Do worry and metacognition drive anxiety during therapy?
In patients treated for stubborn anxiety disorders, worry and constantly scanning for threat sat at the centre of the week-to-week web of symptoms in both therapies, more than the specific catastrophic thoughts that CBT targets.
Source
Metacognitive Therapy Versus Cognitive Behavioral Therapy:A Network Approach
Study at a glance
- Design
- Other — Secondary analysis of weekly questionnaire data from a randomised trial; multilevel vector autoregressive (mlVAR) network models estimated separately for each therapy arm
- N
- N=74 · 74 patients who started treatment (38 CBT, 36 MCT); 67 completed all sessions
- Population
- Adults with treatment-resistant PTSD, social phobia or panic disorder at a specialised inpatient anxiety unit in Norway, heavily comorbid
- Outcome
- Structure and centrality of temporal, contemporaneous and between-person networks linking 11 weekly-rated anxiety, depression, cognition and metacognition items
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The authors reused data from a trial in which 74 patients with treatment-resistant anxiety disorders were randomised to metacognitive therapy (MCT) or disorder-specific cognitive behavioural therapy (CBT). Every week patients rated questionnaire items, and the authors picked 11 of them covering anxiety and depression symptoms, catastrophic cognitions, worry, threat-monitoring and the belief that thoughts are uncontrollable. They then fitted network models, separately for each therapy, showing which items predicted which others from one week to the next, within the same week, and across people.
What they found
In the MCT network, the belief that one's thoughts are uncontrollable predicted later threat-monitoring, and worry was the most central node. The CBT network had more connections; a bodily symptom (heart pounding) was central, but worry, sleep problems and threat-monitoring also had high influence. Neither catastrophic cognition had strong outgoing influence in the CBT network, which runs against what CBT theory would predict. Across all network types, worry and threat-monitoring were consistently central, which fits the self-regulatory executive function (S-REF) model behind MCT.
The limits
What it doesn't show
The two therapies' networks were never statistically compared with each other, so the claim that they differ rests on visual inspection. No correction was made for the many significance tests, there was no replication sample, and network stability could not be checked with the sample available. The items were picked from general questionnaires, so two anxiety-inventory items stand in for 'catastrophic cognitions' and a single belief item stands in for 'metacognition', which limits how well each theory's mechanisms were measured. Network edges show which items predict one another over time, but that does not prove one symptom causes another.
Key terms
- Network approach to psychopathology
- The view that a disorder is a web of symptoms that directly influence one another, rather than symptoms all being caused by one hidden underlying disorder.
- Metacognitive therapy (MCT)
- A therapy that targets beliefs about thinking itself (for example 'my worry is uncontrollable') and the habits of worry, rumination and threat-monitoring, rather than the content of individual thoughts.
- Temporal network
- A network in which an arrow means that a person's level on one item this week predicts their level on another item next week, after accounting for last week's values.
- Node strength / out-strength
- Centrality measures summing a node's connections; high out-strength means changes in that node tend to spread to many others, making it a candidate treatment target.
- Threat-monitoring
- Keeping attention fixed on possible dangers, one of the maintaining processes in the metacognitive model.
Flashcards
0 of 11 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
In a temporal network from an mlVAR model, what does an arrow from worry to shakiness mean?
Common questions
Does this study show MCT works better than CBT?
No. It describes how symptoms and mechanisms were connected within each therapy. The authors note that an earlier paper on the same trial found MCT more effective, but this analysis did not test treatment effectiveness.
Why separate within-person from between-person effects?
Therapists care about whether a change in a given patient's worry is followed by a change in that same patient's symptoms. Between-person associations only say that people who worry more on average also have more symptoms on average, which can differ from the within-person pattern.
Why might lack of interest have been central in an anxiety sample?
The patients were highly comorbid, with several diagnoses each on average, so depressive symptoms such as loss of interest were common and could link to many other symptoms.
More on Anxiety