ارزیابی یک روش کوتاه آموزش روانی 4 جلسه ای برای افراد با نگرانی بالا مبتنی بر فرضیه خلق و خوی به عنوان ورودی

Abstract Background & objectives Given the ubiquity of worrying as a consuming and distressing activity at both clinical and sub-clinical levels, it is important to develop theory-driven procedures that address worrying and allow worriers to manage this activity. This paper describes the development and testing of a psychoeducation procedure based on mood-as-input hypothesis, which is a transdiagnostic model that describes a proximal mechanism for perseverative worrying. The study used nonclinical participants meeting IAPT criteria indicating GAD symptomatology. Methods In 4 sessions, participants in experimental groups received psychoeducation about the basic principles of the mood-as-input hypothesis and received guidance on how to identify and change worry-relevant goal-directed decision rules and negative moods. Participants in the psychoeducation conditions were compared with participants in a befriending control group. Results Psychoeducation about the model significantly reduced PSWQ scores at follow-up compared with the befriending control condition (a between-groups large effect size, Cohen's d = 1.05), and the homework tasks undertaken by the psychoeducation groups raised mood and reduced worry immediately. At follow up 48.2% of participants in the psychoeducation groups were below the recommended cut-off for identifying GAD symptomatology compared with 20% of participants in the control condition. Limitations This study was conducted on a small sample, high-worry student population, without a formal diagnosis. Conclusions This brief, low-intensity procedure is potentially adaptable to online or self-help procedures, and can be integrated into fuller cognitive therapy packages.

1. Introduction Worry is a consuming and distressing activity at both clinical and sub-clinical levels across a range of anxious psychopathologies. As such, it would seem important to develop theory-driven procedures that specifically address worrying and allow worriers to manage this activity. Compared with non-therapy controls, generic cognitive therapy (CT) techniques (i.e. any psychotherapeutic approach that is founded on a theory which aims to modify the cognitions that are deemed to play an important role in maintaining symptoms – see Hanrahan, Field, Jones, & Davey, 2013) appear to be effective in reducing pathological worry for diagnosable disorders such as Generalized Anxiety Disorder (GAD), but are still associated with arguably modest recovery rates of 57% at 12-months follow-up (Hanrahan et al., 2013). Additional therapeutic procedures may be required to boost recovery rates. Given that an effective model for the successful amelioration of pathological worrying is likely to include elements from many theoretically valuable approaches (see Hanrahan et al., 2013), the aim of the present study was to test the effectiveness of a psychoeducation procedure based on a further theoretical approach to pathological and perseverative worrying, namely the mood-as-input (MAI) model (Meeten & Davey, 2011). The mood-as-input hypothesis views decisions about whether to continue or terminate a task as based on interactions between the individual's ‘stop rules’ or decision rules for the task (i.e. what rules have been explicitly or implicitly deployed to define the goals of the task) and the real-time availability of information about whether those goals have been met (see Meeten & Davey, 2011 for a review). The hypothesis argues that perseverative activities such as worrying are frequently associated with goal-oriented decision rules that specify that the task must be completed as thoroughly and extensively as possible (known as “as many as can” stop rules or decision rules). However, the mood-as-input hypothesis specifies that an individual's concurrent mood is an important source of information by which goal-achievement is assessed. When applied to excessive or pathological worrying, the mood-as-input hypothesis predicts that worriers begin worrying by deploying goal-directed “as many as can” decision rules specifying that the task must be completed as thoroughly as possible, but the worrier's negative mood provides information that this has not been achieved, so worrying continues. These predictions have been substantiated in a number of analogue studies of worrying (see Meeten & Davey, 2011; for a review), and suggest that procedures designed to both identify and change goal-directed “as many as can” decision rules or alleviate negative mood should have the effect of reducing worry perseveration and severity. A consequence of this model is that designing an intervention that can shift a worrier away from the use of goal-directed “as many as can” stop rules and also develop strategies for managing negative mood will both help to alleviate the length and frequency of perseverative worry bouts. The present paper describes the results of a psychoeducation procedure based on the mood-as-input model for excessive worrying in participants experiencing clinically-significant levels of worry. In a 4-session procedure, this study aimed to provide psychoeducation to participants about the basic principles of the mood-as-input hypothesis, provide guidance on how to identify worry-relevant goal-directed decision rules and negative moods, and provide advice about how to change their default decision rules and manage their moods. Participants in psychoeducation conditions were predicted to score significantly lower on Penn State Worry Questionnaire (PSWQ) scores than a befriending control group (Sensky et al., 2000) at the end of the 4-session procedure and at a 4-week follow-up. In particular, analyses were undertaken that would determine whether (1) psychoeducation to the mood-as-input model in itself leads to a reduction in worry, and (2) mood and stop rule interventions (following psychoeducation) have a greater effect than psychoeducation alone.

