2.2.5. Protective and Risk Factors: External Factors

Psychological support was assessed by the following question: "How much psychological support do you currently receive from your family, relatives, friends, and others around you?" Practical support was assessed by the following question: "How much practical support do you currently receive from your family, relatives, friends, and others around you?" The responses to both questions were assessed using a 10-point Likert scale (1 = not at all, 10 = receive enough support).

The family adaptability and cohesion evaluation scale III (FACES-III) was developed by Olson, Portner, and Lavee [42] to assess family function for family cohesion and adaptability. In this study, a measure validated in Korean by Lim, Lee, Oh, Kwak, Lee, and Yoon [43] was used. It consists of two sub-scale units of adaptability and cohesion. Family adaptability indicates the degree to which the family system can change in response to current and developmental stresses facing the family system, and family cohesion represents emotional bonds between family members. Each consists of 10 questions on a 5-point Likert scale. The higher the score, the higher the cohesion and adaptability of the family. The Cronbach's α of family adaptation was 0.91 and family cohesion was 0.9 in the present study.

### *2.3. Statistical Analysis*

Descriptive statistics were conducted to describe the sociodemographic characteristics of the data at the baseline time. Repeated measures analysis of variance (ANOVA) was conducted to reveal the annual differences of each variable, using SPSS 20.0 (IBM SPSS Statistics for Windows, version 20.0; IBM: Armonk, NY, USA, 2011). For panel regression analysis, the longitudinal data (wide type) collected over three years were merged first, and the data converted to long type data according to the ID and utilized for analysis. Additionally, a Hausman test confirmed the suitability of the fixed effect model and the random effect model. The Hausman test considers it more appropriate to apply a random effect model if the null hypothesis is rejected, rather than applying a fixed effect model [44]. The Hausman test and panel regression analysis were performed using STATA 14.0 (StataCorp LCC, version 14.0; Texas, USA, 2015.) with a statistical significance level of α = 0.05.

#### **3. Results**

Table 1 presents the characteristics of the 64 participants. The sample was composed of 40 females (62.5%) and 24 males (37.5%) aged 13 to 23 years (mean = 16.89, SD = 1.64 years). At the time of their first enrollment, the average number of years they had lived in South Korea was 3.06 years (SD = 2.54), from less than one year to 12 years. Meanwhile, the average childhood trauma experience was 1.63 (SD = 2.38), with 42.2% reporting no experience of trauma.


**Table 1.** Characteristics of participants at the baseline point.

Note SD: standard deviation accounted for.

Descriptive statistics and repeated measures ANOVA were performed to identify the annual characteristics of variables; results are presented in Table 2. The trend for depression symptoms increased year by year, and was also statistically significant (F = 3.09, *p* < 0.05). Specifically, the proportion of people who can be classified as having depression (cut-off score 21) increased steadily from 45.3% in the first year, to 53.1% in the second year, and 59.4% in the third year. Satisfaction with life was not statistically significant, but was observed to be decreasing, and practical support was also perceived to be decreasing by participants compared to the first year, indicating that the statistical differences are significant (F = 3.516, *p* < 0.05).

**Table 2.** Characteristics of variables, mean (SD).


\* *p* < 0.05.

A panel regression analysis was performed to determine how much change in the independent variables over time affects the degree of change in depression, and the results are presented in Table 3. Prior to the panel regression analysis, the Hausman test was performed. As a result, the random effect model was adopted in this study, indicating that the probability of significance was greater than 0.05 (χ <sup>2</sup> = 14.69, *p* = 0.1). The overall model's explanatory capacity was 37%, suggesting that the model's explanatory power was high (R<sup>2</sup> = 0.37, *p* < 0.01). First, it was shown that suppression of emotional expression affects depression as the period of living in South Korea increases. In other words, if emotional expression suppression increases by one unit over time, depression increases by 0.89 (B = 0.886, *p* < 0.05). On the other hand, resilience and present life satisfaction have been shown have an opposite effect over time, and with each unit increase in resilience and life satisfaction, depression decreased by 0.87 and 0.8 (B = 0.867, *p* < 0.001; B = 0.798, *p* < 0.05). The correlation between the independent variables and depression at each time point is presented in the Supplementary Table S1. Meanwhile, external factors (i.e., emotional and practical support, family adaptability, and cohesion) were not significantly related to changes over time in the depression of NKRYs.


**Table 3.** The result of random-effects GLS regression.

Note: Coef: coefficient, Std. Err: standard error, \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001.

#### **4. Discussion**

This study aimed to identify the factors affecting changes in depression over time by tracking NKRYs for three years. The findings and implications of this study are presented below.

First, the depressive symptoms of NKRYs increased significantly over the three years. Consistent with our results, a three-year follow-up study conducted by Cho et al. [27] with NKRs also showed that the level of depression increased significantly over three years after settling in South Korea. In addition, the number of participants who could be classified as depressed based on the cut-off score also increased every year, especially in the third year, with a high rate of about 60%. This is higher than that reported in previous cross-sectional studies with depression rates of approximately 30% to 48% [13,19,45], indicating that the psychological vulnerability of NKRYs can be considered high. Several studies have pointed to academic and socio-cultural differences as some of the reasons why NKRYs experience difficulties in adapting [46,47]. Despite the importance of academic achievement due to their developmental age, many NKRYs give up their regular academic courses because of academic maladjustment and the burden of academic achievement, which is likely to be high [46]. In addition, they are reported to have difficulty in forming peer relationships because of acculturation stresses based on different values and forms of expression of opinions, which may consequently lead to deterioration in mental health [48]. After arriving in South Korea, they receive short-term adaptation education and are immediately deployed to South Korean society, and their psychological stress seems to be increasing as they enter a boundlessly competitive system with South Korean teenagers without psychological stability and adaptation. [14]. In particular, given that the average residence period in South Korea is only 3.06 years, special attention is required for the mental health of NKRYs in the early stages of settlement.

