Comparing Cognitive, Metacognitive, and Acceptance and Commitment Therapy Models of Depression: A Longitudinal Study Survey

Full reference: Ruiz, F. J., & Odriozola-González, P. (2015). Comparing cognitive, metacognitive, and acceptance and commitment therapy models of depression: A longitudinal study survey. Spanish Journal of Psychology, 18, e39. doi:10.1017/sjp.2015.31

Study type: Longitudinal survey study (self-report measures at two time points with a 9-month interval)

Background and objectives

Unipolar depression is one of the most frequent psychiatric complaints and the leading cause of disability worldwide (Murray & López, 1996). Over the past decades, diverse psychological therapies with their respective etiological models of depression have been proposed. Cognitive Therapy (CT) by Beck is based on a diathesis-stress model proposing that depressogenic schemas—inflexible beliefs centered on the need to be perfect and obtain others' approval to be happy—constitute the primary cognitive vulnerability to depression. According to this model, these schemas generate distortions in information processing that lead to negative automatic thoughts.

Metacognitive Therapy (MCT) by Wells (2009) is based on the metacognitive model of emotional disorders. This approach proposes that the origin of disorders is not the presence of negative thoughts and emotions per se, but the activation of a specific thinking pattern called the Cognitive Attentional Syndrome (CAS). The CAS consists of repetitive thinking in the form of worry and rumination, excessive attentional focus on thoughts and feelings, and coping behaviors such as avoidance and thought suppression. This syndrome is problematic because it extends negative thinking, leads to reduced attentional flexibility, and causes failure to exercise appropriate control over negative experiences.

Acceptance and Commitment Therapy (ACT; Hayes, Strosahl, & Wilson, 1999) posits that psychological inflexibility lies at the core of psychopathology and behavioral ineffectiveness. Psychological inflexibility involves the dominance of private experiences (thoughts, emotions, sensations, memories) over chosen values and contingencies in guiding action. It is described in terms of interrelated processes: cognitive fusion (attachment to private experiences), experiential avoidance (deliberate efforts to avoid unwanted experiences), and lack of values clarity.

Although the MCT and ACT models originate from different philosophical and theoretical standpoints, they share important similarities. The MCT concept of CAS (which includes perseverative worry and rumination, excessive attention to thoughts and feelings, and counterproductive coping behaviors) appears very similar to ACT concepts of experiential avoidance and cognitive fusion. Worry and rumination have been proposed as experiential avoidance strategies, and excessive attentional focus to thoughts and feelings resembles cognitive fusion. Like experiential avoidance, the CAS is thought to have counterproductive effects leading to greater emotional disturbance. However, little research has explored the interrelations among key constructs of the CT, MCT, and ACT models in predicting depressive symptoms.

The current study was designed to explore the interrelationships between key constructs of CT (depressogenic schemas), MCT (dysfunctional metacognitive beliefs), and ACT (psychological inflexibility). Two predictions were made. First, according to previous research (Cristea et al., 2013; Ruiz & Odriozola-González, in press), it was predicted that psychological inflexibility would longitudinally mediate the relationship between depressogenic schemas and depressive symptoms. Second, given the similarities between the CAS and psychological inflexibility, it was hypothesized that psychological inflexibility would also mediate the relationship between dysfunctional metacognitive beliefs and depressive symptoms.

Method

Participants

The sample consisted of 289 participants (59.5% females) with ages ranging from 22 to 82 years (M = 35.38, SD = 8.63). Relative educational level was: 7.3% primary studies, 32.8% mid-level study graduates, and 59.9% college graduates. Participants responded to an anonymous internet survey distributed through social media. All were Spanish speakers. Thirty-six percent reported having received psychological or psychiatric treatment at some time, but only 6.6% were currently in treatment. Also, 4.8% of participants reported consumption of some psychotropic medication.

