Background
The covid-19 pandemic has impacted the lives of people around the world and increased risk factors for mental health problems in an unprecedented way in modern times [
1]. Social interactions are pivotal for the well-being of humans and a lack of social interactions can have a profound effect on our lives. Although the full scope of the long-term psychosocial health effects of the pandemic has yet to be seen, research in the aftermath of covid-19 suggests that children and adolescents are particularly affected by the psychosocial consequences of covid-19 [
2‐
4]. Although less burdened by somatic complications caused by the virus, factors such as societal restrictions, closed schools and family distress have contributed to the lives of children and adolescents being severely affected by covid-19 [
5,
6]. In adolescents, elevated levels of depression, anxiety and stress, for example, have been found during the pandemic [
1,
4,
7], as well as increased risk behaviors [
8].
Different adverse psychosocial consequences of covid-19, such as social isolation and home schooling, are likely influenced by pre covid-19 status. In several mental health conditions there is a bidirectional relationship between vulnerability and stress, where existing symptoms of mental health problems influence how the stress situation is perceived and dealt with, and the stressful experience reciprocally increases symptoms and further risk for mental health problems [
9]. Recent research does in fact suggest that the psychosocial effects of covid-19 have been more stressful for already vulnerable groups with existing mental health problems [
10], which also seems to be the case for adolescents [
7]. Increased loneliness and depressive symptoms in the wake of covid-19 have, for example, shown to be especially evident in those experiencing pre-pandemic risk [
11]. In particular, social isolation, parental stress, low socioeconomic status, preexisting mental health conditions and disabilities contribute to increased impairment from covid-19 [
12]. There are, however, some inconsistencies in the literature so far. Contrary to expectation, Hamza and colleagues, for example, found that students without preexisting mental health problems experienced worse consequences of the pandemic [
13]. Still, the knowledge of how different vulnerable groups, such as adolescents with nonsuicidal self-injury (NSSI), experience the psychosocial impact of the covid-19 pandemic is scarce.
NSSI is often triggered by stress, such as social exclusion, and is conceptualized as a coping strategy [
14,
15]. In the light of this, concern has been raised as to how covid-19 has impacted NSSI [
5,
10,
16]. NSSI is a common behavior in adolescents [
17] and prevalence rates in community samples are reported to be around 20% [
18], while rates in clinical groups of adolescents are even higher, around 40–80% [
19]. Depression and anxiety are common forms of comorbidity [
20]. NSSI has its onset at age 12–14, peaks at 15–16 years of age and decreases in late adolescence [
17], which implies that adolescence is an especially important time period in which to study NSSI. NSSI is somewhat more common in girls, especially in clinical samples [
19]. The fact that NSSI is a risk factor for suicide and is associated with long-term risk [
21], makes it a highly relevant behavior to address.
NSSI is associated with emotion dysregulation [
22], and most often the behavior is engaged in to relieve negative emotions and thoughts [
23,
24]. In addition to regulating internal, emotional states, NSSI is also engaged in to influence the external, social environment [
25]. Some examples are: hoping to receive attention, help and understanding, solving an interpersonal conflict, increasing a sense of belonging and reducing social demands, typically when other forms of social communication have not been successful [
26]. Individuals with NSSI also perceive themselves as more interpersonally sensitive [
27], experience more loneliness [
28], have a negative social bias [
29] and poorer interpersonal communication and problem-solving skills compared to those without NSSI [
14]. According to the vulnerability-stress model [
30], vulnerabilities in adolescents with NSSI could potentially make them more susceptible to covid-19 related stress, which in turn negatively affects the symptomatology. It is therefore of interest to illustrate the impact of the psychosocial consequences associated with covid-19 on adolescents with NSSI.
There are some emerging empirical data indicating that adolescents with NSSI are in fact more burdened by the stress of covid-19. In a recent longitudinal study from Italy, adolescents [
N = 1061, mean age = 15.5 (SD = 0.8), 52.4% females] with earlier NSSI, internalizing symptoms and difficulties with emotion regulation (ER) before the outbreak of covid-19, experienced more covid-19 related stress, which increased the risk of further NSSI [
31]. Also, in another longitudinal study from the US, 91 young female adolescents ages 12–16 years were assessed pre and mid pandemic, and those who persisted with NSSI during covid-19 reported more loneliness and stress compared to those who had no experience of NSSI and those who had prior but not ongoing experience of NSSI during the pandemic [
32]. The role of ER in relationship to NSSI and covid-19 has also been examined in adolescents [
31,
33]. In a Canadian online survey of 809 adolescents (56% females), mean age 15.7 (
SD = 1.4), Robillard and colleagues, for example, found that difficulties with ER fully mediated the relationship between covid-19 stress and deliberate self-harm, irrespective of intent [
33]. From an international perspective, Sweden has had less lock-down and restrictions compared to many other countries, especially where children were concerned, who continued pre-school and school during the pandemic. Swedish high-school students, however, were twice affected by restrictions with closing of schools and distance education during 2020 and 2021 [
34].
