Ethics Statement
This research was conducted in accordance with the Helsinki Declaration (1989) and was approved by the Ethics Committee of the Faculty of Psychology at the University of Warsaw. All subjects were informed about the study and provided informed consent.
Recruitment
The study was conducted in June 2020. It was a part of a larger project aimed at the evaluation of the effectiveness of online mindfulness therapeutic intervention for people experiencing adjustment disorder due to the COVID-19 pandemic. During the current study assessment, Poland had passed through the first wave of COVID-19 and restrictions were being loosened.
Participants were recruited via the Internet. Advertisements were posted on Facebook’s psychological support groups, psychological fan pages promoting well-being, students’ groups, Instagram’s lifestyle, and psychological accounts, and in addition, the invitation to the study was sent in some university newsletters.
On the dedicated platform www.covid.stress-less.pl, as well as in the advertisements, it was explained that the study is designed for people experiencing emotional difficulties related to the COVID-19 pandemic and its consequences and that registering for the study didn’t guarantee participation in the intervention since an individual may not fulfil inclusion criteria. Acceptance of informed consent was mandatory to take part in the study.
Participants
The webpage initially registered 790 people; 564 participants of whom filled in all the obligatory screening questionnaires. Participants were selected if they met the inclusion criteria: diagnosis of AjD (higher than cut-off score (47,5) in ADNM-20 (Lorenz, Bachem & Maercker, 2016)) and they met criteria of emotional disorder (a cutoff score of ≥ 8 for both scales (anxiety and depression) in HADS, Zigmond & Snaith, 1983). These criteria led to a final sample of 308 participants. Characteristics of the sample are shown in Table 1.
----insert table 1----
Measures
Background information of participants. Participants answered questions about their birth year, sex, education level, their professional and financial situation, marital status, the current level of their socioeconomic status related to the COVID-19 pandemic, and any indication of the use of therapeutic/pharmacological methods in the past or present.
Adjustment disorder (The Adjustment Disorder New Module – 20, ADNM-20; Lorenz, Bachem & Maercker, 2016). The ADNM-20 questionnaire is used to assess adjustment disorder. It consists of two parts – a stressor list (which includes a range of acute and chronic life events of the past two years) and an item list (which evaluates the symptoms in response to the most distressing events). Taking into consideration the aim of this research, we added an extra item to a standard list of stressors: the COVID-19 pandemic. Participants responded on a Likert-type scale about how often they have experienced adjustment disorder symptoms in the past two weeks (from 0 – never to 3 – often). The ADNM-20 questionnaire consists of six subscales (preoccupation, failure to adapt, avoidance, depressive mood, anxiety, and impulse disturbance), with preoccupation and failure to adapt as core symptoms, and avoidance, depressive mood, anxiety, and impulse disturbance as accessory symptoms of adjustment disorder diagnosis.
The internal consistency of this questionnaire is high (Cronbach’s α = 0.94), as well as the core symptoms summed in one scale (Cronbach's α = 0.90) and separately (Cronbach's α = 0.88 for preoccupation and Cronbach's α = 0.80 for failure to adapt). The subscale for accessory symptoms also showed a high internal consistency (Cronbach's α = 0.89).
Depression (Patient Health Questionnaire-9; PHQ-9; Kroenke et al., 2001; Polish adaptation - Kokoszka, A., Jastrzębski, A., & Obrębski, M., 2016). PHQ-9 is used to assess the level of depressive symptoms. It consists of 9 statements derived from the DSM-IV criteria of depressive disorder and an additional statement regarding the severity of existing symptoms in daily life. We also added an extra question: “To what extent these behaviors and feelings are related to the COVID-19 epidemic?”. In case participants indicated any of the symptoms, they had to answer this question on a scale from 1 (definitely not related) to 5 (definitely related). The PHQ-9 in the current study showed a good internal consistency (Cronbach's α = 0.83).
Anxiety (Generalized Anxiety Disorder Scale-7, GAD-7; Spitzer, Kroenke, Williams Williams, et al., 2006). GAD-7 is a measure used to assess the level of anxiety symptoms in generalized anxiety disorder (GAD) defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV). It is a scale that contains 7 items that relate to characteristics of GAD (feeling anxious, worrying too much, having difficulties relaxing etc.). The psychometric properties of this questionnaire are strong. The internal consistency of this scale was found to be very good (Cronbach α = .92), as well as test-retest reliability (intraclass correlation = 0.83).
