Development and Validation of an Online Intervention Program to Augment Psychological Resilience: A Randomized Waitlist-Controlled Trial
Article information
Abstract
Objective
Resilience is the ability to bounce back from stressful events using individual and environmental resources. The online intervention program, MoVemEnT, which includes mindfulness training and emotion regulation, was developed to enhance resilience among young adults.
Methods
Persons without current medical or psychiatric diseases, aged 20–40 years, were recruited. Ninety-six participants were randomly assigned to either the intervention group or a waiting list control (WLC) group, with 45 participants per group completing the program.
Results
Results demonstrated significant improvements in the intervention group in measures of emotion regulation (Difficulties in Emotion Regulation Scale, DERS) and clinical symptoms of depression (Center for Epidemiological Studies-Depression Scale, CES-D) and anxiety (Generalized Anxiety Disorder-7) compared to the WLC group (all p<0.05). Correlation analyses indicated that improvements in DERS scores were associated with enhanced resilience, as measured by the Korean Resilience Quotient-53 (KRQ-53) and reduced CES-D scores. Reduction in DERS scores can reduce depressive symptoms (CES-D) through direct and indirect pathways via an increase in KRQ-53, according to mediation analyses.
Conclusion
Our findings reported positive effects of the MoVemEnT program, which includes short videos and brief homework assignments, on resilience in young adults. This study suggests that online intervention programs could offer an opportunity to simultaneously improve general mental health among a large population and may help reduce the burden of face-to-face psychological interventions.
INTRODUCTION
Resilience, the ability to bounce back from adversity and adapt to challenges, is essential for mental health. This ability is not merely a personal trait; rather, it is a skill that can be nurtured and developed over time [1]. Resilience enables individuals to manage tumultuous events by harnessing individual and environmental resources [2]. It empowers individuals to maintain psychological health, even in the face of significant stressors. For example, resilience could be a protective factor against depressive symptoms related to adverse experiences, such as maltreatment [3,4]. Therefore, increasing resilience capacity is crucial for recovering from and reducing negative consequences after adverse events.
Negative developmental experiences, such as maltreatment, can impede the development of resilience. Childhood maltreatment, including emotional, physical, and sexual abuse, as well as neglect, is highly prevalent across countries [5,6]. These experiences are closely associated with physical health problems7 and mental illnesses [8,9]. Maltreatment experiences can be related to resilience. Overall, lower exposure to violence is related to higher resilience in children [10]. Among various types of maltreatment, neglect, physical, psychological, and internet sexual abuse were significantly associated with lower resilience in a recent longitudinal study [11]. In another study, although all maltreatment types were negatively correlated with resilience capacity, only emotional abuse showed a significant relationship with resilience after adjusting for current psychological distress [12].
Additionally, several pieces of evidence explain the relationship between maltreatment experiences and resilience. Resilience comprises components of emotion, cognition, and behavior, all of which are affected by traumatic experiences. Maltreatment experiences have been shown to disrupt the recognition of emotional stimuli [13,14] and impede the development of appropriate emotional regulation skills [15]. Emotion regulation ability is significantly associated with resilience [16,17]. A lack of emotional awareness can moderate the relationship between maltreatment and resilience [18]. Additionally, maltreatment is linked to a lower use of cognitive reappraisal [19], which in turn moderates the relationship between maltreatment and depressive symptoms [20]. In several previous studies, cognitive reappraisal has shown a positive relationship with resilience [16,17]. Moreover, maltreatment experiences exhibited an adverse association with daily physical activity [21,22], potentially correlating with quality of life22 and resilience [23].