3. Results The characteristics of the groups are shown in Table 1. Mean baseline PSWQ scores for the four groups were MAI-1 65.10 (SD 5.13), MAI-2 67.30 (SD 7.35), MAI_Bf 67.10 (SD 4.43), Bf 62.30 (SD 5.38). A one-way ANOVA indicated that there was no significant difference between the groups at baseline, F(3, 36) = 1.68, p = .19. Table 1. Group characteristics and PSWQ scores. MAI-1 – psychoeducation plus MAI exercises (mood exercises first, followed by decision rule exercises); MAI-2 – psychoeducation plus MAI exercises (decision rule exercises first, followed by mood exercises), MAI+Bf – psychoeducation followed by befriending; Bf – befriending only. MAI-1 MAI-2 MAI+Bf Bf Number of participants 10 9 10 10 Number of males 0 1 1 1 Age of participants 20.60 (1.17) 20.80 (1.23) 21.40 (1.58) 20.20 (0.92) PSWQ score at baseline 65.10 (5.13) 67.30 (7.35) 67.10 (4.43) 62.30 (5.38) PSWQ score at outcome 60.40 (9.13) 63.44 (10.05) 64.40 (4.95) 64.80 (4.16) PSWQ score at follow-up 59.80 (7.74) 58.89 (9.47) 62.90 (7.85) 63.80 (4.39) PSWQ difference score: baseline minus outcome −4.70 (7.20) −4.67 (7.70) −2.70 (4.40) 2.50 (5.54) PSWQ difference score: baseline minus follow-up −5.30 (7.42) −9.22 (8.41) −4.20 (5.96) 1.50 (6.20) Table options 3.1. Penn State Worry Questionnaire A mixed two-way 4(Group) × 3(Time: baseline, outcome, and follow-up) ANOVA was conducted, and the descriptive statistics are reported in Table 1. There was a significant main effect of Time on PSWQ scores, F(1.84, 64.34) = 9.98, p < .001, ηp2 = .22, but no significant main effect of Group, F(3, 35) = 0.44, p = .73, ηp2 = .04. The main effect of Time is clarified by a significant Group × Time interaction, F(5.52, 64.34) = 3.04, p = .01, ηp2 = .21, indicating a large effect ( Stevens, 2002). A series of one-way repeated measures ANOVAs were conducted with the data split by group in order to examine how each group varied across the three time points (baseline, outcome, and follow-up) on the PSWQ. The repeated measures ANOVAs indicated that there was a significant effect of time in Groups MAI-1 (p = .04), MAI-2 (p = .004) and MAI-Bf (p = .04). When a Bonferroni-correction is applied to control for multiple ANOVAs, p < .0125 is significant, thus MAI-2 showed a significant change in PSWQ scores across time, and there is a trend indicated in groups MAI-1 and groups MAI-Bf. Importantly, the control group, Bf, did not show a significant effect of time on PSWQ scores (p = .36). As is shown in Table 1, the three groups who underwent MAI psychoeducation (Groups MAI-1, MAI-2 and MAI_Bf) showed mean decreases in PSWQ scores from baseline to outcome, and these decreases are maintained to follow-up. In contrast, the befriending group (Group Bf) showed an increase in PSWQ scores from baseline to outcome, and from baseline to follow-up. 3.2. PSWQ difference scores To further examine the nature of the interaction, difference scores were computed between baseline and outcome, and baseline and follow-up (see Table 1 for the PSWQ difference score descriptive statistics). Outcome: A one-way ANOVA indicated that there was an effect of Group on the PSWQ ‘baseline minus outcome’ difference scores F(3, 35) = 2.85, p = .05, ηp2 = .20. Planned contrasts indicated that the MAI-psychoeducation groups (Groups MAI-1, MAI-2 and MAI_Bf) significantly differed from the befriending group (p = .008). There was no significant difference between psychoeducation-only (Group MAI_Bf) compared to psychoeducation-plus-techniques (Groups MAI-1 and MAI-2) (p = .43), or between the two psychoeducation-plus-techniques groups (Groups MAI-1 and MAI-2) (p = .99). Given this, a comparison was also conducted to compare difference scores between the