ: Erlbaum. Collimore, K. C., McCabe, R. E., Carleton, R. N., & Asmundson, G. J. G. (2008). Media exposure and dimensions of anxiety sensitivity: differential associations with PTSD symptom clusters. Journal of Anxiety Disorders, 22, 10211028. Connor, K. M., Davidson, J. R. T., Churchill, L. E., Sherwood, A., Foa, E., & Weisler, R. H. (2000). Psychometric properties of the social phobia inventory (SPIN): new selfrating scale. British Journal of Psychiatry, 176, 379386. Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of test statistics to nonnormality and specication error in conrmatory factor analysis. Psychological Methods, 1, 1629. Dugas, M. J., Gosselin, P., & Landouceur, R. (2001). Intolerance of uncertainty and worry: investigating specicity in a nonclinical sample. Cognitive Therapy and Research, 25, 551558. Dugas, M. J., Marchand, A., & Ladouceur, R. (2005). Further validation of a cognitivebehavioral model of generalized anxiety disorder: diagnostic and symptom specicity. Journal of Anxiety Disorders, 19, 329343. Eysenck, S. B., Eysenck, H. J., & Barrett, P. (1985). A revised version of the psychoticism scale. Personality and Individual Differences, 6, 2129. aume, J., Letarte, H., Dugas, M. J., & Ladouceur, R. (1994). Why do Freeston, M., Rhe people worry? Personality and Individual Differences, 17, 791802. Gosling, S. D., Vazire, S., Srivastava, S., & John, O. P. (2004). Should we trust web-based studies? A comparative analysis of six preconceptions about internet questionnaires. American Psychologist, 59, 93104. Gosselin, P., Ladouceur, R., Evers, A., Laverdiere, A., Routhier, S., & Tremblay-Picard, M. (2008). Evaluation of intolerance of uncertainty: development and validation of a new self-report measure. Journal of Anxiety Disorders, 22, 14271439. Greco, V., & Roger, D. (2001). Coping with uncertainty: the construction and validation of a new measure. Personality and Individual Differences, 31, 519534. Greco, V., & Roger, D. (2003). Uncertainty, stress and health. Personality and Individual Differences, 34, 10571068. Heydayati, M., Dugas, M. J., Buhr, K., & Francis, K. (2003). The relationship between intolerance of uncertainty and the interpretation of ambiguous and unambiguous information. Paper presented at the annual convention of the association for advancement of behaviour therapy. High, A. C., & Caplan, S. E. (2009). Social anxiety and computer-mediated communication during initial interactions: implications for the hyperpersonal perspective. Computers in Human Behavior, 25, 475482. Holaway, R. M., Heimberg, R. G., & Coles, M. E. (2006). A comparison of intolerance of uncertainty in analogue obsessive-compulsive disorder and generalized anxiety disorder. Journal of Anxiety Disorders, 20, 158174. Judd, C. M., McClelland, G. H., & Culhane, S. E. (1995). Data analysis: continuing issues in the everyday analysis of psychological data. Annual Review of Psychology, 46, 433 465.

