Examining Positive and Negative Affect as Outcomes and Moderators of Cognitive-Behavioral Therapy and Acceptance and Commitment Therapy for Social Anxiety Disorder.

The third wave in CBT shifted away from directly decreasing negative affect and toward acceptance and relinquishing control over maladaptive mood regulation attempts.

“The recent ‘third wave’ in cognitive behavioral therapy has shifted away from directly decreasing negative affect and instead focused on acceptance and relinquishing control over maladaptive attempts to regulate internal mood states.”

• cites Hayes & Hofmann, 2017 sourcePartly supported

  The Abstract says third-wave CBT emphasized acceptance and focused more on a person’s relationship to thoughts and emotions than on their content. It does not specifically describe a shift away from directly decreasing negative affect or toward relinquishing control over maladaptive attempts to regulate mood; it also says newer methods coexist with established ones. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Early CBT approaches for anxiety disorders aimed almost entirely at lowering negative affect and related impairment.

“In alignment with this defensive system, ‘first’ and ‘second-wave’ cognitive behavioral approaches for anxiety disorders almost exclusively focused on achieving reductions in negative affect – including fear and anxiety – and associated functional impairment (Barlow, Ellard, Sauer-Zavala, Bullis, & Carl, 2014; Carl, Soskin, Kerns, & Barlow, 2013; Craske, Meuret, Ritz, Treanor, & Dour, 2016).”

• cites Barlow, Ellard, Sauer-Zavala, Bullis, & Carl, 2014 sourceNot checked — no full text of the source
• cites Carl, Soskin, Kerns, & Barlow, 2013 sourceNot checked — no full text of the source
• cites Craske, Meuret, Ritz, Treanor, & Dour, 2016 sourceSupported

Low positive affect appears particularly characteristic of social anxiety disorder.

“Low positive affect appears to be particularly characteristic of social anxiety disorder (SAD; Watson & Naragon-Gainey, 2010).”

• cites Watson & Naragon-Gainey, 2010 sourceSupported

Exposure therapy reduces negative affect through fear extinction.

“For example, exposure therapy works to lessen negative affect through a mechanism principal to the defensive system – fear extinction (Craske, Treanor, Conway, Zbozinek, & Vervliet, 2014).”

• cites Craske, Treanor, Conway, Zbozinek, & Vervliet, 2014 sourceSupported

These findings are paralleled by results from longitudinal datasets, experimental designs, and structural equation modelling.

“These findings are paralleled with results harnessing longitudinal datasets, experimental designs, and structural equation modelling methods (Brown, Chorpita, & Barlow, 1998; Kashdan, Weeks, & Savostyanova, 2011; Khazanov & Ruscio, 2016; Sewart et al., 2019).”

• cites Brown, Chorpita, & Barlow, 1998 sourceNot checked — no full text of the source
• cites Kashdan, Weeks, & Savostyanova, 2011 sourceNot checked — no full text of the source
• cites Khazanov & Ruscio, 2016 sourceSupported

Previous findings show that cognitive behavioral and mindfulness-based psychotherapies are largely ineffective in heightening positive affect for depression.

“This converges with previous findings that cognitive behavioral and mindfulness-based psychotherapies are largely ineffective in heightening PA for depression (Boumparis et al., 2016; Dichter et al., 2009; Moore et al., 2013).”

• cites Boumparis et al., 2016 sourceNot checked — no full text of the source
• cites Dichter et al., 2009 sourceCould not verify

  The Methods describe a study of behavioral activation therapy, not cognitive behavioral or mindfulness-based psychotherapy. The Results and Conclusions report changes in neural responses to rewards, but do not assess whether those other therapies heighten positive affect (searched: positive affect, mindfulness, CBT, and anhedonia). (LLM, conf=0.78)

  Experimental — accuracy not yet measured
• cites Moore et al., 2013 sourcePartly supported

  The Results (page not provided) found no significant difference in positive affect between behavioral activation and information-support therapy among dementia caregivers. That supports a narrower finding than the citing sentence: this trial did not test mindfulness-based psychotherapy or establish that cognitive behavioral and mindfulness-based therapies are broadly ineffective at increasing positive affect in people with depression. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Baseline positive affect did not moderate the theory-relevant treatment outcomes.