At follow-up (T2), 9 months after initial assessment (T1), 106 participants completed the study (36.7% of the original sample). There were no statistically significant differences between participants who completed the study and those who did not respond at T2 on any of the variables of interest.

Design

This was a longitudinal survey design with two measurement occasions separated by a nine-month interval. There was no manipulation of an independent variable; observational data were collected through self-report measures at both waves.

Intervention / Conditions

There was no intervention in this study. Participants completed questionnaires at two points in time without receiving any experimental intervention or treatment.

Outcome measures

The following instruments were used:

1. Acceptance and Action Questionnaire – II (AAQ-II; Bond et al., 2011): A general measure of experiential avoidance or psychological inflexibility. It consists of 7 items rated on a 7-point Likert-type scale (1 = never true; 7 = always true). Items reflect unwillingness to experience unwanted emotions and thoughts and inability to be in the present moment and behave according to value-directed actions. The Spanish version by Ruiz, Langer, Luciano, Cangas, and Beltrán (2013) was used, showing excellent internal consistency (mean α = .88).

2. Metacognitions Questionnaire-30 (MCQ-30; Wells & Cartwright-Hatton, 2004): A 30-item scale with a 4-point Likert-type format (1 = do not agree; 4 = agree very much). In this study, only the first three subscales were administered: Positive Beliefs about Worry, Negative Beliefs about Uncontrollability and Danger of Worry, and Beliefs about the Need to Control Thoughts. The Spanish version employed by Odriozola-González (2011) was used, showing good internal consistency in the administered subscales (alphas from .78 to .84).

3. Depression subscale of the Depression Anxiety and Stress Scales-21 (DASS-21; Antony, Bieling, Cox, Enns, & Swinson, 1998): A 21-item, 4-point Likert-type scale (0 = did not apply to me at all; 3 = applied to me very much, or most of the time). It contains three subscales (Depression, Anxiety, and Stress) of 7 items each. The Spanish version by Daza, Novy, Stanley, and Averill (2002) was used, showing good psychometric properties. Only results from the depression subscale are presented.

4. Dysfunctional Attitude Scale – Revised (DAS-R; de Graaf et al., 2009; Weissman & Beck, 1978): A measure designed to evaluate depressogenic schemas. It comprises 40 items rated on a 7-point Likert-type scale (1 = fully disagree; 7 = fully agree). The revised version (DAS-R) contains 17 items with two subscales: Perfectionism/Performance Evaluation (e.g., "It is difficult to be happy unless one is good-looking, intelligent, rich and creative") and Dependency (e.g., "My value as a person depends greatly on what others think of me"). The Spanish version of the DAS by Sanz and Vázquez (1993) was used without the 23 items eliminated by de Graaf et al. (2009) for the DAS-R, which showed good psychometric properties.

The primary outcome variable was depressive symptoms measured by the Depression subscale of the DASS-21 at T2.

Data analysis

Descriptive statistics, Cronbach's alphas, and zero-order correlations between all constructs were computed. Because scores on the constructs of interest did not show normal distribution, Mann-Whitney's U was computed to analyze potential differences between participants who completed the study and those who did not respond at T2.

Independent mediation analyses were conducted with the non-parametric bootstrapping procedure to estimate direct and indirect effects using the PROCESS package (Hayes, 2013). Predictor variables at T1 were depressogenic schemas (as measured by the DAS-R) and each of the specific types of metacognitive beliefs. The outcome was depressive symptoms as measured by the DASS-21 at T2. In all cases, the mediator variable was psychological inflexibility (as measured by the AAQ-II) at T2. To control for previous levels of depressive symptoms, scores on the DASS-21 at T1 were entered as a covariate in addition to demographic variables. Indirect effects were deemed significant if the 95% bias corrected (BC) bootstrap confidence intervals (CI) for those effects based on 20,000 bootstrapped samples did not include zero.

