Abstract
Introduction
The outbreak of COVID-19 has heightened mental health challenges among adolescents worldwide, while also hindering the development of effective emotion regulation strategies. Nevertheless, there exists a pertinent requirement to further investigate the psychological ramifications of COVID-19 on adolescents, as well as to discern disparities in these impacts across various nations. In order to address this gap in research, this study compares the prevalence of NSSI (characterized by purposefully injuring one's own body tissues without suicidal intention or engaging in behaviors that are not socially approved or sanctioned.), emotion regulation strategies, and psychopathology (internalization and externalization) among Belgian and Iranian adolescents pre- and during the COVID-19 pandemic.
Method
Adolescents between the ages of 13-21 from Iran and Belgium participated in a two-wave longitudinal design. In Iran, the sample consisted of 117 adolescents in wave 1 (April 2019) and 142 adolescents in wave 2 (Nov 2020). The Belgian sample consisted of 376 participants in wave 1 and 356 adolescents in wave 2. NSSI was assessed using the Self-Harm Inventory (SHI). Emotion Regulation was examined via the Emotion Regulation Inventory (ERI). The Strengths and Difficulties Questionnaire for Youth (SDQ) was used to assess psychological symptoms.
Result
Across the whole sample, lifetime NSSI prevalence was 35% in the first wave and 43.8% in the second wave. Longitudinal analysis across waves and counties did not show an increase in the prevalence of NSSI but did find that emotion dysregulation and internalizing symptoms are important risk factors for NSSI across countries. Emotional suppression and emotional dysregulation were the most prevalent emotion regulation strategies in Iran, whereas emotional integration was most prevalent in the Belgian sample.
Conclusion
Our study highlights cultural differences on the impact of COVID-19 on adolescent mental health. But it also indicates the importance of certain universal risk factors, for example, emotional dysregulation. These findings can inform developers to tailor programs for (early) intervention culturally.
Introduction
International research has consistently shown the detrimental impact of the COVID-19 pandemic on mental health worldwide.1-6 The outbreak of the COVID-19 pandemic led to a significant increase in feelings of anxiety and depression, particularly among adolescents.7,8 These, in turn, have been linked to increases in concentration problems, disrupted sleep, and elevated worry about academic performance, loved ones, and the overall influence of COVID-19 on daily life.9-12 For example, a cross-sectional study in February 2021 with 3000 Austrian adolescents (aged 14-20) reported a significant decrease in mental wellbeing, especially among girls. In total, 55% of the adolescents reported clinically relevant levels of depressive symptoms, 47% reported anxiety symptoms, 23% insomnia, 64% eating disorder symptoms and 16% repeated suicidal ideation (9% almost every day). 13 Hamza et al 14 was among the first to perform a longitudinal study (T1 May 19 - T2 May 20) among college students (Mage = 18.52, SD = .73) and found that college students without pre-existing mental health concerns were at an elevated risk for experiencing mental health difficulties during COVID-19 times. Surprisingly, this was not the case for students who already had mental health concerns pre-Covid: No significant increases in mental health difficulties were reported in this group. Although the psychological impact of the COVID-19 pandemic is examined in several populations and countries (e.g., adults in Western Countries), there are only a paucity of studies on the psychological ill-/wellbeing of adolescents in countries such as Belgian and Iran.
One study 15 from the SIGMA cohort reported a decrease and stabilization in psychopathological symptoms, measured with the Brief Symptom Inventory, and compared incidences in 2018, 2019 and 2020. Another Belgian study 16 with 7000 university students reported similar trends: Based on structured interviews based on the Kessler Psychological Distress scale (K6), the results show no significant increase in psychopathological symptoms during COVID-19 pandemic. These Belgian findings are not in line with other international studies and meta-analyses, where an increase in psychopathological symptoms during the COVID-19 pandemic is seen in adolescent samples. 17 Future research needs to determine the potential cross-cultural differences in psychological impact of COVID-19 on adolescents, especially in understudied countries, such as Belgium and Iran.