“Neither PA nor NA were found to moderate theoretically-relevant treatment outcomes.”

CBT did not reduce negative affect more than the waitlist condition.

“Although ACT outperformed WL in reducing NA, this effect was not observed for CBT.”

ACT did not increase positive affect more than the waitlist condition.

“PA increased significantly in both CBT and ACT from pre to post-treatment, with neither ACT nor CBT outperforming WL in increasing PA.”

Baseline negative affect did not moderate the theory-relevant treatment outcomes.

“Neither PA nor NA were found to moderate theoretically-relevant treatment outcomes.”

CBT did not increase positive affect more than the waitlist condition.

“PA increased significantly in both CBT and ACT from pre to post-treatment, with neither ACT nor CBT outperforming WL in increasing PA.”

ACT reduced negative affect more than the waitlist condition.

“Although ACT outperformed WL in reducing NA, this effect was not observed for CBT.”

Both CBT and ACT showed significant reductions in negative affect from before to after treatment.

“NA decreased significantly in both CBT and ACT from pre to post-treatment.”

Positive affect increased significantly from pre- to post-treatment (S1) in ACT.

“From pre to post-treatment (S1), positive affect increased significantly for CBT (b = 0.3; CI = 0.1 to 0.6; p = .009) and ACT (b = 0.3; CI = 0.01 to 0.5; p = .05), but not WL (b = 0.1; CI = −0.1 to 0.4; p = .37).”

At 12-month follow-up, negative affect for ACT averaged at the 69th percentile.

“Twelve months following treatment, negative affect of those who had those received CBT were averaged at the 74th percentile, while those receiving ACT were at the 69th percentile.”

Positive affect did not moderate quality of life for CBT or ACT during S1.

“Nor did positive affect moderate quality of life for CBT or ACT for S1 (b = −0.2; CI = −1.0 to 0.5; p = .54) or S2 (b = 0.2; CI = −0.2 to 0.7; p = .28).”

Regardless of treatment condition, positive affect did not significantly predict social anxiety outcome during S1.

“Further testing revealed that regardless of treatment condition neither negative affect nor positive affect significantly predicted outcome for S1 (negative affect b = 0.1; CI = −0.1 to 0.3; p = .50; positive affect b = −0.1; CI = −0.4 to 0.1; p = .28) or S2 (negative affect b = 0.01; CI = −0.1 to 0.1; p = .88; positive affect b = −0.01; CI = −0.1 to 0.4; p = .27).”

At pre-treatment, positive affect for CBT was at the 24th percentile of normative values.

“Positive affect was below normative values at pretreatment as indicated by average values in the 24th and 28th percentile for CBT and ACT, respectively.”

Regardless of treatment condition, negative affect did not significantly predict social anxiety outcome during S2.

“Further testing revealed that regardless of treatment condition neither negative affect nor positive affect significantly predicted outcome for S1 (negative affect b = 0.1; CI = −0.1 to 0.3; p = .50; positive affect b = −0.1; CI = −0.4 to 0.1; p = .28) or S2 (negative affect b = 0.01; CI = −0.1 to 0.1; p = .88; positive affect b = −0.01; CI = −0.1 to 0.4; p = .27).”

At 12-month follow-up, negative affect for CBT averaged at the 74th percentile.

“Twelve months following treatment, negative affect of those who had those received CBT were averaged at the 74th percentile, while those receiving ACT were at the 69th percentile.”

At pre-treatment, negative affect for ACT was at the 84th percentile of normative values.

“At pre-treatment, levels of negative affect were notably above normative values as indicated by average values in the 90nd and 84th percentile for CBT and ACT, respectively.”

CBT did not show larger increases in positive affect than waitlist from pre- to post-treatment (S1).

“For S1 pair-wise slope comparisons, neither ACT nor CBT demonstrated larger increases in positive affect than WL (ACT b = 0.1; CI = −0.2 to 0.5; p = .48; CBT b = −0.2; CI= −0.6 to 0.2; p = .26).”

Positive affect did not increase significantly from pre- to post-treatment (S1) in the waitlist condition.

“From pre to post-treatment (S1), positive affect increased significantly for CBT (b = 0.3; CI = 0.1 to 0.6; p = .009) and ACT (b = 0.3; CI = 0.01 to 0.5; p = .05), but not WL (b = 0.1; CI = −0.1 to 0.4; p = .37).”