Anxiety and depression (Hospital Anxiety and Depression Scale; HADS; Zigmond, Snaith, 1983; Polish adaptation Nezlek, Rusanowska, Holas, Krejtz, 2021). HADS is a self-report measure used to assess depressive and anxiety symptoms. The questionnaire is built up of 14 items, 7 related to anxiety, e.g. “I feel tense or wound up” and 7 related to depression, e.g. „I look forward with enjoyment to things”. Factor analyses of the two subscales show a two-factor solution in good correspondence with the HADS subscales for Anxiety (HADS-A) and Depression (HADS-D), respectively. Cronbach's alpha for HADS-A varies from .68 to .93 (mean .83) and for HADS-D from .67 to .90 (mean .82).
Self-compassion (Self-Compassion Scale Short Form, SCS-SF; Raes et al., 2010; P olish adaptation Holas et al., accepted). Self-Compassion Scale consists of 26 items, which are grouped into six subscales: self-kindness, self-judgment, common humanity, isolation, mindfulness, and over-identification. The internal consistency of the questionnaire is high (Cronbach α = .92.), and test-retest validity is adequate (with a correlation of .93 for the SCS overall score).
Perceived Health and Life Risk of COVID-19 scale (PHLRC, Gambin et al., 2021). The scale consists of six questions assessing the subjective risk of covid-19 infection, serious adverse health effects and complications due to a coronavirus infection, and a threat to life as a result of an infection. Each of these areas was assessed using two items - one relating to oneself and a second one to loved ones. The six items were rated on a five-point scale from 1 - very low to 5 - very high. Cronbach's alpha was α = 0.92.
Experiential avoidance (The Acceptance and Action Questionnaire, AAQ-II, Bond, et al., 2011). AAQ-II is the most widely used measure of psychological inflexibility and experiential avoidance. Lower scores of AAQ-II are also indicators of psychological flexibility. AAQ-II consists of 7 items (e.g., “I am afraid of my feelings”, “I worry about not being able to control my worries and feelings”) rated from 0 (never true) to 7 (always true) on an 8-point Likert scale. Developed originally as a measure of psychological inflexibility, its items tend to emphasize experiential avoidance and some have suggested that it be interpreted that way (Tyndall et al., 2019). In the current study Cronbach's alpha was α =0.90.
Statistical Analyses
The first step of the statistical analysis was to generate descriptive statistics and to conduct a Pearson correlation analysis between the analysed variables. Next a moderation analysis was performed in which experiential avoidance and self-compassion were analysed as moderators of the relationship between PHLRC and ADNM-20. The moderation was analysed with the use of Hayes macro Process 3.5.3 in the model no.1 (Hayes, 2017). The sample in the current study was equal to 308 participants. Assuming statistical power to be .8 and planning to perform moderation analysis with one moderator and one explaining variable in a single statistical model one needs to encounter the interaction effect equal to at least .03 in terms of Cohen's f2 effect size measure to detect it as statistically significant. According to Cohen (1988) effect size of f2=.02 is to be considered small. Therefore, the sample in the present study is not sensitive enough to detect it. The moderation effect needs to be a bit stronger. As a result of limited statistical power, we formulated hypotheses of the attenuation effect instead of the buffering effect. The possible interaction between the two moderators was assessed in the model no. 3. Johnson-Neyman procedure was applied to interpret the acquired interactions. Further, cluster analysis based on k-means method was performed to extract groups of participants with different profiles of AAQ and self-compassion (IBM SPSS Statistics 28.0). A four cluster solution was used in order to allow for extracting four possible combinations of AAQ and self-compassion, i.e. low levels of both, high levels of both, low level of AAQ and high level of SC, high level of AAQ and low level of SC. In addition, a four cluster solution led to the extraction of four groups that were balanced regarding the number of participants. The extracted subgroups were then compared in terms of ADNM, GAD and PHQ with the use of one-way ANOVA followed by Games-Howell post-hoc test.