Researchers worldwide have issued warnings regarding the potential repercussions of the COVID-19 pandemic, emphasizing not only the escalation of psychopathological symptoms but also proposing that heightened distress during this period may impede individuals' coping mechanisms and hinder the development of effective emotion regulation strategies,18,19 especially during adolescence, as this is a period where emotion regulation strategies are still in development. 20 Emotion regulation can be defined as “the extrinsic and intrinsic processes responsible for monitoring, evaluating, and modifying emotional reactions, especially their intensive and temporal features, to accomplish one’s goals”21(27-28). Adolescents who have not yet acquired the necessary skills for effective emotion regulation often resort to utilizing maladaptive strategies to manage their emotions (i.e., emotion dysregulation 22 ) which in turn increases the risk of psychopathological symptoms. 23 Non-suicidal self-injury (NSSI), 24 the direct and deliberate damaging of body tissue without the primary intent to die and not socially sanctioned, is one maladaptive strategy that has been linked to emotion dysregulation 25 in early to mid-adolescence. 26 NSSI includes behaviors such as cutting, scratching, hitting and burning oneself. 27 Hasking et al 28 and Plener 19 have hypothesized an increase in non-suicidal self-injury (NSSI) during COVID-19 times due to the related increase of several NSSI risk factors such as loneliness, elevated levels of negative emotions, decreased social support, and increased academic stress. Although only paucity of empirical studies focused on the incidence of NSSI during the COVID-19 outbreak. For example, during the COVID-19 pandemic Tang et al 29 investigated prevalence rates of NSSI in a cross-sectional study among 1060 Taiwanese junior high school students and found an increased prevalence of NSSI of 40%. Similarly, Zetterqvist et al 30 compared prevalence rates of two studies pre-COVID and one study peri-COVID-19: Opposed to general prevalence rates of 17.2%–17.7% pre-COVID, 27.6% engaged in NSSI in comparable populations during COVID-19. In contrast to these findings, a longitudinal study in an adult sample examining NSSI prevalence rates during the COVID-19 pandemic (T1 May 20 – T2 August 20) among a nationally representative sample of 1032 Irish adults (aged 18-64 years), researchers found no significant changes in peri-COVID-times rates of NSSI. 31 Similarly, Steinhoff et al 32 conducted a longitudinal study of NSSI prevalence among 786 Swiss young adults, with one pre-COVID (i.e., 2018, aged 20 years) and four peri-COVID-19 assessments (one in April 2020 the fourth week following strict lockdown, one in early May 2020, third in late May 2020 and the last one in September 2020 after first restrict lockdown, aged 22 years) and found no significant increases in prevalence rates in this adult sample. As research to date shows some inconsistent findings, more longitudinal research is needed to examine the impact of COVID-19 on incidence and frequency of NSSI in adolescence. Furthermore, as Carosella et al 33 point out, the currently available research is mainly focused on western samples. Nonetheless, cross-sectional designs in non-western sample seem to indicate increases in NSSI incidence since COVID-19 (e.g., 29). No study to date has examined cross-cultural differences in NSSI incidence in a longitudinal design beyond western countries. Culture plays a crucial role in shaping people's emotions, thoughts, and behaviors. 34 Specifically, in collectivistic cultures, such as Japanese culture, close family relationships have a significant impact on the experience of specific emotions. This influence is known as co-regulation, where mothers play a crucial role in shaping their children's emotional experiences. In these cultures, mothers prioritize discussing and reflecting upon events that are considered important by others, rather than focusing on individual self-interest. As a result, they instill in their children an understanding of the interconnected nature of emotions.
On the other hand, individualistic cultures structure individuals' emotional lives based on personal choices in specific situations. These cultures create a social environment that encourages uniqueness and personal happiness through independent self-expression. This is in contrast to collectivistic cultures, like Japanese culture, where emotional expression is often suppressed in social settings to cultivate a more tranquil emotional state in individuals' lives. 35
One significant aspect influenced by culture is emotion regulation, including the motivation to regulate emotions and the use of various strategies for doing so. Extensive research has examined cultural differences in emotion regulation strategies and their adaptiveness,36-38 ultimately impacting psychological well-being. As culture promotes behaviors aligned with social values, engaging in behaviors consistent with these values fosters mental health. 39 Consequently, when emotion regulation aligns with cultural values, it tends to be adaptive, and vice versa. 40
To address this gap in research, the current study will examine potential cross-cultural differences in NSSI incidence in a longitudinal study examining a Western and non-Western population of adolescents. Pre-vs peri- COVID-19 times.