Negative affect did not differentially moderate quality of life for CBT or ACT during S1.

“Negative affect was not found to differentially moderate quality of life for CBT or ACT for S1 (b = 0.1; CI = −0.7 to 0.8; p = .81) or S2 (b = −0.01; CI = −0.5 to 0.4; p = .96).”

Positive affect did not predict quality of life outcome during S2.

“In contrast, as shown in Figure 2, positive affect predicted outcome for S1 (b = −0.6; CI = −1.0 to −0.2; p = .002), but not for S2 (b = 0.1; CI = −0.2 to 0.2; p = .99).”

The S2 change slope for negative affect was not significant in CBT.

“S2 change slopes were not significant in CBT (p = .19) or ACT (p = .37).”

Negative affect did not differentially moderate social anxiety symptoms for CBT or ACT during S2.

“Negative affect was not found to differentially moderate social anxiety symptoms for CBT or ACT for S1 (b = 0.4; CI = −0.1 to 0.8; p = .09) or S2 (b = 0.1; CI = −0.2 to 0.2; p = .85).”

Negative affect decreased significantly from pre- to post-treatment (S1) in ACT.

“From pre to post-treatment (S1), negative affect decreased significantly for both CBT (b = −0.5; CI = −0.8 to −0.3; p <. 001) and ACT (b = −0.7; CI = −0.9 to −0.3; p <. 001), but not WL (b = −0.2; CI = −0.5 to −0.1; p = .18).”

ACT showed larger reductions in negative affect than waitlist from pre- to post-treatment (S1).

“For S1 pair-wise slope comparisons, ACT demonstrated larger reductions in negative affect than WL (b = −0.5; CI = −0.9 to −0.1; p = .01).”

Positive affect increased significantly from pre- to post-treatment (S1) in CBT.

“From pre to post-treatment (S1), positive affect increased significantly for CBT (b = 0.3; CI = 0.1 to 0.6; p = .009) and ACT (b = 0.3; CI = 0.01 to 0.5; p = .05), but not WL (b = 0.1; CI = −0.1 to 0.4; p = .37).”

Positive affect did not moderate social anxiety symptoms for CBT or ACT during S2.

“Nor did positive affect moderate social anxiety symptoms for CBT or ACT for S1 (b = 0.04; CI = −0.4 to 0.5; p = .88) or S2 (b= 0.2; CI = −0.1 to 0.4; p = .30).”

The S2 change slope for negative affect was not significant in ACT.

“S2 change slopes were not significant in CBT (p = .19) or ACT (p = .37).”

The slope change in quality of life was not significant for individuals with high positive affect (1 SD above the mean).

“Slope change was not found to be significant for individuals with high positive affect (1 SD above the mean; slope = 0.05; CI = −0.4 to 0.5; p=.82).”

At 12-month follow-up, positive affect for ACT averaged at the 36th percentile.

“Twelve months following treatment, positive affect of those who had those received CBT were averaged at the 41st percentile, while those receiving ACT were at the 36th percentile.”

Regardless of treatment condition, negative affect did not significantly predict social anxiety outcome during S1.

“Further testing revealed that regardless of treatment condition neither negative affect nor positive affect significantly predicted outcome for S1 (negative affect b = 0.1; CI = −0.1 to 0.3; p = .50; positive affect b = −0.1; CI = −0.4 to 0.1; p = .28) or S2 (negative affect b = 0.01; CI = −0.1 to 0.1; p = .88; positive affect b = −0.01; CI = −0.1 to 0.4; p = .27).”

At 12-month follow-up, positive affect for CBT averaged at the 41st percentile.

“Twelve months following treatment, positive affect of those who had those received CBT were averaged at the 41st percentile, while those receiving ACT were at the 36th percentile.”

The S2 change slope for positive affect was not significant in ACT.

“S2 change slopes were not significant in CBT (p = .62) or ACT (p = .88).”

Regardless of treatment condition, negative affect did not significantly predict quality of life outcome during S1.

“Regardless of treatment condition, negative affect did not significantly predict outcome for S1 (b = −0.01; CI = −0.4 to 0.4; p = .97) or S2 (b = −0.1; CI = −0.3 to 0.1; p = .48).”