Results

Descriptive data, internal consistencies and zero-order correlations

Participants' mean scores on all measures did not differ significantly from scores obtained with nonclinical populations in other studies. The internal consistencies of the DASS-21, AAQ-II, and DAS-R were excellent, ranging from .89 to .93. The internal consistencies of the MCQ-30 factors were acceptable, ranging from .73 to .87.

Participants who completed the study did not show statistically significant differences from those who did not respond at T2 in psychological inflexibility (completers: M = 19.56, SD = 7.54; noncompleters: M = 19.36, SD = 7.98; U = 9546.5, p = .82). There were also no differences in depressogenic schemas (completers: M = 40.81, SD = 16.05; noncompleters: M = 43.48, SD = 19.08; U = 9161.5, p = .43), depressive symptoms (completers: M = 3.22, SD = 3.61; noncompleters: M = 4.15, SD = 4.32; U = 8568, p = .095), positive metacognitive beliefs (completers: M = 9.20, SD = 3.25; noncompleters: M = 9.08, SD = 2.89; U = 9677.5, p = .98), negative metacognitive beliefs (completers: M = 12.19, SD = 3.17; noncompleters: M = 11.96, SD = 3.50; U = 9254.5, p = .51), and metacognitive beliefs about the need to control thoughts (completers: M = 11.17, SD = 3.26; noncompleters: M = 11.45, SD = 3.57; U = 9334.5, p = .59).

Mediation analysis of the effect of depressogenic schemas on depressive symptoms

The mediation analysis revealed that psychological inflexibility, as measured at T2, acted as a mediator in the relationship between depressogenic schemas at T1 and depressive symptoms at T2. Depressogenic schemas significantly predicted the proposed mediator variable (psychological inflexibility, path a: TE = .166, SE = .044, p < .001) but not the dependent variable (depressive symptoms; path c or total effect: TE = .019, SE = .018, p = .29). However, psychological inflexibility at T2 significantly predicted depressive symptoms (path b: TE = .140, SE = .039, p = .0005). The indirect effect of depressogenic schemas on depressive symptoms through psychological inflexibility was statistically significant (path ab), with a point estimate of .023 (SE = .010; 95% BC CI [.008, .048]).

Mediation analysis of the effect of specific types of dysfunctional metacognitive beliefs on depressive symptoms

In the first mediation analysis, psychological inflexibility was shown to be a statistically significant mediator of the relationship between positive metacognitive beliefs at T1 and depressive symptoms at T2. Positive metacognitive beliefs marginally predicted psychological inflexibility (path a: TE = .377, SE = .204, p = .07) but did not predict depressive symptoms (path c or total effect: TE = .028, SE = .080, p = .73). However, psychological inflexibility at T2 was a significant predictor of depressive symptoms (path b: TE = .139, SE = .037, p = .0003). The point estimate of the indirect effect (path ab) was .052 (SE = .031) and was statistically significant with 95% BC CI [.005, .134].

In the second mediation analysis, psychological inflexibility acted as a mediator in the relationship between negative metacognitive beliefs at T1 and depressive symptoms at T2. Negative metacognitive beliefs at T1 predicted psychological inflexibility (path a: TE = .674, SE = .234, p < .01) and marginally predicted depressive symptoms (path c or total effect: TE = .161, SE = .093, p = .09) at T2. The latter prediction lost significance when psychological inflexibility was included in the model (path c' or direct effect: TE = .075, SE = .092, p = .42); however, psychological inflexibility significantly predicted depressive symptoms (path b: TE = .129, SE = .038, p = .0009). The indirect effect was significant (path ab), with a point estimate of .087 (SE = .049, 95% BC CI [.016, .214]).

