Efficacy of acceptance and commitment therapy focused on repetitive negative thinking in fibromyalgia: A randomized multiple-baseline design

Context and Objectives

Fibromyalgia (FM) is a medical condition of unknown origin characterized by chronic and generalized musculoskeletal pain, usually accompanied by stiffness, fatigue, and sleep disturbances. It affects approximately 2.10% of the world population, with a prevalence 4 to 7 times higher in women than in men, and with no significant differences between different ethnic groups. FM diagnosis is typically established in women between 35 and 60 years of age.

Patients with FM experience higher pain intensity than other rheumatological diseases, and this disease has the most significant impact on quality of life in terms of social, family, intellectual, and health outcomes. A frequent comorbid symptom in FM is the presence of psychological symptoms such as anxiety and depression. Additionally, many patients with FM experience sleep disturbances, which constitute a risk factor for FM development.

The goal of FM treatment should be to improve physical function and quality of life because it is considered a chronic condition. According to recent treatment guidelines, active nonpharmacological therapies are the mainstay for FM treatment. Treatment should include supervised exercise programs and cognitive-behavioral therapy (CBT) when emotional symptoms and sleep disturbances are present. Acceptance-based interventions have been increasingly tested in recent years for FM treatment.

A characteristic factor frequently observed in FM patients is repetitive negative thinking (RNT) in the form of worry and rumination. RNT can be considered a variable of special clinical interest in FM treatment. Brief RNT-focused ACT interventions have been recently tested for the treatment of emotional disorders, finding very large effect sizes in reducing emotional symptoms and RNT.

The present study aims to analyze the efficacy of a brief Acceptance and Commitment Therapy (ACT) protocol focused on reducing RNT in four women diagnosed with FM. This is a preliminary investigation to test the effect of a brief, individual, 4-session intervention focused on RNT in FM patients.

Method

Participants

The sample was recruited through social media advertisements and referrals from other professionals. Inclusion criteria were: (a) medical certificate of fibromyalgia diagnosis with at least six months duration; (b) age between 35 and 60 years; (c) scores of 15 or higher on the General Health Questionnaire (GHQ-12); (d) scores above 25 on the Depression, Anxiety and Stress Scale (DASS-21); (e) stable adherence of two or more months to pharmacological treatment (if applicable) and perceived lack of improvement in FM symptoms; (f) commitment not to initiate psychological treatment during the investigation; (g) stable adherence of two or more months to any alternative treatment (acupuncture, yoga, hydrotherapy, etc.) and perceived lack of improvement in FM symptoms.

Exclusion criteria were: (a) high suicidal risk, psychotic disorder, or abuse of psychoactive substances other than analgesics; (b) receiving some type of psychological intervention; (c) not presenting significant improvement trends at baseline.

Four participants met the inclusion criteria. All had FM diagnosis certified by a physician specializing in rheumatology. Their ages ranged from 39 to 55 years. Three of them were single and one was married. All had technical or professional education and were residents of Bogotá. Table 1 presents the sociodemographic data of the participants, as well as their hierarchical triggers for worry/rumination, experiential avoidance strategies, and medical treatments received.

Design

A noncon concurrent, randomized, across-participants, multiple-baseline design was used. Participants were randomized to the intervention phase between the third and sixth week of baseline collection. The independent variable was a brief protocol of approximately 60-minute ACT sessions focused on RNT. Dependent variables were divided into outcome and process measures. Outcome variables were emotional symptomatology, sleep quality, and quality of life. Process variables were pathological worry, cognitive fusion, and valued actions.

Intervention

The intervention consisted of a brief protocol of 4 individual sessions of approximately 1 hour each focused on RNT.

Session 1 included: introduction of the intervention rationale; functional analysis of the psychological inflexibility pattern with central focus on worry/rumination identifying hierarchical triggers, main contents in the repetitive thought chain, and additional experiential avoidance strategies; identification and amplification of the consequences of the inflexible pattern and opening of a flexible alternative using a physical metaphor (pushing/letting go of worries); discrimination training of the worry/rumination process and its contents, contact with its consequences, and training in distancing from triggers with a physical metaphor (circling the chair); Audio 1 (exercise aimed at developing the skill to differentiate between engaging in RNT or taking distance from it while choosing a valued direction).

Session 2 included: review of progress and difficulties experienced since Session 1; training in multiple examples in distancing from worry/rumination through (a) putting zoom to thoughts exercise, (b) free association exercise, (c) conscious fantasizing and worrying exercise, and (d) "I can't possibly..." exercise; Audio 2 (exercise of observing thoughts in balloons from a hierarchical perspective).

