How are triggers for repetitive negative thinking organized? A relational frame analysis

Full reference: Gil-Luciano, B., Calderón-Hurtado, T., Tovar, D., Sebastián, B., & Ruiz, F. J. (2019). How are triggers for repetitive negative thinking organized? A relational frame analysis. Psicothema, 31(1), 53-59. https://doi.org/10.7334/psicothema2018.133

Study type: Experimental descriptive study of derived relations

Background and objectives

Repetitive negative thinking (RNT) has been identified as an important transdiagnostic process present across multiple psychological disorders. Although research has advanced in identifying RNT as a central process in psychopathology, in distinguishing RNT from other forms of recurrent cognition (such as obsessions and intrusive memories), and in identifying metacognitive beliefs that support it, one area has been scarcely analyzed: the relationships among triggers (negative thoughts) for RNT. From the perspective of Relational Frame Theory (RFT), it has recently been suggested that these triggers typically become related hierarchically across the individual's learning history, meaning that some thoughts acquire aversive functions because they are part of a more overarching trigger that signals an abstract negative reinforcer at the top of a hierarchical network. This study aimed to initiate an empirical analysis of the relationships among RNT triggers by testing the hypothesis that triggers are typically organized in hierarchical relations.

Method

Participants

One hundred undergraduate students (71 women; age range = 18-32 years; M = 20.33, SD = 1.77) participated in this study. None had previous experience with the procedures or theory (RFT) involved in the study. Participants were recruited through class advertisements and received 20,000 Colombian pesos (approximately 6 euros) as compensation. Seventy-five participants were studying Psychology, 8 International Business Studies, 7 Engineering, 2 Mathematics, and 2 Law. The remaining participants were studying other careers such as Marketing, Physical Education, Philosophy, and Arts.

Experimental design

The study employed an observational descriptive design in which participants evaluated their triggers for RNT and selected one of three suggested models representing different ways their triggers could be related: coordination, comparison, and hierarchy. The primary variables were the type of diagram selected (coordination, comparison, or hierarchy) and scores on RNT measures. The main comparison was between participants who selected HIERARCHY versus COMPARISON in relation to their scores on the Perseverative Thinking Questionnaire (PTQ) and the Penn State Worry Questionnaire-11 (PSWQ-11).

Materials and apparatus

Several standardized instruments were used: (1) Mini International Neuropsychiatric Interview (MINI; Sheehan et al., 1998) for diagnostic assessment; (2) Personal Health Questionnaire-9 (PHQ-9; Kroenke, Spitzer, & Williams, 2001); (3) Generalized Anxiety Disorder-7 Scale (GAD-7; Spitzer, Kroenke, Williams, & Löwe, 2006); (4) Depression Anxiety and Stress Scales-21 (DASS-21; Lovibond & Lovibond, 1995); (5) Perseverative Thinking Questionnaire (PTQ; Ehring et al., 2011); (6) Penn State Worry Questionnaire-11 (PSWQ-11; Meyer, Miller, Metzger, & Borkovec, 1990); (7) List of Triggers for RNT (LT-RNT), a 70-item list designed for this study assessing whether participants typically experience thoughts with avoidance discriminative functions; and (8) Three diagrams designed for the study representing ways RNT triggers could be organized. The diagrams used circles and figures of gangsters to represent triggers: the COORDINATION diagram presented circles and figures of equal size; the COMPARISON diagram presented circles and figures with increasing sizes; and the HIERARCHY diagram presented a central circle with other smaller circles orbiting around it and miniatures of gangster figures at the top, middle, and bottom.

Procedure

The study was conducted in the Laboratory of Clinical Psychology of a Colombian university. All participants received detailed information about the study and signed informed consent. Subsequently, the MINI interview was administered. Following this, participants completed the PHQ-9, GAD-7, DASS-21, PTQ, and PSWQ-11, in that order. Then the LT-RNT was administered and the three diagrams representing ways RNT triggers could be organized were explained. Participants were asked to select the diagram that best described them. The interviewer asked participants to justify their selection to ensure they had understood the diagrams and also asked them to depict the selected diagram with their six most problematic thoughts. Finally, participants received complete information about the study's objectives. Participants experiencing psychological difficulties according to the MINI, PHQ-9, or GAD-7 were offered participation in a clinical study or alternatives for inexpensive psychological services.

Data analysis

Analyses were computed with free software JASP 0.8.6. First, descriptive data were calculated. Second, a Multinomial Test was conducted to analyze whether participants selected one of the three types of diagrams (COORDINATION, COMPARISON, or HIERARCHY) more or less frequently than expected by chance (33.33% of selections). Ninety-five percent confidence intervals (CI) were computed for the selected proportion of each diagram type, and the p-value calibration introduced by Sellke, Bayarri, and Berger (2001), also known as the Vovk-Sellke maximum p-ratio (VS-MPR), was calculated. The VS-MPR is interpreted as the maximum possible odds in favor of the alternative hypothesis over the null hypothesis (a type of Bayes Factor). Third, because variables did not follow a normal distribution, one-tailed Mann-Whitney U-tests were computed to test the hypothesis that participants who selected HIERARCHY would show higher scores on RNT measures (PTQ and PSWQ-11) than participants who selected COMPARISON. COORDINATION was excluded from this analysis because only two participants selected it. Finally, the content of RNT triggers was categorized to explore main topics across participants. Two authors independently observed all diagrams depicted by participants and independently produced topic categories for RNT. They then agreed on the name and number of topics. Subsequently, two independent raters classified each diagram into one of the categories. In cases of discordance in categorization (which occurred in only 17.17% of cases), raters were asked to agree on selecting only one category for each depicted diagram.