Based on Nock, 41 stressful life events hinder an individual’s ability to regulate emotions in an adaptive way which in turn increases the risk for NSSI which has been identified as a maladaptive way of emotion regulation. Research confirmed this model during the pandemic. According to Robillard et al 42 adolescents experiencing COVID-19-related stresses showed more engagement in NSSI during the pandemic and emotion regulation difficulties mediated this relationship. Moreover, the results of peri-pandemic studies 43 have shown an increase in internalizing symptoms in general and clinical populations of youth. Due to the evidence that emotion regulation difficulties increase the risk of internalizing symptoms such as anxiety and depression,44,45 it is important to investigate the role of emotion regulation difficulties and internalizing symptoms.
The current study addresses these gaps by longitudinally examining the incidence NSSI, in the first wave in April 2019 (pre-COVID19) and the second wave in November 2020 (peri-COVID19), both in a Belgian and Iranian sample of adolescents. Based on Carosella et al, 33 we hypothesize no increase in NSSI in our Western sample but do hypothesize (based on cross-sectional studies such as Tang et al 29 an increase in NSSI in our non-Western sample.
Based on previous studies (e.g., 13), we will examine gender, emotion regulation difficulties and internalizing symptoms as potential predictive risk factors for changes in NSSI. We expect females and youngsters with internalizing and emotion regulation problems to be more at risk for NSSI, in line with findings of Lutz et la 46 and Ara. 47
Methods
Participants
The initial sample for the first wave of data collection in Belgium (April 2019) comprised 376 participants from Belgium, with a mean age of 15.64 (SD = 1.42), and a female representation of 59.3%. Likewise, the Iranian sample in April 2019 consisted of 117 adolescents, with a mean age of 13.85 (SD = .62), and a female representation of 37.6%.
However, the onset of the COVID-19 pandemic disrupted our research timeline, leading to a deviation from the planned 12-month interval for data collection. Instead, in November 2020 after an 18-month hiatus we did data administration, during which a considerable number of participants dropped out (58.25% for Belgium and 62.39% for Iran).
To address this issue, we conducted recruitment for new cohorts with first wave in November 2020 and the second wave in November 2021. Subsequently, for this study a total of 156 students from Belgium and 44 students from Iran participated in both waves of data collection.
The table below denote the subset of participants who provided data for both wave 1 and wave 2, enabling the analysis of longitudinal changes within individual subjects.
Measures
Demographic questionnaire
The researcher created measures of demographic items, including age, sex, country of origin, parents’ birthplace, and religious orientation, which were collected via 9 sociodemographic questions.
Emotion Regulation was examined via the Emotion Regulation Inventory (ERI). 48 The ERI measures 3 emotion regulation sub-scales: integration, suppression, and dysregulation. Each subscale contains six questions for a total of 18 items with Likert-type response scales ranging from “not at all” (1) to “completely” (5). The integration subscale includes items such as: “When I experience negative emotions, I usually try to understand why that is the case” and “When I experience negative emotions, I think it is important to identify the cause as best as possible. to understand”. Suppression is measured with items such as: “When I experience negative emotions, I almost always try to hide them from others” and “I try to pay as little attention as possible to my negative emotions”. Examples of items that attempt to measure dysregulation are: “My negative feelings sometimes make me do things I don't really want to do” and “When I experience negative emotions, I lose control of my behavior”. The ERI has been demonstrated to be a reliable and valid tool for assessing the use of different strategies for regulating negative and positive emotions. In this study, psychometric reliability was demonstrated for all sub-scales at each time point in both samples and ranged from .58 to .75.
NSSI was examined using the Self-Harm Inventory (SHI), 49 a self-report questionnaire examining direct and indirect forms of self-harming behaviors. For this study, the 5 NSSI methods (cutting, burning, hitting, scratching, head banging, e.g. “Have you ever intentionally or purposely burned yourself”) and one item examined suicide attempts (“Have you ever attempted suicide?”) were administered. Respondents were considered to be engaging in NSSI if they mention using at least one method of NSSI in the SHI (Have you ever intentionally, or on purpose: cut yourself (item 2), burned yourself (item 3), hit yourself (item 4), banged your head (item 5), and/or scratched yourself (item 8) without suicidal intent).