NSSI often co-exists with other symptoms. In NSSI research it is generally a challenge to disentangle the unique effect of NSSI above and beyond the contribution of other psychosocial symptoms and diagnoses, which therefore also need to be taken into account. Furthermore, NSSI in adolescents is often hidden from others, and far from all cases come to the attention of health care [
35]. It is therefore necessary to investigate the effects of covid-19 in community samples.
In this cross-sectional study, we examine how a vulnerable group of adolescents with lifetime experience of NSSI, from a large community-based sample, perceive the psychosocial consequences of covid-19 compared to adolescents without NSSI, taking other symptoms of mental health problems into account, by examining the following research questions:
First, do adolescents with experience of lifetime NSSI differ in the way they perceive the psychosocial impact of the covid-19 pandemic compared to adolescents without NSSI, and second, is NSSI experience associated with perceived psychosocial covid-19 consequences, when demographic variables and symptoms of mental health problems (symptoms of depression and anxiety, and difficulties with emotion regulation) are controlled for?
We will further build on these results and go on to examine potential interaction effects between NSSI and the mental health problems that turn out to be significantly associated with covid-19 related stress in the analysis.
Based on a vulnerability and stress perspective and earlier research that suggest that the psychosocial effects of covid-19 are potentially more stressful for vulnerable groups, we hypothesized that those with NSSI would experience worse consequences of covid-19. In this cross-sectional study, we also hypothesized that NSSI experience, together with symptoms of depression and anxiety and difficulties with emotion regulation (ER), would be associated with perceived psychosocial impact of covid-19 in adolescents.
Discussion
In this cross-sectional study of a large community sample of older adolescents (N = 1602), negative psychosocial impact of the covid-19 pandemic was examined in individuals with and without NSSI. Adverse effects of covid-19 were more pronounced among adolescents with NSSI. Those with NSSI experienced more covid-19 related stress, such as loneliness, and less ability to reach out to friends and adults outside of the home if needed. Furthermore, more adolescents with NSSI had experienced an interruption in treatment and that home schooling had not worked well. Those with NSSI experience also reported higher levels of anxiety, depressive symptoms and ER difficulties compared to those without NSSI. When demographic variables and symptoms of depression, anxiety and emotion dysregulation were controlled for in a hierarchical regression, NSSI experience was not, however, independently associated with psychosocial covid-19 consequences. In the final model, experiencing the family’s financial situation as poor and neither good nor bad, studying a theoretical high school program, difficulties with ER and symptoms of depression were statistically significant. Furthermore, there was an interaction effect between NSSI and depressive symptoms, and a trend (p = 0.05) for an interaction effect between NSSI experience and ER difficulties on covid-19 consequences, where higher levels of depressive symptoms and ER difficulties weakened the effect of NSSI experience.
First, higher levels of covid-19 stress were reported by individuals with NSSI. That individuals with NSSI reported more adverse effects of covid-19 than their non-self-injuring peers confirms our hypothesis and also earlier studies that have shown that covid-19 stress has a more adverse effect on already vulnerable populations [
7,
11]. Increased loneliness, for example, in association with covid-19 has previously been found to be especially evident in those experiencing pre-pandemic risk [
11]. In the current study, more adolescents with NSSI also reported loneliness during the covid-19 pandemic. Furthermore, adolescents with NSSI reported that the pandemic had made it more difficult for them to connect with peers and adults outside the family if anything worried them. This is especially concerning since social support has shown to have a beneficial effect on NSSI, whereas perceived low social support, especially in the face of adversities, is associated with an increased risk for NSSI [
44]. That treatment had been interrupted is also disconcerting and potentially implies that transition to online resources, for instance, has not been unproblematic for this vulnerable group. The need for digital options for support and treatment of NSSI in the context of the pandemic has previously been highlighted [
45].
Adolescents with experience of NSSI thus perceived more covid-19 related stress compared to those without NSSI, as well as reporting higher levels of depressive and anxiety symptoms and difficulties with ER, which is consistent with previous research [
20,
22,
46]. Adolescents without vulnerabilities and mental health problems are potentially better equipped with alternative coping strategies, and thus less susceptible to the negative consequences of the stressor, according to the vulnerability-stress model.
Another potentially negative consequence for adolescents with NSSI related to covid-19 could be that prescribed restrictions resulted in limited access to necessary health-promoting activities. Different forms of behavioral activation, engaging in social activities, meeting friends and pursuing hobbies and interests, for example, are activities that can be used to help prevent an episode of NSSI, and also contribute to a positive identity and sense of empowerment [
47]. Limited opportunity to engage in these primarily social activities could potentially lead to a vicious cycle, with increased risk for further mental health problems in vulnerable groups.
Second, we examined if NSSI independently was associated with covid-19 related stress, even when other variables were controlled for. There is recent empirical support that adolescent girls might be more negatively impacted by the covid-19 pandemic, with a more risky life style [
8], more stress [
1] and increased presentation to health care for self-harm, irrespective of intent [
48], but our results did not support any sex differences in how covid-19 was perceived when symptoms of depression, anxiety and ER difficulties were taken into account.