In line with previous research, several neuroimaging studies have supported the biological mechanisms underlying the relationship between maltreatment and resilience. Hyperactivity in the amygdala, an important area for recognizing negative emotional stimuli, has been linked to experiences of maltreatment [24,25]. Increased amygdala activity could indicate stress vulnerability [26]. Additionally, maltreatment experiences have led to a reduction in amygdala inhibition by the medial prefrontal cortex (mPFC) [27] or alterations in the development of mPFCamygdala connectivity [28], which is an important circuit for emotion regulation. Previous studies have reported that increased inhibition of the amygdala by the mPFC could be a protective factor after traumatic experiences [28]. Furthermore, activities in semantic networks, potentially related to cognitive reappraisal, showed a negative relationship with emotional maltreatment experiences [29]. Activities in salience networks, involved in measuring the salience of stimuli, were positively related to maltreatment experiences during thought-action fusion, which is one of the maladaptive metacognitive beliefs [30].
Based on previous literature, maltreatment experiences can weaken resilience through emotional, cognitive, and behavioral contexts. Firstly, maltreatment experiences make it difficult to recognize or regulate emotions appropriately. Secondly, these experiences disturb the ability to observe or reappraise thoughts accurately. Thirdly, physical activities may be diminished in individuals with maltreatment experiences. While various individual factors, such as cognitive flexibility [31], optimism [32], self-regulation [33], and problem-solving skills [34] can enhance resilience, the authors hypothesized that an intervention program targeting the aforementioned factors, particularly in individuals with maltreatment experiences, could effectively improve resilience. This approach might also prevent the development of psychiatric illnesses linked to resilience in those with maltreatment experiences [3]. Recent reviews suggest that intervention programs with flexible schedules [35] or combining mindfulness and cognitive behavioral therapy [36] were more effective. Considering these aspects comprehensively, this study aimed to develop and evaluate an online intervention program to enhance psychological resilience, grounded in scientific research.
METHODS
Study design
This study was a single-blinded, randomized, controlled trial conducted at a single center. Participants were allocated to either the online intervention or waiting list control (WLC) groups in a 1:1 ratio with gender matching. Following allocation, the intervention group was promptly scheduled to receive the online intervention developed by the authors. The WLC group, on the other hand, received the online intervention after a 4-week waiting period. Participants who completed the experiment received 100,000 Korean won (approximately USD 80). Approval for this study was obtained from the institutional review board of Kyungpook National University Chilgok Hospital under reference number 2023-02-013.
Participants
Persons without current medical or psychiatric diseases, aged 20–40 years, were recruited through online advertisements on several university boards located in Daegu, Korea. Patients with acute medical disorders that could affect daily function or other major psychiatric illnesses, including major depressive disorders, were excluded. A total of 107 people participated in the assessment of eligibility, of which 11 were excluded based on the study’s exclusion criteria. Four had current medical or psychiatric illnesses. Seven individuals strongly indicated major depressive disorder (n=6) or generalized anxiety disorder (n=1), reporting both positive screening results on the MINI patient health survey [37] and severe symptom measurements (Center for Epidemiological Studies-Depression Scale [CES-D] score of 25 or higher for depression and Generalized Anxiety Disorder-7 [GAD-7] score of 15 or higher for generalized anxiety disorder). After obtaining written informed consent, 96 participants were randomly allocated to either the intervention or WLC groups. Three participants from the WLC group withdrew during the waiting period, and three from the intervention group did not complete the video sessions or homework due to personal schedules, leading to their dropout. Finally, 90 individuals (45 in the intervention group and 45 in the WLC group) successfully completed our research protocol, including assessments conducted during pre-treatment and post-treatment periods, and were therefore included in the analysis.
Development of an online intervention program
A 4-week online intervention program named the MoVemEnT program was developed, consisting of two videos (about 15 minutes) per week along with brief homework assignments (Figure 1). This program was then administered. The content of week 1 explains how negative previous experiences or memories affect patterned suffering in daily life. Additionally, the importance of mindfulness is explained, with training in mindfulness breathing and eating. The content of week 2 focuses on emotional processing. One video addresses recognizing and appropriately labeling one’s emotions, while the other explains how to accept one’s emotions through self-compassion. The central topic of week 3 is thoughts. One video discusses how to balance one’s thoughts with cognitive reappraisal, while the other explains concepts of cognitive defusion, which can help in gaining some distance from one’s thoughts. During week 4, two videos demonstrate the importance of personal values, suggest identifying one’s own values, and guide participants in making small activity plans based on those values. Participants were asked to complete brief homework assignments for each video and submit them to the author. Mindfulness breathing was consistently suggested throughout the four weeks of the MoVemEnT program. The MoVemEnT program was entirely developed and provided in Korean, with video content totaling approximately 2 hours.