Session 3 included: review of progress and difficulties experienced since Session 2; observer exercise (modified to identify triggers of worry/rumination and contact with values); clarification of values and identification of valued actions (garden metaphor); Audio 3 (life movie exercise - contact with values and adversities).

Session 4 included: review of progress and difficulties experienced since Session 3; general review of the work done in the intervention; exploration of the consequences of monitoring physical sensations as habitual behavior; experiential exercise of monitoring symptoms versus not monitoring; identification of valued actions; closing of the intervention.

Outcome Measures

Depression, Anxiety and Stress Scale - 21 (DASS-21; Ruiz et al., 2017b). The DASS-21 is a scale that measures the main emotional symptoms: depression, anxiety, and stress. It consists of 21 items describing negative symptoms evaluated on a four-point Likert-type scale (3 = it has happened to me a lot, or most of the time; 0 = it has not happened to me).
General Health Questionnaire - 12 (GHQ-12; Ruiz et al., 2017a). The GHQ-12 is a self-report that measures mental health status in recent weeks. The scale consists of 12 items answered on a Likert-type scale (0-1-2-3) and has six items in the positive direction and six in the negative direction. Scores equal to or greater than 12 indicate the possibility of experiencing an emotional disorder.
36-Item Short Form Survey (SF-36; García et al., 2014). The SF-36 is a questionnaire that measures quality of life in relation to health, applicable to patients with different medical conditions and the general population. It consists of 36 items that measure both positive and negative health. The instrument contains eight subscales: Physical Function, Physical Role, Body Pain, General Health, Vitality, Social Function, Emotional Role, and Mental Health.
Sleep Quality Report. Four questions were used to measure sleep parameters: sleep latency, maintenance, degree of rest the following morning, and sleep quality. The question on sleep latency was, "In general, during the last week, how long did it take you to fall asleep?" Response options were: (a) it took little or no time to fall asleep, (b) it took me a while to fall asleep, but I managed to fall asleep between half an hour and an hour, and (c) it took me a long time to fall asleep (more than an hour). The question on sleep maintenance was, "On average, how many hours did you sleep during the night?" Response options were: (a) less than one hour, (b) one hour, (c) 2-3 hours, (d) 4-5 hours, (e) 6-8 hours, (f) more than 8 hours. The question on the degree of rest was, "On average, how tired did you feel the next morning on a scale of 1-5?" (1 = extremely tired; 5 = very rested and energetic). Finally, sleep quality was measured by the following item, "Overall, on a scale of 1 to 5, how would you rate the quality of your sleep during the past week?" (1 = very bad, 5 = very good).

Process Measures

Penn State Worry Questionnaire - 11 (PSWQ; Ruiz et al., 2018b). The PSWQ assesses the general tendency to worry or worry trait and is widely used to measure the degree of worry characteristic of Generalized Anxiety Disorder. The scale has 11 items answered through a Likert-type scale from 1 to 5 (5 = much; 1 = not at all).
Cognitive Fusion Questionnaire (CFQ; Ruiz et al., 2017c). The CFQ is a questionnaire consisting of seven items answered on a seven-point Likert scale (7 = always true; 1 = never true) that measures cognitive fusion. Cognitive fusion refers to becoming entangled in thoughts, evaluations, judgments, and memories, behaving accordingly.
Valuing Questionnaire (VQ; Ruiz et al., 2022). The VQ consists of 10 items that evaluate how much one has acted in accordance with personal directions during the last week. The VQ is answered on a Likert-type scale with seven response options (6 = completely true; 0 = not at all) and consists of two subscales: Progress in valued directions and Obstructing behaviors in those directions.

Data Analysis

Data were analyzed both individually and globally for the four participants. For intrasubject analysis, the nonparametric Tau-U test (Parker et al., 2011) was used. This test was computed to perform intrasubject analysis on weekly measures through the calculator http://single-caseresearch.org/calculators/tau-u. This test provides a nonoverlapping effect size between baseline and intervention data. One advantage of Tau-U compared to other nonparametric effect sizes is that it allows correcting for significant trends during baseline. Additionally, Tau-U allows hypothesis testing to determine if there were statistically significant differences between phases. This study presents all effect sizes in the intervention phase as positive, regardless of whether their scores should decrease or increase.