three groups containing psychoeducation (MAI-1, MAI-2 and MAI-Bf) collapsed together with the befriending only control group (Bf). A significant difference was found, t(37) = 2.87, p = .007, d = 1.05. This represents a large effect size ( Cohen, 1992). Follow-up: A one-way ANOVA indicated an effect of group on the PSWQ ‘baseline minus follow-up’ difference scores, F(3, 35) = 3.81, p = .02, ηp2 = .25. Again, planned contrasts were conducted. These indicated that there was a significant difference between the MAI-psychoeducation groups (Groups MAI-1, MAI-2 and MAI_Bf) and the befriending group (p = .005). There was no significant difference between psychoeducation-only (Group MAI_Bf) compared to psychoeducation-plus-techniques (Groups MAI-1 and MAI-2) (p = .27), or between the two psychoeducation-plus-techniques groups (Groups MAI-1 and MAI-2) (p = .23). 3.3. Understanding of the MAI model Changes in knowledge about the mood-as-input hypothesis as a result of the psychoeducation procedure were assessed through a quiz testing knowledge of the mood-as-input model. There was a significant difference in the performance of the groups on the quiz, F(3, 35) = 7.58, p < .001, ηp2 = .39. The three groups who received psychoeducation scored significantly higher than the befriending control group (p < .001) (Group MAI-1, M = 8.50, SD = 1.27; MAI-2, M = 8.44, SD = 2.30; MAI+Bf, M = 8.80, SD = 0.92; Bf, M = 5.70, SD = 1.89). There was no significant difference between the three groups who received psychoeducation (Groups MAI-1, MAI-2 and MAI_Bf) (ps > .62). This can be considered as a manipulation check indicating that the psychoeducation manipulation was successful. Although there was a significant difference between psychoeducation groups and the befriending group in terms of knowledge of the mood-as-input model, scores on the measure did not correlate with PSWQ difference scores between session 1 and outcome, r = .15, p = .35, or with PSWQ difference score between session 1 and the 4-week follow-up, r = .21, p = 19. 3.4. MAI experiments The lifting mood experiments and the changing decision rule experiments were carried out by groups MAI-1 and MAI-2. In each case, the measures used are detailed in Materials section. Mood and the level of worry were rated using Likert-type scales. 3.4.1. Lifting mood Three-way 3(attempt: 1, 2, 3) × 2(time: pre, post) × 2(condition: mood session first, decision rule session first) ANOVAs were conducted on mood and worry reports during the lifting mood homework task. Descriptive statistics are shown in Table 2. Participants were asked to try out the mood lifting and decision rule exercises three times each, during the week after their session on these techniques. The variable ‘attempt’ represents these three different attempts at lifting their mood and changing their decision rule. The variable ‘time’ codes whether the ratings came from immediately before the experiment or immediately after it. In both analyses, only the main effect of time was significant, ps ≤ .001, ηp2 ≥ .75. Paired sample t-tests between mood ratings pre and post the mood lifting activity indicated that there was a significant difference for all three attempts, with mood levels higher following the mood lift activity, ts ≥ 3.50, ps ≤ .003, rs .40–.64. Paired sample t-tests also indicated that worry levels were significantly lower after the lifting mood activity than before, ts ≥ 4.34, ps < .001, rs .51–.75. Mood was measured on a single scale (−10 = very negative, 0 = neutral, +10 = very positive), thus it is not possible to separate changes in negative and positive mood specifically. There were no significant differences across the three attempts at lifting mood, suggesting