Structure of the MoVemEnt program.
Before starting the program, participants attended an initial face-to-face orientation, which included psychological assessments designed using SurveyMonkey (www.surveymonkey. com). Weekly announcements were sent via an online messenger to promote engagement. A clinical psychologist reviewed and encouraged homework completion. Participants were required to complete post-experimental assessments within one week after the four-week program.
Measures
Resilience-related measures
The level of difficulty in emotional regulation was evaluated using the Difficulties in Emotion Regulation Scale (DERS) [38]. The DERS assesses six domains of emotional difficulty in adults: nonacceptance of emotional responses, lack of emotional awareness, impulse control difficulties, difficulty engaging in goaldirected behavior, lack of emotional clarity, and limited access to emotion regulation strategies. High scores on the DERS indicate a greater degree of emotional difficulty. The Korean version of the scale was standardized with high internal consistency (Cronbach’s alpha=0.92) [39]. We also used the Cognitive Emotion Regulation Questionnaire (CERQ), a self-report questionnaire, to evaluate nine cognitive emotion regulation strategies. These strategies are categorized into five adaptive strategies (acceptance, refocus on planning, putting into perspective, positive refocusing, and positive reappraisal) and four less adaptive strategies (self-blame, other-blame, rumination, and catastrophizing) [40]. Higher scores indicate a greater use of each regulation strategy. The Korean version of the CERQ was validated with good internal consistency (Cronbach’s alpha=0.92) [41]. Based on the contents of the MoVemEnT program, the scores of DERS and CERQ were selected as primary outcome measures. Additionally, overall resilience was assessed using the Korean Resilience Quotient-53 (KRQ-53) [42]. Comprising 53 questions on a 5-point Likert scale, the KRQ-53 evaluates diverse aspects of resilience, such as controllability, positivity, and sociability.
Clinical symptoms measures
To measure depressive and anxiety symptoms, CES-D and GAD-7 were used, respectively. The 20-item CES-D was designed to assess depression symptoms in the general population [43]. The Korean version of the CES-D demonstrated strong internal consistency in the normal population (Cronbach’s alpha=0.91) [44]. The GAD-7 scale was developed to briefly measure general symptoms of anxiety with seven questions on a 4-point scale [45], and the scale was validated in the Korean language with good internal consistency (Cronbach’s alpha=0.92) [46].
Statistical analyses
Differences between groups in demographic characteristics, such as age, sex, education level, and symptoms of anxiety and depression, were assessed using t-tests for continuous variables and chi-squared tests for categorical variables. The effects of the group on changes in resilience-related and clinical symptom measures over time (pre-treatment and post-treatment) were examined using linear mixed-effect models. This analysis included group, time, and their interactions as fixed effects, with participants included as random intercepts to account for repeated measures. Post-hoc pairwise comparisons were conducted to calculate between- and within-group effect sizes (Cohen’s d), using independent t-tests for between-group differences and paired t-tests for within-group differences. Analysis of covariance was additionally employed to evaluate group differences in the primary outcomes over time while controlling for baseline scores. Furthermore, Pearson correlation analysis and mediation analyses were conducted to investigate factors related to changes in resilience.
Mediation analysis using structural equation modeling (SEM) was performed to determine the associations between difficulties in emotion regulation, resilience, and depressive symptoms. Based on the results of correlation analysis with multiple comparisons, we selected the independent variable (emotion dysregulation) as a significant predictor of the dependent variable (depressive symptoms) and the mediator (resilience). The significance of the SEM analysis was supported by 1,000 times bootstrapping iterations. The statistical significance level used in this study was p<0.05. All data analyses were performed using JASP (https://jasp-stats.org/).