The extent of clinically significant changes was analyzed following a proposal similar to that presented by de Ruiz et al. (2018a). To claim the presence of clinically significant changes required: (a) the Tau-U value to be statistically significant in favor of the intervention phase, and (b) to show a score at follow-up closer to the mean of the nonclinical population than to the clinical population. To test this latter criterion, descriptive data obtained from scale validation studies by Ruiz et al. (2018a) were consulted.

Intrasubject changes in GHQ-12 were analyzed by computing the reliable change index and clinically significant change according to guidelines suggested by Jacobson and Truax (1991). It was established that a 10-point change represents a significant change in the SF-36. Finally, changes in sleep quality parameters were considered significant when they changed in each aspect evaluated.

Between-subjects data analysis was performed by computing the effect size for multiple-baseline designs developed by Hedges et al. (2012). This analysis provides a standardized mean difference that shares the same metric as Cohen's d frequently used in group designs. This allows comparison of effect sizes obtained in single-case experimental designs with those reported in group designs.

Results

Outcome Results

The evolution of DASS-21 scores and each of its subscales can be seen in Figure 1. Visual analysis reveals relatively stable baseline trends for most participants and measures. Although some baselines showed improvement or deterioration, none were statistically significant.

P1 showed the lowest DASS-Total scores at baseline. However, after the intervention, scores tended to decrease to minimal levels. Scores on DASS-Depression showed a spike at the 3-month follow-up, which may have been affected by the hospitalization of a family member in serious condition. Table 3 shows that effect sizes were statistically significant for all variables and all changes were clinically significant. Similarly, P1 showed clinically significant changes on the GHQ-12 during post-treatment and all three follow-ups (see Table 4), with scores below the 12-point cutoff indicating possible presence of emotional disorder. Table 5 shows that this participant also showed significant improvements in sleep latency, tiredness, and sleep quality, although the number of hours slept during the night decreased. Finally, P1 experienced significant increases in 5 of the 8 subscales of the SF-36.

P2 showed very intense emotional symptoms during baseline, mainly in depression and anxiety. However, after the introduction of the intervention, he showed noticeable changes in both. Table 3 shows that effect sizes were statistically significant for all variables except DASS-Stress. However, P2 only experienced clinically significant changes in DASS-Depression because his total and DASS-Anxiety scores were still closer to the average of the clinical population than to the nonclinical population. On the other hand, the participant did show significant changes on the GHQ-12 during post-treatment and all three follow-ups (see Table 4), with scores below the 12-point cutoff indicating possible presence of emotional disorder. Table 5 shows that P2 reported highly significant changes in all sleep parameters assessed. Finally, the participant showed significant improvements in all sleep parameters assessed (Table 5) and in all SF-36 subscales except Vitality (Table 6).

P3 showed high scores on DASS-21 during baseline, especially on the anxiety subscale. The participant showed immediate changes upon introduction of the intervention that were maintained at the 1-month follow-up. However, at the 2-month follow-up, he showed an increase in scores associated with being in a grief period. Scores decreased again at the 3-month follow-up. Changes were statistically significant for all DASS-21 indicators, although clinical significance was not reached for DASS-Stress (see Table 3). GHQ-12 scores followed a similar evolution. The participant experienced clinically significant changes on all measured parameters except at the 1-month follow-up on GHQ-12. Similarly, all sleep parameters showed significant improvements (Table 4). Finally, the participant experienced significant changes in all SF-36 subscales at the 3-month follow-up except Physical Function.

P4 showed high emotional symptoms at baseline, especially depression and anxiety. Scores decreased dramatically at the start of the intervention. All Tau-U values were statistically significant, with clinically significant changes in all subscales (Table 3). A similar conclusion can be reached by looking at GHQ-12 scores in Table 4. Also, Table 5 shows that the participant improved in all sleep parameters assessed. On the SF-36, the participant experienced significant improvement in 5 of the 8 subscales, although his score decreased in Emotional Role.

Process Results

Figure 2 presents the evolution of scores on process measures. All four participants showed high scores on pathological worry (PSWQ) during baseline, mainly P3 and P4. All participants showed statistically and clinically significant changes after the introduction of the intervention.

Cognitive Fusion (CFQ) scores were also high during baseline. P1, P2, and P4 showed statistically and clinically significant changes during the intervention. P3 showed declining scores during the follow-up period and ended with a low score at the end of follow-up. However, given the slow evolution of scores, the Tau-U value did not reach a statistically significant level.

