The purpose of this study is to investigate if an internet-based metacognitive therapy reduces negative metacognitions and if reductions negative metacognitions mediates reductions in worry.

The aim is primary mechanistic i.e. we want to investigate if a clear change in negative metacognitions in one group (internet-based metacognitive therapy; I-MCT) relative to another (waiting list) mediates subsequent reductions in worry.

Our hypotheses are the following:

1. I-MCT reduces both negative metacognitions (Beliefs about uncontrollability and danger of worry) and worry from baseline to week 10,

2. reductions in negative metacognitions will significantly mediate subsequent reductions in worry.

   Additionally we hypothesize that patients who score low at baseline in negative metacognitions will not show this process pattern i.e.

3. patients who score low at baseline in negative metacognitions will benefit less from treatment (moderator hypothesis)

   and consequently

4. will not show the same mediation response as stipulated in hypothesis 2 (moderated mediator hypothesis).

Trial Design: Randomized controlled trial with waitlist control. Duration: Ten weeks Primary Endpoint: Change in worry symptoms and negative metacognitions from baseline to Week 10. Long term follow-up is also investigated (baseline to 6-months after treatment completion and baseline to 12-months after treatment completion).

Efficacy Parameters: Penn State Worry Questionnaire (PSWQ) and the negative metacognitions subscale (negative beliefs about uncontrollability of thoughts and danger) of the Meta-Cognitions Questionnaire 30 items.

Safety Parameters: Adverse Events is assessed at week 10.

Description of Trial Subjects: Patients > 18 years old with a PSWQ score more than 56 points Number of Subjects: Anticipated 140

Analysis plan:

To address the hypotheses in the study growth modeling analysis using the expectation-maximization algorithm and maximum-likelihood estimation will be employed. The treatment effect on negative metacognitions and worry will examined by comparing average growth rates between treatment and control group over the assessment period. Growth modeling for longitudinal mediation will be employed to test (a) the overall mediated effect on worry using individual trajectories of change on the proposed mediator (i.e., negative metacognitions) and outcome (i.e., worry), and (b) mediated baseline by treatment moderation effect using the initial assessment of negative metacognitions as the moderator of the indirect (i.e., mediated) and direct effect of treatment. Competing mediator is depressive symptoms using the PHQ-2.

Inclusion Criteria:

• Outpatients
• ≥ 18 years
• Situated in Sweden
• Informed consent
• PSWQ score more than 56 points

Exclusion Criteria:

• Substance dependence during the last six months
• Post traumatic stress disorder, bipolar disorder or psychosis
• Symptoms better explained by axis 2 diagnosis (e.g. autism or borderline personality disorder)
• MADRS-S score above 25 points
• Psychotropic medication changes within two months prior to treatment that could affect target symptoms.
• Received metacognitive therapy for pathological worry the last 2 years.