RESULTS
Sample characteristics
Table 1 presents the pre-treatment demographic, psychological, and clinical characteristics of the participants selected for the study. The mean ages of the online intervention and WLC groups were 26.6±5.5 and 25.3±4.5 years, respectively. There were no significant differences between the groups in terms of age, sex, or education level. Additionally, no differences were observed in psychological measures, such as resilience, or clinical measures, including depression and anxiety, prior to treatment initiation.
Differences in baseline demographic, psychological, and clinical characteristics between intervention and control groups
Resilience-related outcomes
Linear mixed-effects model analyses revealed significant differences between the online intervention and WLC groups in DERS, CERQ, and KRQ-53 scores from pre-treatment to post-treatment. Post-hoc pairwise comparisons indicated that, compared to the WLC group, the online intervention group demonstrated significant improvements in emotion regulation, as measured by the DERS, at the post-treatment point (d=-0.90). However, CERQ and KRQ-53 did not show group differences (Table 2).
Means, SDs, and ESs across study time points and group-by-time interactions (resilience-related outcomes)
Additionally, within the online intervention group, overall scores on the DERS (d=-0.87) and CERQ (d=0.44 for adaptive and d=-0.59 for non-adaptive) demonstrated improvement between the pre- and post-treatment points (Table 2). Among the subscales of the primary outcomes, most of the DERS subscales (excluding impulse control difficulties and lack of emotional awareness) and some of the CERQ subscales (positive refocusing and catastrophizing) displayed significant improvement, with effect sizes exceeding medium thresholds (d>0.5) (Supplementary Tables 1 and 2). Additionally, scores on the KRQ-53, which represents overall resilience, showed improvement between pre-treatment and post-treatment points (d=0.51). While some subscales of the KRQ-53, including emotion regulation, life satisfaction, and appreciation, showed significant improvement, the effect sizes were relatively small (d<0.5) (Supplementary Table 3).
Clinical outcomes
The results of the clinical measures before and after the treatment are presented in Table 3. In the linear mixed-effects model analyses, significant group-by-time interactions were observed in the CES-D and GAD-7 scores. Post-hoc pairwise comparisons showed that, compared to the WLC group, the online intervention group demonstrated significant improvements in depressive symptoms as measured by CES-D (d=-0.99) and anxiety symptoms as measured by GAD-7 (d=-0.93) at posttreatment. Within the online intervention group, all measurements related to symptoms of depression (d=-0.78) and anxiety (d=-0.68) were reduced at post-treatment with medium to large effect sizes.
Means, SDs, and ESs across study time points and group-by-time interactions (clinical outcomes)
Relationship between changes in resilience-related measures and clinical measures
Pearson correlation analyses revealed that changes in the DERS scores (post-treatment minus pre-treatment) were significantly correlated with reductions in other resilience-related measures, including CERQ (r=-0.36, p=0.015 for adaptive and r=0.35, p=0.019 for non-adaptive) and KRQ-53 (r=-0.49, p<0.001). Also, changes in the DERS scores were significantly related to clinical outcomes, including CES-D (r=0.46, p=0.001) and GAD-7 (r=0.42, p=0.004). All significant results were maintained after multiple comparison analyses using the Benjamini-Hochberg procedure. All correlation results are described in Table 4.
Results of Pearson correlation analyses among changes of resilience-related measures and clinical measures
Mediation analysis using SEM
Changes in DERS scores were selected as the independent variable in mediating regression analyses, as DERS scores had a significant relationship with KRQ-53 scores (mediator) and CES-D scores (dependent variable) in correlation analyses with multiple comparisons. In our model, both direct and indirect pathways were significant. The standardized direct effect was 0.18 (p=0.033), and the standardized indirect effect was 0.28 (p=0.048). These results indicate that resilience partially mediates the effects of emotion dysregulation on depressive symptoms. The model represented perfect fit indices since it was saturated [χ2(0)=0.00, GFI=1, NFI=1, RMSEA=0]. Detailed results of mediational analyses are reported in Table 5.