Negative affect decreased significantly from pre- to post-treatment (S1) in CBT.

“From pre to post-treatment (S1), negative affect decreased significantly for both CBT (b = −0.5; CI = −0.8 to −0.3; p <. 001) and ACT (b = −0.7; CI = −0.9 to −0.3; p <. 001), but not WL (b = −0.2; CI = −0.5 to −0.1; p = .18).”

At post-treatment, negative affect for CBT had reduced to the 78th percentile.

“At post-treatment negative affect had reduced to the 78th percentile for CBT and 63rd percentile for ACT.”

Among participants with average positive affect, the slope change in quality of life was significant.

“Tests of simple effects revealed that among participants with low (1 SD below the mean) and average (mean) positive affect, the slope change in quality of life was significant (low positive affect slope= 1.1; CI = 0.6 to 1.6; p < .001; average positive affect slope = 0.5; CI = 0.3 to 0.9; p < .001).”

Negative affect did not decrease significantly from pre- to post-treatment (S1) in the waitlist condition.

“From pre to post-treatment (S1), negative affect decreased significantly for both CBT (b = −0.5; CI = −0.8 to −0.3; p <. 001) and ACT (b = −0.7; CI = −0.9 to −0.3; p <. 001), but not WL (b = −0.2; CI = −0.5 to −0.1; p = .18).”

Negative affect did not differentially moderate quality of life for CBT or ACT during S2.

“Negative affect was not found to differentially moderate quality of life for CBT or ACT for S1 (b = 0.1; CI = −0.7 to 0.8; p = .81) or S2 (b = −0.01; CI = −0.5 to 0.4; p = .96).”

Individuals with high positive affect already had high quality of life at pre-treatment.

“However, high positive affect individuals already demonstrated high quality of life at pre-treatment.”

Regardless of treatment condition, positive affect did not significantly predict social anxiety outcome during S2.

“Further testing revealed that regardless of treatment condition neither negative affect nor positive affect significantly predicted outcome for S1 (negative affect b = 0.1; CI = −0.1 to 0.3; p = .50; positive affect b = −0.1; CI = −0.4 to 0.1; p = .28) or S2 (negative affect b = 0.01; CI = −0.1 to 0.1; p = .88; positive affect b = −0.01; CI = −0.1 to 0.4; p = .27).”

Regardless of treatment condition, negative affect did not significantly predict quality of life outcome during S2.

“Regardless of treatment condition, negative affect did not significantly predict outcome for S1 (b = −0.01; CI = −0.4 to 0.4; p = .97) or S2 (b = −0.1; CI = −0.3 to 0.1; p = .48).”

At pre-treatment, positive affect for ACT was at the 28th percentile of normative values.

“Positive affect was below normative values at pretreatment as indicated by average values in the 24th and 28th percentile for CBT and ACT, respectively.”

At pre-treatment, negative affect for CBT was at the 90th percentile of normative values.

“At pre-treatment, levels of negative affect were notably above normative values as indicated by average values in the 90nd and 84th percentile for CBT and ACT, respectively.”

Positive affect did not moderate social anxiety symptoms for CBT or ACT during S1.

“Nor did positive affect moderate social anxiety symptoms for CBT or ACT for S1 (b = 0.04; CI = −0.4 to 0.5; p = .88) or S2 (b= 0.2; CI = −0.1 to 0.4; p = .30).”

Positive affect predicted quality of life outcome during S1.

“In contrast, as shown in Figure 2, positive affect predicted outcome for S1 (b = −0.6; CI = −1.0 to −0.2; p = .002), but not for S2 (b = 0.1; CI = −0.2 to 0.2; p = .99).”

The S2 change slope for positive affect was not significant in CBT.

“S2 change slopes were not significant in CBT (p = .62) or ACT (p = .88).”

Positive affect did not moderate quality of life for CBT or ACT during S2.

“Nor did positive affect moderate quality of life for CBT or ACT for S1 (b = −0.2; CI = −1.0 to 0.5; p = .54) or S2 (b = 0.2; CI = −0.2 to 0.7; p = .28).”

CBT and ACT did not significantly differ from each other in their S1 slopes for negative affect.