The Strengths and Difficulties: Questionnaire for Youth (SDQ; 50 Dutch version 51 ) measures the presence of psychosocial problems, strengths and the influence of psychosocial problems on daily functioning. The questionnaire contains a total of 25 items related to the following five subscales: ‘hyperactivity/attention deficit’, ‘emotional problems', ‘peer problems', ‘conduct problems' and ‘pro-social behaviour’. The internalizing score is the sum of the ‘emotional problems' and ‘peer problems' subscales and externalizing score is the sum of ‘hyperactivity/attention deficit’ and the ‘conduct problems' subscales. The 25 items are formulated on the basis of statements and relate to the past six months. The answer options are: ‘not true’, ‘somewhat true’ and ‘certainly true’. For example, to measure the hyperactivity/attention deficit subscale, the following question is asked: “I am restless, I cannot sit still for long”. The emotional problems subscale is measured by, for example, the following item: “I get very angry and often have a temper.” Proble.ms with peers are mapped by, for example, “I get along better with adults than with young people my age”. “I often fight. “I manage to get other people to do what I want” is a sample item to map the conduct problems subscale, for the pro-social behavior subscale it is, for example, “I help someone who is hurt, upset or feeling sick”. Research shows that the SDQ is a reliable and valid instrument for identifying psychosocial problems in adolescents. 52 In this study, only the total score on psychological difficulties, internalizing and externalizing problems were used due to good internal consistency. Previous studies report good internal consistency for the total difficulties scale ranges and good construct validity (r = .70 for the total score) has been reported. In the current study, Cronbach’s alpha for the total difficulties scale is .75 for the Iranian sample and .75 (Belgium first wave; .72 Belgium second wave). Cronbach alpha for internalizing symptoms is .50 in the Iranian sample (Belgium first wave; .65 Belgium second wave) and for externalizing symptoms is .77 in the Iranian sample and .71 (Belgium first wave; .68 Belgium second wave).
Procedure
This study is part of a cross-cultural longitudinal study focusing on cultural differences in emotion (dys) regulation (Vatandoost, Baetens, Erjaee & Van Heel, in press). This unique longitudinal and cross-cultural dataset gives insights into pre-, and peri-COVID-19 impacts on mental health difficulties and maladaptive symptoms in a Belgian and Iranian sample of adolescents. Originally this study was conceptualized as a two-time point longitudinal study, with a first-time point in April 2019 (T1_cohort1) and a second time point in April 2020, but due to the closure of the schools in the spring of 2020, we administered T2_cohort1 in November 2020 (18 months post T1). As many students already finished their secondary school at T2, we administered data from a new cohort of students in November 2020 (T1_cohort2) to ensure enough power.
To ascertain the absence of any notable disparities in sociodemographic characteristics between participants who discontinued their involvement and those who remained in the study, we conducted a comprehensive analysis of the variations, as presented in the tabulated data below.
Data was administered in Belgium and Irian simultaneously. For the Iranian sample, 6 schools were contacted by email or telephone with the request to participate in this study. These schools were contacted because they reflect the metropolitan context of a large city in Iran. Two schools agreed to participate in this study, with students from the second and third grades in high school both from general and vocational tracks. In total, 68.80% of the 234 addressed Iranian students participated in this study. For the Belgian sample, 8 schools were contacted due to their geographic location in and nearby a metropolitan city in Belgium. The final sample consists of three secondary schools with students from the second and third grades, general education (ASO), technical education (TSO) and vocational education (BSO).
The research was evaluated and approved by the Medical Ethics Committee of UZ Brussel (2019/073) and Yazd Shahid Sadoughi University of Medical Sciences in Iran (IR.SSU.REC.1400.159). Prior to the study, the parents were informed about the study by means of passive informed consent. The students signed an active informed consent before filling in the questionnaire before each administration. For the second wave of the study, the students had to complete the questionnaires in the classroom while respecting the COVID-19 precautions (e.g., mouth mask, disinfecting, distance, etc.). This procedure was therefore different from the first measurement without COVID (mostly administered in large classrooms of students). The anonymity of the survey was guaranteed by assigning a unique code to each participant; this was used to link all longitudinal data at each time point. Student participants were compensated with a small gift (a flowerpot and seeds), as well as a resource list of available local mental health services.