The consequences of covid-19 are unfairly distributed in the population [
49], and in this study, perceiving the family financial situation as poor and neither good nor bad were associated with negative psychosocial consequences of the covid-19 pandemic. That the effects of the pandemic particularly affected socioeconomically burdened groups has been shown earlier in children [
50]. Studying a theoretical high school program was also associated with higher levels of covid-19 related stress. Few other studies have investigated this, but one tentative speculation could be that the pandemic caused more stress for adolescents studying theoretical programs, due to its potentially negative impact on their chance to applying to university.
When symptoms of depression, anxiety and difficulties with ER were controlled for, NSSI experience in itself, entered last in the model, was not an independent significant predictor in the total model on covid-19 consequences, and our hypothesis was thus not confirmed. Symptoms of depression and difficulties with ER explained more of the variance in covid-19 psychosocial stress than NSSI experience, and thus seem more important than NSSI itself when it comes to covid-19 consequences. Symptoms of depression and difficulties with ER were thus significant predictors in the multiple regression model, which was in line with our hypothesis. Symptoms of anxiety, however, were not associated with covid-19 stress, contrary to our hypothesis. It has previously been shown that adolescents with depressive symptoms [
7] have been more affected by the pandemic, which our results confirm. ER difficulties have also earlier been found to be important in relation to covid-19 stress [
31,
33]. The total explained variance in the model (23%) in the current study was low, indicating that other variables than those entered into the model also contribute to the understanding of the psychosocial impact of covid-19. Thus, how the covid-19 stressor is perceived does not seem to be limited to NSSI experience in itself, but rather to other symptoms, such as symptoms of depression and difficulties with ER. An earlier longitudinal study of Italian adolescents showed that earlier NSSI, together with internalizing symptoms and difficulties with ER, predicted more covid-19 related stress [
31].
Since NSSI in itself was not sufficient to predict perceived covid-19 related stress when other variables were controlled for in our model, and symptoms of depression and difficulties with ER were significant, we went on to further examine the relationship between NSSI, symptoms and covid-19 related stress by testing interaction effects. Results showed that depressive symptoms weakened the effect of NSSI experience. The impact of NSSI experience on covid-19 stress depended on whether levels of depressive symptoms were higher or lower. The effect of lifetime NSSI experience on covid-19 stress was higher when depressive symptoms were lower. When depressive symptoms were higher, on the other hand, the importance of NSSI experience decreased. NSSI experience thus interacted differently with covid-19 stress depending on the level of depressive symptoms. Higher symptoms of depression thus played a more important role in relationship to covid-19 related stress, which suggests that higher depressive symptoms, which are more pervasive in nature than lifetime prevalence of NSSI behavior, implies a greater burden, while the effect of NSSI was higher when depressive symptoms were lower. There was, however, no significant interaction effect between NSSI experience and ER difficulties. There was, however, a trend (p = 0.05), where ER difficulties weakened the effect of NSSI experience. When levels of ER difficulties were higher, the effect of NSSI was weakened. Since the interaction was only trend significant, caution has to be taken when interpreting these results. In conclusion, NSSI and depressive symptoms interacted to explain how covid-19 stress was perceived, i.e., with higher depressive symptoms, the effect of NSSI disappeared.
By international comparison, Sweden has had less lock-down and restrictions and there might be cultural differences in how much the pandemic affects adolescents [
51], which limits comparisons between countries. The lives of high-school students in Sweden have, however, been affected with distance education. Despite including a large sample of Swedish high school adolescents that enabled sub-group analysis, the study is not without limitations. The main limitation is the cross-sectional design, which does not allow for causal inferences. There is potentially a bidirectional relationship between vulnerability, symptoms and experience of covid-19 stress, but the design cannot cater for such complexity. The precise nature of these relationships is therefore unclear. Another limitation in the current study is related to the aspect of time. The data only allow for analysis of lifetime prevalence of NSSI, while ER difficulties and symptoms of depression and anxiety were rated currently, and this imbalance could potentially influence the results. The single item measurement of lifetime NSSI, which does not allow for analyses of NSSI frequency, severity or whether NSSI was ongoing or in the past, is another limitation. The fact that the study was conducted in the midst of the pandemic is a strength, contributing important information about this unique time period. However, at the same time it also means that measures of covid-19 impact had not yet been well validated. The measure of psychosocial consequences of covid-19 was developed for the purpose of the study, which calls for some caution. However, a factor analysis showed that all seven items loaded on one underlying factor. Furthermore, the response rate was low, which limits generalization to all high school students. Lack of time and covid-19 were most often given as reasons for schools not participating in the data collection. However, the data do not allow for further analysis of potential systematic bias in school drop-out. Similarly, data on the adolescents who did not choose to take part in the study are also lacking, although those reluctant to take part generally tend to be a more burdened group, which most likely would increase group differences had they been included. The results cannot be generalized beyond community samples of older adolescents. However, schools were selected based on information from the national school register and stratified to represent a normal population of third year Swedish high school students regarding school size and study program where the geographical distribution across Sweden was also considered.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.