“Inconsistent with our hypothesis, CBT and ACT S1 slopes did not significantly differ from each other (b = −0.2; CI = −0.5 to 0.2; p = .39).”

At post-treatment, negative affect for ACT had reduced to the 63rd percentile.

“At post-treatment negative affect had reduced to the 78th percentile for CBT and 63rd percentile for ACT.”

Among participants with low positive affect (1 SD below the mean), the slope change in quality of life was significant.

“Tests of simple effects revealed that among participants with low (1 SD below the mean) and average (mean) positive affect, the slope change in quality of life was significant (low positive affect slope= 1.1; CI = 0.6 to 1.6; p < .001; average positive affect slope = 0.5; CI = 0.3 to 0.9; p < .001).”

At post-treatment, positive affect increased to the 41st percentile for both CBT and ACT.

“At post-treatment, positive affect had increased to the 41st percentile for both CBT and ACT.”

ACT and CBT did not significantly differ from each other in their S1 slopes for positive affect.

“Inconsistent with our hypothesis, ACT and CBT S1 slopes did not significantly differ from each other (b = −0.01; CI = −0.4 to 0.3; p = .70).”

ACT did not show larger increases in positive affect than waitlist from pre- to post-treatment (S1).

“For S1 pair-wise slope comparisons, neither ACT nor CBT demonstrated larger increases in positive affect than WL (ACT b = 0.1; CI = −0.2 to 0.5; p = .48; CBT b = −0.2; CI= −0.6 to 0.2; p = .26).”

Negative affect did not differentially moderate social anxiety symptoms for CBT or ACT during S1.

“Negative affect was not found to differentially moderate social anxiety symptoms for CBT or ACT for S1 (b = 0.4; CI = −0.1 to 0.8; p = .09) or S2 (b = 0.1; CI = −0.2 to 0.2; p = .85).”

Traditional CBT has largely ignored positive affect.

“These traditional cognitive behavioral approaches (referred to as ‘CBT’), however, have largely disregarded positive affect – one’s pleasurable engagement with the environment which results in positive emotional states (e.g., joy, interest, and enthusiasm) – an affective dimension which is generated by a separate motivational system related to appetitive and goal directed behaviors (Davidson, 1994”

In that prior study, positive affect improved in two of seven participants.

“Positive affect was found to improve in only two of seven participants, and worsened in five of seven – even as negative affect improved over the course of treatment.”

• cites Bosley, Fisher, & Taylor, 2018 sourceNot checked — no full text of the source

ACT and CBT likely share common treatment mechanisms, making them more similar than distinct.

“Findings suggest that ACT and CBT share common treatment mechanisms making them more similar than distinct.”

A meta-analysis reported a moderate negative association between social anxiety and positive affect (r = −.36; 95% CI: −.31 to −.40).

“A meta-analysis examining cross-sectional and experience sampling studies found a moderate, inverse relationship between social anxiety and positive affect (r = −.36; 95% CI: −.31 to −.40) which remained consistent after covarying for depressive symptoms (r = −.21; 95% CI: −.16 to −.26; Kashdan, 2007).”

• cites Kashdan, 2007 sourceNot checked — no full text of the source

In that prior study, positive affect worsened in five of seven participants.

“Positive affect was found to improve in only two of seven participants, and worsened in five of seven – even as negative affect improved over the course of treatment.”

• cites Bosley, Fisher, & Taylor, 2018 sourceNot checked — no full text of the source

In that prior study, negative affect improved over treatment.

“Positive affect was found to improve in only two of seven participants, and worsened in five of seven – even as negative affect improved over the course of treatment.”

• cites Bosley, Fisher, & Taylor, 2018 sourceNot checked — no full text of the source

That inverse relationship remained after controlling for depressive symptoms (r = −.21; 95% CI: −.16 to −.26).

“A meta-analysis examining cross-sectional and experience sampling studies found a moderate, inverse relationship between social anxiety and positive affect (r = −.36; 95% CI: −.31 to −.40) which remained consistent after covarying for depressive symptoms (r = −.21; 95% CI: −.16 to −.26; Kashdan, 2007).”

• cites Kashdan, 2007 sourceNot checked — no full text of the source

Compared with CBT, ACT's treatment philosophy is more consistent with appetitive and goal-directed behavior.

