Increasingly certain about uncertainty: Intolerance of uncertainty across anxiety and depression

Increasingly certain about uncertainty: Intolerance of uncertainty across anxiety and depression

Reference: Carleton, R. N., Mulvogue, M. K., Thibodeau, M. A., McCabe, R. E., Antony, M. M., & Asmundson, G. J. G. (2012). Increasingly certain about uncertainty: Intolerance of uncertainty across anxiety and depression. Journal of Anxiety Disorders, 26, 468-479. doi:10.1016/j.janxdis.2012.01.011

Aims of the Current Study

·         This study is the first systematic attempt to examine intolerance of uncertainty (IU) endorsement rates and response patters across undergraduate, community, and clinical samples. IU is a potentially important transdiagnostic construct involved in the development and maintenance of anxiety and mood disorders.

Introduction and Literature Review

·         Identifying and understanding constructs common to anxiety and depression, such as IU, may provide directions for investigating transdiagnostic vulnerabilities and for developing transdiagnostic treatment protocols.

·         IU is defined as a negative response to ambiguity/uncertainty, beliefs about uncertainty as threatening, and the inability to cope with ambiguity and change

·         Two dimensions of IU

o   Prospective IU = the cognitively focused dimension of IU (e.g., “Unforeseen events upset me greatly”) à expected to be more related to worry and obsessive compulsive symptoms

o   Inhibitory IU = the behaviorally focused dimension of IU (e.g., “The smallest doubt can stop me from acting”) àexpected to be more related to social anxiety, panic, agoraphobia, and depression.

·         Growing evidence that IU may be a transdiagnostic cognitive construct across undergraduate, community, and clinical populations

o   Especially GAD, but also OCD, social anxiety disorder (SAD), panic disorder with or without agoraphobia (PDA), and major depressive disorder (MDD)

Hypotheses

·         Endorsement rates:

o   IU endorsement is expected to be higher in clinical sample relative to both the undergraduate and community samples

o   Within the clinical sample: OCD and GAD > MDD, PDA, and SAD

·         2-factor solution of the IUS-12 was expected

·         Prospective IU was hypothesized to be more related to worry and OCD symptoms

·         Inhibitory IU was hypothesized to be more related to SAD, PDA, and MDD

Method

·         Participants

o   376 outpatients: SAD (n = 120, 32%), PDA (n = 89, 24%), GAD (n = 63, 17%), OCD (n = 60, 16%), and MDD (n = 26, 7%)

o   428 undergraduates

o   571 community members

·         Measures

o   Intolerance of Uncertainty Scale, short form (IUS-12)

o   Structured Clinical Interview for DSM-IV (SCID)

·         Analyses

o   Empirical distributions of the IUS-12 scores across diagnostic, community, and undergraduate groups were studied using Kernel density estimation curves

o   Two ANOVAs comparing total and subscale means of the IUS-12 across groups

§  ANOVA 1: primary GAD vs. primary OAD + co-occurring GAD vs. primary OAD (w/o GAD) vs. primary MDD vs. undergraduates vs. community (Tukey post-hoc comparisons, α = .01)

§  ANOVA 2: SAD vs. PDA vs. GAD vs. OCD vs. MDD vs. undergraduates vs. community (Tukey post-hoc comparisons)

o   Confirmatory factor analysis (CFA) for 3 purposes:

§  (1) Assess the fit of the presumed 2-factor structure for each of the clinical, undergraduate, and community samples

§  (2) Assess if the IUS-12 factor structure, measurement weights, and structural covariances differed for men vs. women

§  Assess if the IUS-12 factor structure, measurement weights, and structural covariances differed for diagnostic groups

Results

·         Some demographic differences between groups (e.g., age, gender)

·         Items and summed scale alternatives demonstrated acceptable skew and kurtosis

·         Almost no statistically significant differences between men and women on the total or subscale scores of the IUS-12 in any sample