The evolution of scores on valued actions (VQ) was less clear. Participants showed average scores similar to the nonclinical population on VQ-Progress. Only P2 showed clinically significant changes in this subscale of the VQ. For VQ-Obstruction, P1 and P3 had baseline scores close to the average of the nonclinical population. These participants showed no change in this variable. P2 and P4, who showed clinical scores at baseline, showed clinically significant changes.

Discussion and Conclusions

This study aimed to analyze the efficacy of a 4-session Acceptance and Commitment Therapy (ACT) protocol focused on reducing repetitive negative thinking in relation to emotional symptoms, sleep quality, and health-related quality of life. The four participants showed clinically significant changes in at least one of the DASS-21 subscales. All participants showed clinically significant changes in DASS-Depression and three in DASS-Anxiety. DASS-Stress scores were not as high and relevant in this study, although participants also showed clinically significant changes. As for process measures, all participants showed statistically and clinically significant changes in pathological worry and three of them in cognitive fusion. However, changes in valued actions were clearly more modest: only one participant showed clinically significant changes in VQ-Progress and two in VQ-Obstruction. This difference in intervention effect could be related to the fact that participants did not show excessive changes in values according to VQ scores.

Comparable effect sizes across designs were large and statistically significant for DASS-Total, DASS-Depression, pathological worry, cognitive fusion, and obstruction in values. These results are promising and compare very well with the weighted effect sizes found in recent meta-analyses of ACT efficacy in FM (Haugmark et al., 2019). This is important because brief psychological interventions are especially appropriate for FM patients, given the frequent difficulty and cost of treatment involvement in interventions (Gómez-Pérez et al., 2020).

The high effect of the protocol in reducing pathological worry and cognitive fusion is similar to previous studies that analyzed the efficacy of brief RNT-focused ACT interventions (Dereix-Calonge et al., 2019; Ruiz et al., 2016, 2018a, 2019, 2020a, 2020b). These studies have found very large effect sizes in reducing emotional symptoms and RNT. The current results preliminarily test the effect of a brief, individual, 4-session protocol focused on RNT in FM patients.

It is noted that a significant process of change of the intervention could be the reduction of RNT, as found in Dereix-Calonge et al. (2019). However, a mediation analysis of the intervention effect could not be conducted because the number of assessment points was not sufficient. Future studies could make use of more intensive assessment through ecological momentary assessment. Additionally, the large effect size found in the reduction of RNT is especially important in this context, given its relevance to FM compared to other chronic pain syndromes (Ricci et al., 2017).

Some limitations of the study merit attention. First, the sample used was small and further replications are needed to confirm the efficacy of brief RNT-focused ACT interventions. Second, the effect of the intervention was only assessed through self-report measures. Future studies could analyze the effect of the intervention in a clinical interview and behavioral tests of pain tolerance. Third, pain intensity and direct interference were not measured as is typical in studies investigating the efficacy of psychological interventions in chronic pain conditions. Future studies should analyze the effect of RNT-ACT interventions on pain intensity and interference and their relationship with RNT using ecological momentary assessment. Fourth, because the recruitment process was spread over several weeks, a nonconcurrent multiple-baseline design was conducted. These designs have lower internal validity than concurrent designs, although additional limitations would not be particularly problematic in this type of study (see discussion in Ruiz et al., 2018b). Fifth, all participants were women, which reduces the generalizability of the results. However, this bias in recruitment is understandable given the higher prevalence of FM in women.

In conclusion, this study should be considered a preliminary investigation of the effect of brief RNT-focused ACT interventions in patients suffering from FM. This type of intervention holds promise, and the conduct of further studies is warranted. Furthermore, this study extends the applicability of brief RNT-ACT focused interventions to health conditions and has found results similar to previous studies that analyzed their effect on emotional disorders.

Significance and contribution

This study contributes important findings on the efficacy of a brief Acceptance and Commitment Therapy protocol focused on repetitive negative thinking for treating fibromyalgia. Demonstrating that a 4-session intervention produces large effect sizes in emotional symptoms, sleep quality, and quality of life suggests that brief psychological interventions centered on specific processes can be especially viable for patients with chronic conditions facing adherence and cost barriers. The findings extend the applicability of brief contextual behavioral interventions to health conditions beyond emotional disorders.

Potential Mechanism of Change: The finding that RNT was reduced as a result of the intervention suggests that this is a key mechanism in the treatment of FM through ACT, aligning with theory of the psychological inflexibility model.