Results

One participant discontinued participation after the MINI administration due to a personal emergency. Therefore, 99 participants completed the entire study. Participants showed mean scores within the range of nonclinical samples in other studies conducted in the authors' laboratory.

Figure 5 presents the percentage of participants who selected COORDINATION (2.02%, 95% CI [0.2%, 7.1%]), COMPARISON (18.18%, 95% CI [11.1%, 27.2%]), and HIERARCHY (79.80%, 95% CI [70.5%, 87.2%]). The Multinomial Test showed statistically significant results (χ²(2) = 100.06, p < .001). The VS-MPR value was 3.93 × 10¹⁹, indicating that the observed data provide overwhelming evidence in favor of the alternative hypothesis of a different percentage of diagram selection from the 33.33% that would be expected by chance.

Participants who selected HIERARCHY showed statistically significantly higher scores on the PSWQ-11 (W = 466, p = .012, d = 0.46) and the PTQ (W = 427.5, p = .004, d = 0.63) compared to those who selected COMPARISON, with medium effect sizes. The VS-MPR values were 7.13 and 15.68 for the PSWQ-11 and PTQ, respectively. These values indicate that the observed data moderately (PSWQ-11) and strongly (PTQ) support the alternative hypothesis that participants who selected HIERARCHY would show a higher tendency to engage in RNT than those who selected COMPARISON.

Six main topic categories for RNT were identified in the diagrams. Failure was the most prevalent category, presented by 56.6% of participants, encompassing fear of failing in important goals, feelings of being a failure, lack of confidence in oneself, and self-blame for committing errors. Loneliness/Social criticism was the second most prevalent category at 19.2%, involving fear of being alone, feelings of loneliness, fear of criticism, and thinking about others' judgments. Fear of thoughts/emotions/madness was the third category at 12.1%, involving fear of certain feelings and thoughts and fear of losing control and going mad. Need to control/perfection was the fourth category at 9.1%, involving needing to have everything under control, intolerance of uncertainty, and needing to be perfect. The world/Life is unfair and Fear of death/illness were the least prevalent categories, each at 8.1%, involving feeling that things should work differently and fears related to one's own or a loved one's death or illness.

Discussion and conclusions

The results of this study are consistent with the RFT conceptualization that triggers for RNT are usually hierarchically organized around a superordinate trigger that sustains the remaining ones. The theory suggests that all types of relational framing are involved in deriving self-contents, but hierarchical framing appears to be a key aspect in the development and expansion of self-contents. As the individual develops fluency in hierarchical framing, many actions are hierarchically related to goals, and these goals to more abstract positive reinforcers called values. The individual then learns to derive thoughts and feelings with appetitive functions when framing in coordination the deictic "I" with the achievement of or advancement in those goals linked to values. Simultaneously, the individual also learns to derive aversive thoughts when the deictic "I" is framed in opposition to those goals and values. This analysis presents thoughts with appetitive and aversive functions as two sides of the same coin: one side constituted by a hierarchical network of appetitive thoughts linked to personal values, and the other side formed by a hierarchical network of thoughts that acquired aversive functions because they are framed in opposition to the other network.

This study was a first attempt to initiate an empirical analysis of the relationships among RNT triggers. After conducting an assessment procedure that included selecting individuals' most frequent RNT triggers, ninety-nine participants were presented with three diagrams depicting ways RNT triggers could be organized (COORDINATION, COMPARISON, or HIERARCHY). The majority of participants selected HIERARCHY (79.80%) as the diagram best describing them, while COMPARISON was selected by 18.18% and COORDINATION by only 2.02%. Participants who selected HIERARCHY obtained higher scores on RNT than those who selected COMPARISON. This indicates that an overwhelming majority of participants showing at least an average level of RNT selected the HIERARCHY diagram.

The authors acknowledge several limitations. First, this study is an initial step toward analyzing how RNT triggers might be related; however, it is only descriptive, and future studies should experimentally analyze how RNT triggers are built in learning history. Second, all participants were undergraduates; future studies might replicate these findings with more diverse samples in terms of age and educational level. Third, the procedure used depends on participants' ability to provide functional descriptions of their behavior; future studies might develop alternative ways that depend less on this ability.

In conclusion, the preliminary results obtained in this study are consistent with the RFT conceptualization that RNT triggers are hierarchically organized around an overarching trigger that sustains the remaining ones. These results might have relevant clinical implications. If RNT triggers are hierarchically related, interventions directed at altering the discriminative functions of the trigger at the top of the hierarchy should also alter the functions of triggers at lower levels of the hierarchy, which might strengthen the intervention's effect. This idea is present in two recent clinical studies that have shown that very brief protocols based on acceptance and commitment therapy can have very large effect sizes in treating emotional disorders.

Significance and contribution

This study provides the first empirical evidence of how triggers for repetitive negative thinking are functionally organized from a behavioral perspective. By demonstrating that the majority of participants perceive their RNT triggers as hierarchically organized (with more central/abstract triggers sustaining more concrete ones), the study validates a conceptual model of problematic self-content organization. The findings have important implications for understanding transdiagnostic processes in psychopathology, suggesting that understanding the structure of cognitive networks could inform the design of interventions aimed at modifying functions at specific hierarchical levels.