195

Leary, M. R. (1983). A brief version of the Fear of Negative Evaluation Scale. Personality and Social Psychology Bulletin, 9, 371375. Mattick, R. P., & Clarke, J. C. (1998). Development and validation of measures of social phobia scrutiny fear and social interaction anxiety. Behaviour Research and Therapy, 36, 455470. Miller, M. B. (1995). Coefcient alpha: a basic introduction from the perspectives of classical test theory and structural equation modeling. Structural Equation Modeling, 2, 255273. Orsillo, S. M., Lilienfeld, S. O., & Heimberg, R. G. (1994). Social phobia and response to challenge procedures: examining the interaction between anxiety sensitivity and trait anxiety. Journal of Anxiety Disorders, 8, 247258. Peterson, R. A., & Reiss, S. (1992). Anxiety sensitivity index manual (2nd ed.). Worthington, OH: International Diagnostic Systems. Rapee, R. M., & Heimberg, R. G. (1997). A cognitive-behavioral model of anxiety in social phobia. Behaviour Research and Therapy, 35, 741756. Rapee, R. M., & Medoro, L. (1994). Fear of physical sensations and trait anxiety as mediators of the response to hyperventilation in nonclinical subjects. Journal of Abnormal Psychology, 103, 693699. Rector, N. A., Szacun-Shimizu, K., & Leybman, M. (2007). Anxiety sensitivity within the anxiety disorders: disorder-specic sensitivities and depression comorbidity. Behaviour Research and Therapy, 45, 19671975. Reiss, S., & McNally, R. J. (1985). The expectancy model of fear. In: S. Reiss & R. R. Bootzin (Eds.), Theoretical issues in behaviour therapy (pp. 107121). New York, NY: Academic Press. Riskind, J. H., Tzur, D., Williams, N. L., Mann, B., & Shahar, G. (2007). Short-term predictive effects of the looming cognitive style on anxiety disorder symptoms under restrictive methodological conditions. Behaviour Research and Therapy, 45, 17651777. Simmons, A., Matthews, S. C., Paulus, M. P., & Stein, M. B. (2008). Intolerance of uncertainty correlates with insula activation during affective ambiguity. Neuroscience Letters, 430, 9297. Spitzer, R. L., Kroenke, K., Williams, J. B., & Lowe, B. (2006). A brief measure for assessing generalized anxiety disorder: the GAD-7. Archives of Internal Medicine, 166, 1092 1097. Stein, M. B., Jang, K. L., & Livesley, W. J. (1999). Heritability of anxiety sensitivity: a twin study. American Journal of Psychiatry, 156, 246251. Swinson, R. P. (2006). The GAD-7 scale was accurate for diagnosing generalised anxiety disorder. Evidence-Based Medicine, 11, 184. Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). New York, NY: Harper and Row. Taylor, S. (1993). The structure of fundamental fears. Journal of Behavior Therapy and Experimental Psychiatry, 24, 289299. Taylor, S. (Ed.). (1999). Anxiety sensitivity: theory, research, and treatment of the fear of anxiety. Mahwah, NJ: Erlbaum. Taylor, S., Koch, W. J., Woody, S., & McLean, P. (1996). Anxiety sensitivity and depression: how are they related? Journal of Abnormal Psychology, 105, 474479. Taylor, S., Zvolensky, M. J., Cox, B. J., Deacon, B., Heimberg, R. G., Ledley, D. R., et al. (2007). Robust dimensions of anxiety sensitivity: development and initial validation of the anxiety sensitivity index-3. Psychological Assessment, 19, 176188. Watson, D. (2000). Mood and temperament. New York, NY: Oxford University Press. Watson, D., & Clark, L. A. (1994). The PANAS-X: manual for the positive and negative affect schedule-expanded form. Iowa City, IA: Watson and Clark. Watson, D., Clark, L. A., & Harkness, A. R. (1994). Structures of personality and their relevance to psychopathology. Journal of Abnormal Psychology, 103, 1831. Watson, D., & Friend, R. (1969). Measurement of social-evaluative anxiety. Journal of Consulting and Clinical Psychology, 33, 448457. Weeks, J. W., Heimberg, R. G., Fresco, D. M., Hart, T. A., Turk, C. L., Schneier, F. R., et al. (2005). Empirical validation and psychometric evaluation of the brief fear of negative evaluation scale in patients with social anxiety disorder. Psychological Assessment, 17, 179190. Weeks, J. W., Heimberg, R. G., & Rodebaugh, T. L. (2008a). The fear of positive evaluation scale: assessing a proposed cognitive component of social anxiety. Journal of Anxiety Disorders, 22, 4455. Weeks, J. W., Heimberg, R. G., Rodebaugh, T. L., & Norton, P. J. (2008b). Exploring the relationship between fear of positive evaluation and social anxiety. Journal of Anxiety Disorders, 22, 386400. Zinbarg, R. E., Barlow, D. H., & Brown, T. A. (1997). Hierarchical structure and general factor saturation of the anxiety sensitivity index. Psychological Assessment, 9, 277 284. Zvolensky, M. J., Kotov, R., Antipova, A. V., Leen-Feldner, E. W., & Schmidt, N. B. (2005). Evaluating anxiety sensitivity, exposure to aversive life conditions, and problematic drinking in Russia: a test using an epidemiological sample. Addictive Behaviors, 30, 567570.