In the third mediation analysis, psychological inflexibility was also a mediator of the relationship between metacognitive beliefs about the need to control thoughts at T1 and depressive symptoms at T2. The need to control thoughts at T1 predicted psychological inflexibility (path a: TE = .627, SE = .204, p < .01), but did not predict depressive symptoms (path c or total effect: TE = .073, SE = .082, p = .38) at T2. This prediction lost significance when psychological inflexibility was included in the model (path c' or direct effect: TE = –.014, SE = .081, p = .87); however, psychological inflexibility was a significant predictor of depressive symptoms (path b: TE = .139, SE = .038, p = .0004). The indirect effect was significant (path ab), with a point estimate of .087 (SE = .051, 95% BC CI [.013, .220]).

Discussion and conclusions

To the authors' knowledge, this is the first study that analyzed the potential longitudinal mediating role of psychological inflexibility in the effect of both metacognitive beliefs and depressogenic schemas on depressive symptoms. Results showed that psychological inflexibility, as measured at T2, acted as a mediator of the effect of depressogenic schemas and dysfunctional metacognitive beliefs at T1 on depressive symptoms at T2 after controlling for T1 level of depressive symptoms and demographic variables.

This study advances over previous cross-sectional studies that showed the relationship between dysfunctional schemas and depression was mediated by psychological inflexibility (Cristea et al., 2013; Ruiz & Odriozola-González, in press). Accordingly, psychological inflexibility seems to play a relevant role in the CT model of depression, and CT theorists might analyze the theoretical and practical implications of these findings. From an ACT perspective, depressogenic schemas can be seen as a type of rule-governed behavior mostly characterized by sensitivity to social whim, which in the ACT model is called generalized pliance.

The mediational results can be seen as relatively consistent with the MCT model if we accept that the CAS significantly overlaps with the ACT constructs of cognitive fusion and experiential avoidance, which are key behavioral processes underlying psychological inflexibility. In the ACT context, dysfunctional metacognitive beliefs can be seen as verbal rules that prompt cognitive fusion with negative thoughts and feelings and the use of experiential avoidance strategies (e.g., rumination, thought suppression) to deal with them. For instance, positive metacognitive beliefs prompt the use of rumination and counterproductive coping behaviors such as suppression, threat monitoring, and avoidance to deal with negative thoughts. Likewise, negative metacognitive beliefs prevent the person from interrupting rumination and promote further engagement in experiential avoidance because depressive experiences are viewed as dangerous.

The authors acknowledge several limitations. First, as all data were obtained using self-report measures, relationships among variables might be artificially inflated. Second, as the sample was composed of nonclinical participants, generalizability of the findings may be limited. Third, only 106 participants completed the study out of 289 who responded at T1; however, no differences in the scores of the psychological constructs of interest were found between completers and noncompleters. Fourth, three subscales of the MCQ-30, which is a measure more relevant to anxiety than depression, were used to assess dysfunctional metacognitive beliefs. There are metacognitive measures that assess beliefs more specific to depression, such as the Positive Beliefs about Rumination Scale (PBRS); however, no Spanish translation was available. Finally, the longitudinal design used does not allow for attributions of causality because no independent variable was manipulated; however, longitudinal studies such as this one provide a way to test the predictive ability and interrelations of key concepts of psychological models of psychological disorders.

In conclusion, this is the first study that longitudinally compared key constructs of CT (depressogenic schemas), MCT (dysfunctional metacognitive beliefs), and ACT (psychological inflexibility) in the prediction of depressive symptoms. Results highlight the relevance of psychological inflexibility as the most proximal predictor of depressive symptoms and warrant examination of further interrelationships between the CT, MCT, and ACT models of depression.

Significance and contribution

This study contributes to the field of depression research by longitudinally analyzing how key constructs from cognitive theory, metacognitive theory, and contextual behavioral science relate to depressive symptoms. The findings suggest that psychological inflexibility operates as a common mechanism through which both depressogenic schemas and dysfunctional metacognitive beliefs affect depressive symptoms. This comparative longitudinal approach provides empirical evidence for understanding depression across theoretical frameworks and offers insights into potential points of intervention targeting psychological flexibility as a transdiagnostic mechanism underlying multiple pathways to depressive psychopathology.

---