Second, panel regression was used to determine factors affecting changes in depression over time, and it was shown that individual internal factors had a statistically significant effect on changes in depression. Specifically, expressive suppression of emotion was shown to increase depression over time. This result is in line with previous studies that mentioned expressive suppression as a risk factor for depression [16,21]. Aldao, Nolen-Hoeksema, and Schweizer [49] suggested that the presence of expressive suppression is likely to be more strongly associated with depression than the absence of cognitive reappraisal. In terms of interpersonal relationships, adolescents seem to choose emotional suppression rather than expression to avoid damaging relationships with peers who show their own symptoms of depression [50].

In addition, resilience was a significant preventative factor for changes in depression in this study. Previous studies on NKRYs focused on resilience as a preventative factor for mental health issues including depression [16,19,24,45,51]. According to McLaughlin, Doane, Costiuc, and Feeny [52], resilience consists of two aspects: heightened psychological vulnerability and adaptation to risk. Based on the above concepts, for NKRYs who are forcibly exposed to various stresses, resilience is likely to be an important internal factor that can increase adaptability by protecting them from the psychological stresses that may appear in the adaptation process.

Additionally, life satisfaction has been shown to have a negative impact on the increase in depression over time. This is consistent with a two-year follow-up study of 189 NKRs [53], which showed that participants' depression increased while their overall life satisfaction decreased. Another study of NKRYs also reported significantly lower life satisfaction in a group with depression than in a group without depression [19]. Several studies on refugees and immigrants mention post-immigration factors, such as experiencing discrimination, not having close friends, and acculturation stress, which affects life satisfaction or depression rather than pre-immigration stress [54–56]. Therefore, the results imply that depression prevention programs should consider ways to enhance resilience and life satisfaction and foster ego strength by recognizing emotions and promoting healthy emotional expression.

Meanwhile, external factors were not significant in the change in depression over three years in the present study. Specifically, psychological and practical support did not directly affect the change in depression. This is contrary to previous studies, which showed that the perception that support can be gained from relatively close family or peer relationships plays a positive psychological role in the cultural adaptation process [23,45,57]. However, according to a study by Jeong and Kang [8] on environmental protection factors, peer support alone does not have a significant impact on cultural adaptation stress, but can be indirectly influenced by personal internal protection factors such as internal locus of control, suggesting the indirect influence of external factors. Regarding family function, no statistically significant association with changes in depression over time was reported in this study. Meanwhile, Nam et al. [24] revealed that family cohesion was significantly associated with depression among NKRs, in contrast to our finding that family adaptability was not associated with depression. However, half (50%) of the participants were reported to be living with friends or in dormitories away from their families, and may therefore be significantly affected by peer relationships rather than family relationships [58].

There are some limitations to this study. First, the sample size was relatively small; therefore, care should be taken in interpreting and applying the results. Given the characteristics of school-based research, the dropout rate was very high because of changes in students enrolled in the school; participants who failed the study were no longer present at school due to reasons such as graduation or suspension of study, and were no longer able to follow up. This was beyond the control of the researchers; therefore, we propose in following studies the need to select a study subject by considering the dropout potential of participants in the research design phase. Second, some scales, such as psychological support and practical support, comprised a single question, which had limitations for obtaining substantial responses. Therefore, future studies will likely need to gather more detailed information through validated questionnaires. Also, the Cronbach's alpha for the measure of expressive suppression was relatively low. We believe this is due to the small total sample size, and one must be careful about understanding the result. Third, although this study explored variables that affect temporal changes in depression, there are limitations that have not been able to compare the relative influence of variables or present structural models. Therefore, for further studies it is recommended to use statistical methods to clarify the temporal causal relationship.

Nevertheless, to the best of our knowledge, there is only one longitudinal study of NKRYs to date [16], and it is therefore a significant achievement for this study to identify individual psychological changes over three years using follow-up observations. It is also

meaningful in that it provides information on the mental health of NKRYs in the initial settlement process, less than five years after settlement. In this study, we found a tendency for depression in NKRYs to increase year by year for three years. Emotional suppression has also been found to be a risk factor that can increase depression, and resilience and life satisfaction have been shown to be protective factors that decrease depression. These results suggest that the development of programs to encourage individual resilience and life satisfaction is required to promote mental health in NKRYs, and to seek ways to promote and safely address the emotions experienced. Further research is required to expand and justify the research results through a larger and more representative sample.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/1660-460 1/18/4/1696/s1, Table S1: Correlation analysis between depression and main variables.

**Author Contributions:** Conceptualization, S.P. and Y.S.; methodology, Y.S. and S.P.; formal analysis, Y.S. and S.P.; investigation, S.P.; writing—original draft preparation, Y.S.; writing—review and editing, Y.S. and S.P.; funding acquisition, S.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean Government [NRF-2016R1D1A1B03931297].

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board of the National Center for Mental Health (No. 116271-2017-11).

**Informed Consent Statement:** Informed consent was obtained from all participants involved in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.