The detailed parameter estimates related to mediation analysis
DISCUSSION
This study demonstrated the efficacy of an online intervention program in improving resilience among young adults. First, the DERS, CERQ, and KRQ-53 scores, which are closely related to resilience, showed significant improvement over the treatment period, whereas the WLC did not report any significant changes. Second, significant improvements in clinical measures, such as depression (CES-D) and anxiety (GAD-7), were observed only in the online intervention group. Third, changes in DERS scores were significantly associated with changes in CES-D and KRQ-53 scores. Furthermore, mediation analyses indicated that the improvement of depressive symptoms was indirectly mediated by the enhancement of overall resilience, which was influenced by emotion regulation. Our results provide crucial findings that serve as a cornerstone for verifying the effectiveness of online interventions in in-creasing resilience.
In terms of the DERS score, the online intervention group showed a significant reduction from pre-treatment (82.7±22.5) to post-treatment (67.4±16.0) points, with a large effect size (d=-0.87). The results suggest that our program has a greater impact on emotion dysregulation than on cognitive changes. The major components of our program included being aware of, allowing, and staying with emotions in a self-compassionate way. Also, mindfulness training was introduced in the first week of the program, with practices encouraged throughout the four weeks. Several studies have reported the effects of online self-compassion training on emotion regulation ability. For example, a study with psychology trainees found that DERS mean scores decreased from 76.8 to 67.4 after the program [47]. Another study reported that emerging adults showed improved emotion regulation after a 6-week online mindful self-compassion program (16.8 to 14.3 in the short form of the DERS), with all subscales showing significant improvement [48]. However, young people with chronic physical conditions did not show significant changes in DERS after a 4-week webbased self-compassion training, although increased well-being and decreased distress were observed [49]. Therefore, our improvement in DERS and its subscales might be due to the selfcompassionate approach to emotion processing. The effects could vary with different participants and should be verified in clinical samples.
Although the CERQ scores did not show significant differences between groups post-treatment, there were significant differences between pre- and post-treatment scores. Compared to intensive face-to-face programs, our program resulted in more minor changes. A recent study using an 8-week program focused on enhancing emotion regulation in patients with depression showed marked improvements in CERQ scores [50]. Similarly, eight sessions of group interventions targeting emotion regulation strategies in breast cancer patients had positive long-term outcomes on emotion regulation skills, lasting up to 2 years [51]. A recent review highlighted that structured interventions incorporating emotion regulation strategies, such as cognitive behavioral therapy, acceptance-commitment therapy, and dialectical behavioral therapy have positively affected emotion regulation skills [52]. While the changes in cognitive regulation skills through our program were relatively small, several factors should be considered. Cognitive reappraisal and cognitive defusion were major components of our program. Interestingly, larger effect sizes were observed in positive refocusing (d=0.78) compared to positive reappraisal (d=0.39). Cognitive defusion has roots in Beck’s cognitive distancing, and distancing is an early step in making cognitive changes, such as reappraisal and restructuring [53]. Positive refocusing, which involves diverting attention to unrelated positive events, is similar to distraction [54]. Therefore, our brief and simple online intervention might help participants create a small gap from negative thoughts rather than reconstruct their thoughts.
Furthermore, our program appears to have positive effects on general symptoms of depression and anxiety. Although our participants reported no previous or present clinical diagnosis of depression or anxiety, they showed reduced symptoms after the program. Various forms of online intervention have proven effective for depression and anxiety. For example, internet-based cognitive behavioral therapy (iCBT) has repeatedly shown positive effects on symptoms of depression [55]. In this meta-analysis, unguided iCBT was found to be similarly effective in individuals with symptoms of mild/subthreshold depression compared to guided iCBT [55]. Another meta-analysis showed that both tailored and transdiagnostic iCBT were effective interventions for anxiety and depressive disorders [56]. Recently, online interventions involving autonomous user engagement, referred to as internet-based self-help interventions, have been reported as effective for improving mental health among adolescents and college students [57]. The nature of the MoVemEnT program is similar to minimally guided internetbased self-help interventions with transdiagnostic content. Considering the nature of the program and our participants, the positive effects on symptoms of depression and anxiety seem consistent with those in previous studies.