Data analysis
Before conducting the statistical analysis, rigorous data cleaning procedures were undertaken. Outliers were identified through the utilization of boxplots. A total of 17 participants' data were deemed as outliers and subsequently excluded from the analysis.
The statistical analyses were performed on two independent datasets, based on the origin of the country where the data is collected (Belgium or Iran). Firstly, Chi-square and Mann-Whitney U test performed to ensure there is no significant difference between the sample of participants who dropped out and those who remained in study in terms of sociodemographic characteristics.
Then, NSSI and SSI prevalence were computed per wave and per country. Furthermore, differences between waves per country were assessed by the paired samples t-tests and follow-up paired t-test were conducted for NSSI vs no-NSSI groups. In a second step paired sample t-tests were conducted to assess the differences between the three emotion regulation strategies as well as internalizing and externalizing problems across two waves and within the same subjects.
Additionally, multilevel logistic regressions were employed to assess the predictive value of emotion dysregulation, gender, and country on the occurrence of NSSI (i.e., showing NSSI, yes or no). Due to the sample size a Bayesian estimator with non-informative priors was used in the regression analyses. Cohen's ƒ2 Cohen’s d were used to measure the effect size. Statistical analyses for this study were performed with IBM SPSS statistics version 28 and MPlus Version 8. 53
Results
Prevalence of NSSI and SSI
Cohorts and drop out rates information per countries.
Analysing demographic disparities between study dropouts and Persisting participants.
Note: for Iran all participants in the study were born in Iran.
aAs the age range in Iran was not wide and as the age range was not normally distributed we used Mann-Whitney to test the mean age’s difference.
Absolute and relative frequencies of NSSI and suicide attempt for the Belgian and Iranian dataset.
Note. N and % represent the (absolute and relative) values of the Belgian and Iranian participants who self-harm and/or commit a suicide attempt.
NSSI prevalence rates total, per countries and per waves. Note: W1 & W2 stand for wave one and wave 2.
SSI prevalence rates total, per countries and per waves. Note: W1,W2, SSI stand for wave one, wave 2 and scale for suicide ideation.
The Belgian prevalence for (at least one) suicide attempt is 8% (N = 30/374) for the first wave and 4.3% (N = 57/356) for the second wave, with no significant differences in (committing at least one) suicide attempt between T1 and T2 (z = -.378, P = .7). For Iran, a prevalence rate for suicide attempts at T1 was 1% (N = 1/117) for the first wave and 6.8% (N = 3/44) for the second wave, indicating a significant increase from T1 to T2 (z = -2.17, P = .03).
Groups’ differences in emotion regulation and psychological symptoms
Regarding the T-tests, Table 3 and Figures 3 and 4 represents the results of the paired samples t-tests, including the Belgian and Iranian participants who participated in both waves. Belgian participants show significantly lower levels of internalizing symptoms over time (t (138) = -3.32, P < .001) and no change in time regarding externalizing symptoms (t (138) = 1.74, P = .84). The internalizing symptoms (t (43) = 07.98, P < .001) and externalizing symptoms (t (43) = -6.44), P < .001) represent a significant increase in the Iranian sample over both waves. In terms of emotion regulation modes cross waves and countries, the sole significant difference observed was a decrease in integrative emotion regulation during wave two in Belgium (t(138) = - 2.99), P < .05) Difference in 3 modes of emotion regulation per countries and per waves. Note: W1 & W2 stand for wave one and wave 2. Difference in internalizing & externalizing symptoms per countries and Per Waves. Note: W1 & W2 stand for wave one and wave 2.
Comparison of total mean differences in ERI & internalizing and externalizing symptoms between both waves for Belgium and Iran.
Independent samples T-tests for ERI between the NSSI and no NSSI group in both Belgian and Iranian Waves.
Note. All variables are normal distributed for the no NSSI group in the second Iranian wave (dysregulation K-S (12) = .21, p<= .14), suppression (K-S (12) = .16, P = .20) and integration symptoms (K-S (12) = .19, P = .20).
Differences between Iranian adolescents vs Belgian sample on modes of emotion regulation and internalization and externalization symptoms.