·         Kernel distribution estimations: 3 general patterns

o   SAD, GAD, OCD, and MDD demonstrated very similar distributions: high range of IUS-12 scores, normal to negatively skewed

o   Undergraduate and community samples demonstrated very similar distributions to each other: lower range of IUS-12 scores, disparate from the clinical groups, positively skewed

o   PDA wider range of IUS-12 scores

·         ANOVA 1

o   Primary GAD, primary OAD + co-occurring GAD, primary OAD (w/o GAD), primary MDD > undergraduates, community

·         ANOVA 2

o   SAD, PDA, GAD, OCD, MDD > undergraduates, community

o   Re-ran ANOVA excluding participants with a comorbid diagnosis à still no statistically significant differences between clinical groups

·         CFA

o   2-factor structure was statistically superior to a unitary structure for men and women in all 3 samples

·         Invariance analyses

o   No differences between men and women based on measurement weights or structural covariances for any of the 3 samples (clinical, undergraduate, community), suggesting response patterns for men and women were comparable

o   No differences between undergraduate and community samples based on measurement weights suggesting response patterns were largely similar; however, the structural covariances were significant different

o   Significant and substantial differences in response patterns based on measurement weights or structural covariances between the clinical sample and each of the undergraduate and community samples

o   No differences across the diagnostic groups based on measurement weights or structural covarainces

Discussion

·         Three important contributions

o   First direct comparative analyses of IU response patterns and empirical distributions, as measured by IUS-12, across clinical, undergraduate, and community samples

§  Empirical distributions of IU for those with principal SAD, GAD, OCD, and MDD very similar (although PDA had relatively large variance of scores)

§  No statistically significant differences between any diagnostic groups

§  IU appears to not be specific to any given disorder (e.g., OCD or GAD) but instead seems to represent a broad transdiagnostic cognitive vulnerability

o   First presentation of normative IUS-12 data for several clinical samples and further assessment of the proposed IUS-12 factor structure

§  2-factor structure was supported and superior to unitary alternative for all three samples

o   Further psychometric support for the IUS-12 as being invariant across sex with a robust factor structure

§  Response patterns of undergraduate and community participants were largely comparable; clinical response patterns were also comparable to one another but were substantially different from undergraduate/community samples

·         Limitations

o   Samples sizes for each diagnostic group within the clinical sample are insufficient for robust CFA (preliminary analyses)

o   No inter-rater reliability information for SCID diagnoses

o   Clinical sample had high levels of comorbidity (not diagnostically “pure”)

o   Undergraduate and community samples were not diagnostically assessed

o   The causality of IU and the development of anxiety/depressive disorders cannot be determined

o   Unable to determine mechanisms or manner in which IU relates to each disorder

o   Sample primarily Caucasian

·         Future directions

o   Do individuals with anxiety and depression experience greater IU in general or IU specific only to their disorder (e.g., to bodily sensations in panic)

o   Incorporating IU-specific treatment components into therapy (e.g., in vivo exposure to uncertainty)

o   Further taxometric investigation of IU as measured by the IUS-12

o   Further studies directly comparing IUS-27 and IUS-12

Discussion Questions

·         Based on the results reported in this study (and reviewed from previous studies), does it seem to you that IU is indeed a transdiagnostic construct across anxiety and depression (vs. unique to OCD and/or GAD)?

o   If so what evidence do you find most compelling?

o   If not, what other information would you need in order to make your decision?

·         What is the next study in this area that you think needs to be done?

·         Based on your research knowledge and clinical experiences, what are your preliminary thoughts/hypotheses about the empirical question raised in the future directions section about whether individuals with anxiety and depression experience greater IU in general or IU specific only to their disorder (e.g., to bodily sensations in panic)?

·         Do any of the limitations mentioned above (e.g., small sample sizes, no diagnostic inter-rater reliability) particularly concern you?

·         For those of you who research disorders/problems outside of anxiety and depression, are there ways that you can see this work on IU extending to other concerns/populations?