In correlation analyses, changes in DERS scores showed a significant relationship with changes in both resilience-related and clinical measures. Previous studies have demonstrated the interrelationship among emotion regulation, resilience, and clinical symptoms, including depression. These studies found a positive relationship between emotion regulation capacity and resilience [18,58]. and a negative relationship between emotion regulation capacity and resilience with depression [3,59,60]. Also, difficulties in emotion regulation could mediate or moderate the relationship between significant stressful events, including childhood trauma, and resilience [18,58,60]. Overall, the results of our correlation and mediation analyses were consistent with previous research. Our findings indicated that improvements in emotion regulation can enhance overall resilience and reduce depressive symptoms through direct and indirect pathways. Therefore, a treatment program that includes skills for emotion regulation could be effective in enhancing resilience.
This study had some limitations. First, the sample size and the inclusion of only young and healthy adults limit the generalizability of the MoVemEnT program’s effects. We focused on young adults because the age of around 20 is the peak onset period for many psychiatric illnesses [61]. Enhancing resilience in this age group could help prevent such illnesses. Further studies with a larger sample size, including clinical populations, may be needed to extend our findings. Second, our study did not demonstrate the long-term effects of our program, as it did not include a follow-up period. Third, although the online intervention was later administered to the WLC group after a waiting period, the psychological effects of anticipating future treatment cannot be excluded. However, we believe these effects were minimal, as the WLC group reported no significant improvement in resilience-related and clinical measures during the waiting period. Fourth, all measures used in this study were self-reported questionnaires, so the results should be interpreted cautiously. For instance, scores on the DERS and its emotion awareness subscale increased significantly during the waiting period in the WLC group, while other measures showed no significant changes. Fifth, this study did not evaluate the degree of participants’ engagement. However, the authors reviewed weekly homework and communicated with participants through an online messenger to encourage sincere engagement in the program.
Despite the limitations of the study, our findings demonstrated the positive effects of the MoVemEnT program, which includes short videos and brief homework assignments with minimal guidance, on resilience in young adults. Our results suggest that online intervention programs could offer an opportunity to improve general mental health among a large population simultaneously and may help reduce the burden of face-to-face psychological interventions.
Supplementary Materials
The Supplement is available with this article at https://doi.org/10.30773/pi.2024.0319.
Means, SDs, and ESs across study time points and group by time interactions (subscales of DERS)
Means, SDs, and ESs across study time points and group by time interactions (subscales of CERQ)
Means, SDs, and ESs across study time points and group by time interactions (subscales of KRQ-53)
Notes
Availability of Data and Material
Data will be made available on an appropriate request.
Conflicts of Interest
The authors have no potential conflicts of interest to disclose.
Author Contributions
Conceptualization: Sang Won Lee. Data curation: Sang Won Lee, Jihyun Nam. Formal analysis: Sang Won Lee. Funding acquisition: Sang Won Lee. Investigation: Sang Won Lee, Jihyun Nam. Methodology: Sang Won Lee. Resources: Sang Won Lee. Supervision: Sang Won Lee. Validation: Sang Won Lee, Jihyun Nam. Writing—original draft: Sang Won Lee, Jihyun Nam. Writing—review & editing: Sang Won Lee.
Funding Statement
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (2021R1F1A1049460 and RS-2023-00241248).
Acknowledgments
Many parts of the MoVemEnT program, such as accepting one’s emotions, cognitive defusion, and identifying values, were based on acceptance-commitment therapy. The first author sincerely thanks Professor Seung Jae Lee for providing the opportunity to experience and learn acceptance- commitment therapy.