NSSI predictors
Results for the multilevel linear regression analysis with NSSI as outcome.
Note: Change stands for change in NSSI between T1 and T2 and INT stands for Internalizing Symptoms.
Results for the multilevel logistic regression analysis with prevalence of NSSI as outcome.
Note. NSSI_PREV = prevalence of NSSI.
Overall, there is less NSSI in Belgium than in Iran. The correlation between NSSI and change in NSSI is negative meaning that when NSSI levels are high, the change is small (no significant difference in NSSI scores pre and peri-Covid).
Next, a multilevel logistic regression has been carried out to examine the role of emotion dysregulation in predicting NSSI. Analysis revealed that Emotion dysregulation significantly predicted an increase in the prevalence of NSSI (OR = 1.132) across countries and genders (see Table 8).
Discussion
Our design offered a unique lens to compare adolescents from two countries with different cultural factors (Iran as a collectivist culture and Belgium as an individualistic culture) with regards to psychological symptoms and emotion regulation strategies pre and peri-COVID-19 time.
In line with our hypothesis, findings indicated no significant increase in the prevalence of NSSI in COVID-19 times in our Western sample, as well as for our overall sample. Incidence of NSSI remains high pre- and peri-COVID in our Western sample, whereas the incidence significantly increases pre vs peri COVID in our Iranian sample, where the original incidence of NSSI was very low (compared to prevalence rates in meta-analyses such as Swannell et al 54 ). Our study shows that both internalizing problems and emotion dysregulation are important risk factors for NSSI, across countries and genders. These results are in line with previous studies that highlighted the role of internalization in NSSI.55,56 More specifically, De Luca et al, 57 using path analysis reported that adolescents with a history of internalizing symptoms experience more COVID-19-related stress which increases the risk for NSSI engagement. Our finding that emotion dysregulation predicts NSSI prevalence across countries confirms the conceptualization of NSSI etiological models that describe NSSI as a maladaptive way of emotion regulation (see the review by Klonsky 58 ). A recent study by Sorgi et al 59 showed that difficulties with emotion regulation significantly predict NSSI history. Our results expand these findings to samples from different cultural backgrounds.
Within our sample from Belgium, we observed that integrative emotion regulation emerged as the predominant strategy employed for the regulation of emotions and it also increased. Notably, despite the COVID-19 pandemic, we did not detect an increase in the occurrence of non-suicidal self-injury (NSSI) within the Belgian sample. This finding is significant because integrative emotion regulation is recognized as an adaptive approach to emotion regulation. Moreover, it aligns with prior research emphasizing the advantages associated with the utilization of integrative emotion regulation strategies in promoting mental well-being.60-62
The higher score of integrative mode of emotion regulation among Belgian sample might have contributed to the absence of a significant increase in NSSI cases during the pandemic. This highlights the potential advantages of incorporating integrative emotion regulation into interventions and treatments targeting mental health.
Furthermore, we found increase of internalizing symptoms in our Belgian subsample during the COVID-19 pandemic which is in line with previous studies. 63
The Iranian results showed that the emotion regulation strategy most used by Iranian adolescents is suppression. Also, the levels of emotion dysregulation were high in the Iranian sample and there was a significant increase in internalizing symptoms. An increase in internalization problems among the Iranian sample can be explained by cultural differences; due to the lack of social interaction during the lock-down and the nature of collectivism, being away from social interactions might be more detrimental in collectivist countries than countries with more individualistic values. As emotion regulation difficulties are a risk factor for NSSI64,65 it could partially explain the high prevalence of NSSI in the Iranian sample.
As mentioned, another objective of this study was to examine the role of gender, emotional dysregulation and internalization as potential predictive risk factors for NSSI. The result showed that there is no significant relationship between gender and NSSI for both samples which is consistent with other studies.66-68
Based on our study in terms of psychological symptoms and emotion regulation strategies during the COVID-19 pandemic, several similarities and differences emerged.
With regard to similarities, we first conclude that both Iran and Belgium share common risk factors for NSSI among adolescents, including internalizing problems and emotion dysregulation, suggesting the universality of these risk factors across cultures. Secondly, both countries emphasize the importance of effective emotion regulation strategies in understanding psychological symptoms and NSSI prevalence, particularly during the COVID-19 pandemic, highlighting the significance of emotion regulation for mental well-being.