“When compared with CBT, ACT treatment philosophy aligns more consistently with appetitive and goal directed behaviors.”

These strategies have consistently reduced negative affect and related symptoms.

“Indeed, these strategies have consistently demonstrated effectiveness in reducing negative affect and related symptomatology (e.g., Craske et al., 2019).”

• cites Craske et al., 2019 sourceNot checked — no full text of the source

Insufficient focus on positive affect might partly account for why CBT for SAD often yields suboptimal responses.

“Lack of explicit attention to the positive dimension of affect may in part explain less-than-optimal treatment response rates for CBT for SAD (see Loerinc et al., 2015).”

• cites Loerinc et al., 2015 sourceNot checked — no full text of the source

In depression, CBT has not shown a significant increase in positive affect (g=.37, 95% CI −0.004-0.77, p < .07).

“Cognitive behavioral therapies have not been found to significantly increase positive affect for individuals with depression (g=.37, 95% CI −0.004–0.77, p < .07; Boumparis, Karyotaki, Kleiboer, Hofmann, & Cuijpers, 2016).”

• cites Boumparis, Karyotaki, Kleiboer, Hofmann, & Cuijpers, 2016 sourceNot checked — no full text of the source

Prior to this study, only one investigation had tested CBT's impact on positive affect in anxiety disorders, using person-specific analyses with seven generalized anxiety disorder patients.

“Thus far, the effect of CBT on positive affect for anxiety disorders has only been examined in one study through person-specific analyses on seven individuals with generalized anxiety disorder (Bosley, Fisher, & Taylor, 2018).”

• cites Bosley, Fisher, & Taylor, 2018 sourceNot checked — no full text of the source

Two validated self-report scales were Z-score combined to yield a more valid and reliable index of self-reported positive and negative affect.

“In an effort to create a more valid and reliable index of self-reported positive and negative affect, scores from two well-validated self-report measures were combined following Z-score transformation.”

The reduction in negative affect observed in ACT is likely a byproduct of ACT-based strategies.

“Although reducing NA is not a focus of ACT, it’s likely that this reduction is a byproduct of ACT-based strategies.”

Anxiety disorders are characterized by abnormal activation of a theorized defensive system that produces elevated negative affect.

“Anxiety disorders have long been characterized by abnormal activation of a theorized defensive system which gives rise to elevated levels of negative affect (NA; e.g., fear; Gray, 1994).”

• cites Gray, 1994Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

The lack of quality-of-life improvement in individuals with high pre-treatment positive affect appears to be a ceiling effect because they already reported high quality of life before treatment.

“As displayed in Fig. 2, this appears to be a ceiling effect due to individuals with high pretreatment PA reporting high levels of QOL prior to initiation of ACT and CBT thus giving less potential for improvement.”

This is the first study to examine changes in positive and negative affect as a result of CBT and ACT for anxiety disorders.

“This is the first study to our knowledge that examines changes in PA and NA as a result of CBT and ACT for anxiety disorders.”

A subset of anxiety disorders, including social anxiety disorder, are characterized by low positive affect resulting from separate deficits in appetitive learning processes.

“However, a subset of anxiety disorders including social anxiety disorder are also characterized by low positive affect (PA) resulting from separate deficits in appetitive learning processes (e.g., Brown et al., 1998).”

• cites Brown et al., 1998 sourceNot checked — no full text of the source

Individuals with low positive affect may benefit from augmenting CBT and ACT with strategies specifically designed to regulate positive emotion.

“Furthermore, our findings suggest that individuals with low PA may benefit from augmenting CBT and ACT with strategies specifically designed to regulate positive emotion.”

Extraversion was found to be a non-specific significant linear predictor of composite symptoms, with higher extraversion associated with fewer symptoms across follow-up assessments.

“Craske et al. 2014 found extraversion to be a non-specific significant linear predictor of composite symptoms (z=1.96, p=.05), wherein participants higher in extraversion reported significantly fewer symptoms collapsed across follow-up assessments (Craske et al., 2014).”

• cites Craske et al., 2014Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

There were no significant pre-treatment group differences between CBT, ACT, and waitlist on the variables of interest.

“No significant pre-treatment group differences were found between CBT, ACT, WL for variables of interest (ps > .23).”