that there was no significant change in the participants' ability to perform this technique across the three attempts. Table 2. Mood and worry levels pre- and post-homework task (lifting mood activity/changing decision rule). Homework task Attempt Mooda pre-task Mooda post-task Worryb pre-task Worryb post-task Mood lift 1 −1.89 (4.11) 2.16 (4.17) 5.95 (2.09) 3.84 (1.83) 2 1.53 (3.76) 4.21 (2.88) 4.63 (2.29) 3.00 (1.91) 3 −3.33 (4.06) 2.28 (4.25) 4.92 (1.90) 3.50 (2.20) Change decision rule 1 0.58 (3.81) 2.18 (3.88) 5.32 (1.80) 3.63 (2.24) 2 0.11 (4.09) 2.53 (3.63) 4.95 (1.72) 2.63 (1.21) 3 0.26 (4.57) 0.79 (4.43) 5.21 (1.99) 2.89 (1.94) a −10 (very negative) to +10 (very positive). b 0 (no worry) to 10 (highly worried). Table options 3.4.2. Changing decision rule Similarly, three-way 3(attempt: 1, 2, 3) × 2(time: pre, post) × 2(condition: mood session first, decision rule session first) ANOVAs were conducted on mood ratings and worry ratings during the changing decision rule homework task. Again, in both analyses, only the main effect of time was significant, ps ≤ .001, ηp2 ≥ .73. Paired sample t-tests between mood pre-decision rule change and mood post-decision rule change indicated that there was a significant difference for all three attempts, with a shift to a more positive mood following the change in decision rule, ts ≥ 2.25, ps ≤ .04, rs .22–.63. Descriptive statistics are in Table 2. Paired sample t-tests also examined level of worry before the decision rule change and after the decision rule change, and indicated that there was a significant decrease in worry for all three attempts, ts ≥ 5.88, ps < .001, rs .63–.81. 3.5. Participant expectation One-way ANOVAs examined whether the groups had different expectations of the usefulness of their sessions on managing their worry. There was no significant difference between the groups at baseline, F(3, 36) = 1.94, p = .14, ηp2 = .14. Changes in expectation throughout the course of the experiment were examined by calculating difference scores. Change in expectation did not significantly different across the groups (baseline expectation minus outcome expectation: F(3, 35) = 0.30, p = .83, ηp2 = .03); (baseline expectation minus follow-up expectation: F(3, 35) = 0.59, p = .63, ηp2 = .05). Nor did change in expectation significantly correlate with PSWQ total scores at outcome or follow up, or difference scores, rs ≤ .20, ps ≥ .23, providing little evidence that expectations alone were a significant contributor to change. 3.6. Clinical significance At follow up 48.2% of participants in the psychoeducation groups were below Behar et al.'s (2003) PSWQ criteria of 62 for GAD symptomatology compared with only 20% of participants in the control condition (See Table 3 for minimum and maximum PSWQ scores at follow-up). Jacobson and Truax's (1991) criteria for reliable and clinically significant change were also applied, using the cut-offs calculated by Fisher (2006); i.e. a participant's score on the PSWQ needed to be below 47 at follow-up and to have decreased by at least 7 points from baseline. According to this more stringent approach, only two participants from the psychoeducation groups and none in the control group showed reliable and clinically significant change at follow-up. Table 3. Minimum and maximum PSWQ scores at follow-up shown by group. MAI-1 – psychoeducation plus MAI exercises (mood exercises first, followed by decision rule exercises); MAI-2 – psychoeducation plus MAI exercises (decision rule exercises first, followed by mood exercises), MAI+Bf – psychoeducation followed by befriending; Bf – befriending only. Group Minimum PSWQ score Maximum PSWQ score MAI_1 44 72 MAI_2 38 66 MAI_Bf 52 73 Bf 56 71