Turning to the differences between Iran and Belgium, we first observe that the incidence of NSSI was initially lower in Iran compared to Belgium, but during the pandemic, there was a significant increase in NSSI prevalence in Iran while no significant increase was observed in Belgium. This suggests that cultural factors and the impact of the pandemic may have different effects on NSSI between the two countries. Additionally, the predominant emotion regulation strategy differed, with integrative emotion regulation being more commonly used in Belgium, while suppression was the primary strategy employed by Iranian adolescents. This indicates cultural variations in preferred approaches to emotion regulation. Lastly, Iranian adolescents displayed higher levels of emotion dysregulation and experienced a significant increase in internalizing symptoms during the pandemic, which may be influenced by cultural factors and the impact of reduced social interactions in a collectivist culture like Iran.This study highlighted the importance of cross-cultural studies to understand the impact of the COVID-19 crisis on adolescent mental health and specifically NSSI. The findings of this study propose more insight into the impacts of the ongoing crisis on youths in different countries which can have implications for interventions and prevention efforts.
Limitations and implications for future studies
The present study highlighted the impacts of the ongoing crisis of COVID-19 on adolescents' mental health outcomes and also the importance of cultural and between-country differences in this regard. The data are unique, but they are derived from an ongoing project which was not originally designed specifically to answer questions about the pandemic. Future studies need to include pandemic-related factors such as how they experienced the pandemic, the role of social isolation loneliness, health anxiety, and Covid-related stresses in total. Including these factors will be helpful in enhancing our understanding of the effects of the pandemic on the mental health of adolescents and the interactions between factors, and as a result, the extracted results can be explained in a multilateral way. Also, the high rate of dropout mainly due to covid and the utilization of convenience sampling may limit the generalizability of the findings. Nevertheless, by including a common subset of participants, recruiting from the same region and schools, and utilizing multilevel regression analyses, we have taken rigorous steps to ensure that the differences observed between wave 1 and wave 2 are primarily attributable to the COVID-19 pandemic or effects of time, rather than sampling differences. However, it is important to acknowledge the potential for enhancing our understanding of the dynamics of the variables under investigation through further research that incorporates more frequent measurements. Such an approach would undoubtedly contribute to a more comprehensive and nuanced comprehension of the research topic.
Another limitations is the restricted measurement of variables on only two occasions. Ideally, a third measurement could be incorporated, preferably post-pandemic, to comprehensively investigate the presence of enduring trends or the enduring impact of the pandemic on the studied variables. This additional measurement would enhance the study's capacity to discern the temporal dynamics and provide a more comprehensive understanding of the observed effects.
Moreover, it is needed to include a follow-up to observe the longer-term effect of COVID-19, because there are still lots of changes going on.
The result of this study is limited to the Iranian and Belgian samples. More comparative research is recommended to increase the knowledge about the impact of cultural factors and policies taken by different countries.
Conclusion
Our study is the first to shed light on cross-cultural differences in COVID-19 impact on mental health in adolescence from a longitudinal design. It highlights cultural differences in for example emotion regulation strategies, where Iranian adolescents are more likely to report emotional suppression strategies, and Belgian adolescents report more integrative emotion regulation strategies. More importantly, this study also underscores risk factors, such as high levels of emotion dysregulation, or internalizing problems that increase the risk for NSSI, across countries, and across genders. These findings may inform developers to be culturally sensitive in the development of for example early intervention programs, but on the other hand also informs policy makers and developers on universal targets (e.g., emotional dysregulation) which are relevant across countries and genders.
Footnotes
Acknowledgements
We extend our heartfelt appreciation to all the participants who took part in this study.
Author Contributions
Conception: V.S., B.I.; Project administration: V.S.,; Design: B.I., V.S.; Analysis: V.H.M., V.D.M.J, I.B., V.S; Data collection: V.S. ; Literature review: V.S., V.H.L.; Writing (Original Draft): V.S., I.B., V.H.L., V.D.M.J.; Writing (Review & Editing): V.S., I.B.; Supervision: I.B.; V.H.M.; All author read and approved final version.
Declaration of conflicting interests:
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding:
The authors declare no financial support for the research, authorship, and/or publication of this article.