Identifying Subgroups of Suicidality Among Adolescents and Influencing Factors Using Latent Class Analysis
Article information
Abstract
Objective
We aimed to classify subgroups of suicidality among adolescents and identify the influencing factors of the classification of these latent classes.
Methods
Suicidal thought, plans, and attempts as well as the feelings of sadness/hopelessness and loneliness were utilized as indicators to derive the suicidality classes. Additionally, health behaviors, such as dietary habits, physical activity, experiences of violence victimization, sexual activity, and deviant behavior, along with demographic factors, such as sex, school year, grades, and household income, were considered as influencing factors. The analysis utilized data from the 18th Youth Health Behavior Survey (2022) conducted by the Korea Disease Control and Prevention Agency, involving 51,850 middle and high school students.
Results
The findings revealed three latent classes of suicidality among adolescents: “active suicidality,” “passive suicidality,” and “non-suicidality.” The influencing factor analysis indicated that all factors, with the exception of high-intensity physical activities, significantly influenced the classification of latent classes of suicidality. Notably, walking exercise and the frequency of exercise during physical education class were found to be factors that differentiated between active and passive suicidality within the suicidality classes.
Conclusion
This study employed nationwide data to identify the exhibited suicidality classes among adolescents and tested the influencing factors necessary for predicting such classes. The study’s findings offer valuable insights for policy development in suicide prevention and suggest the need for developing customized interventions tailored to each identified class.
INTRODUCTION
According to the suicide rate statistics of the Organization for Economic Co-operation and Development (OECD) for the years 2018–2020, South Korea faces a significant issue with the suicide rates, ranking first among the 42 OECD member nations [1]. While this ranking predominantly reflects the broader population and not specifically adolescents, this problem is particularly acute among younger age groups. This is highlighted by the Causes of Death Statistics in 2022 from the Statistics Korea, reporting suicide as the leading cause of death among Korean adolescents for 10 consecutive years [2]. Adolescence represents a period of rapid physical and psychological changes, and adolescents may experience emotional difficulties owing to such changes, which can sometimes lead to suicide attempts. Despite ongoing efforts in suicide prevention, the suicide rate among Korean adolescents continues to rise, exacerbated by the impact of the coronavirus disease-2019 (COVID-19) pandemic [3]. This sustained increase underscores the necessity for a multidimensional understanding of suicidality among adolescents.
Adolescent suicide often results from complex interactions among various factors. It encompasses intricate thoughts or behaviors rather than being solely the consequence of a single actions [4]. Therefore, behavioral aspects related to suicide, including suicidal thoughts, plans, and attempts, as well as various psycho-emotional characteristics of adolescents should be examined to identify whether they are associated with suicidality [5]. Many studies have focused on analyzing the reasons for attempting a suicide as well as psycho-emotional variables, such as depression, anxiety, sadness, and hopelessness, as indicators, in addition to suicidal thoughts, plans, and attempts, to differentiate suicidality classes among adolescents [5-7]. Although not included for differentiating suicidality classes in previous studies, the present study intended to focus on loneliness, which is considered another risk factor for suicide. Loneliness can be defined as a quantitative or qualitative lack of a social relationship network [8,9], which is reportedly associated with both suicidal thoughts and behavior [10,11].
Most previous studies on factors that influence the differentiation of suicidality classes have analyzed the influence of specific factors or focused on the psycho-social characteristics of adolescents. For example, some studies have examined the influence of the experience of child sexual violence on suicide behavior patterns [12] and psycho-social variables, such as impulsivity, self-esteem, self-harm, abuse experience, and deviant behavior, as predictors for differentiating suicide behavior patterns [6]. Another study examined sex, academic grades, sex identity, and self-perceived overweight as factors for predicting suicidality population [7]. The present study examined variables that showed a significant association with suicide behavior for comprehensive consideration of how health behavior factors, including activities of daily living (ADLs), of adolescents discussed in existing studies can influence the suicidality among adolescents. Studies have reported that dietary habits and physical activities have a significant influence on depression and suicidal thought [13-15]; school victimization increases the likelihood of suicidal thought and attempt [16,17], and sexual activity experience, smoking, and drinking are significant influencing factors that increase depression, suicidal thought, and suicidal attempts [18]. Based on the predictors identified in existing studies, dietary habits (breakfast status and frequency of eating fast foods), physical activities (high-intensity activities, moderate-intensity activities, muscle strengthening exercises, walking, and physical exercises [PEs]), violence victimization experience, sexual activity experience, deviant behavior (smoking, and drinking), and general characteristics (sex, year in school, grades, and household income) were included as influencing factors for differentiating suicidality classes.
In the present study, latent class analysis (LCA) was used to differentiate suicidality classes among adolescents and prepare interventions specialized for each class. LCA is based on a person-oriented approach, and its subgroups are derived by analyzing the response patterns of individuals. This approach offers the advantage of analyzing heterogeneity within individuals that remains undetected in variable-centered analysis.
Accordingly, this study aimed to explore adolescent suicidality by examining suicidal thoughts, plans, and attempts, and psycho-emotional factors such as sadness/hopelessness and loneliness, thus broadening the understanding of suicidality. It focused on health-behavior factors, including ADLs, to predict suicidality classes among adolescents and provide data for customized intervention methods in suicide prevention education. The research questions are: “How many latent classes of suicidality exist among adolescents?” and “What factors influence these classes?”
METHODS
Study population
The present study used data from the 18th Youth Health Behavior Survey (2022). The Youth Health Behavior Survey was conducted by the Korea Disease Control and Prevention Agency (KDCA) and the Ministry of Education for the purpose of identifying the health behaviors of youths and deriving health indicators for youths. The data from the 18th survey used in the present study consisted of those from 51,850 students from Korea (26,397 boys and 25,453 girls).
Measures
Suicidality
For differentiating the latent classes of suicidality among adolescents, suicidal thought, suicidal plan, suicidal attempt, sadness/hopelessness, and loneliness were used as indicators. To measure suicidal thought, suicidal plan, suicidal attempt, and sadness/hopelessness, Yes-or-No questions were used to ask regarding such experiences: “Have you seriously thought about suicide in the past 12 months?;” “Have you made any specific plans to commit suicide in the past 12 months?;” “Have you attempted suicide in the past 12 months?;” and “Have you felt sad or hopeless to the point of stop doing your ADLs for 2 weeks straight in the past 12 months?” In addition, loneliness was measured using a question that asked how often the respondents felt lonely in the past 12 months and the question was rated on a 5-point Likert scale (1: never; 2: almost never; 3: sometimes; 4: often; and 5: always). In the present study, adolescents with scores of 1 (never felt lonely) or 2 points (almost never felt lonely) were defined as the not lonely group, while those with scores of 3 (sometimes felt lonely) to 5 points (always felt lonely) were defined as the lonely group.
Predictors
For factors that influence the classification of latent classes of suicidality, adolescents’ health behaviors, such as dietary habits and physical activities, violence victimization experience, sexual activity experience, health risk behaviors, and general characteristics, were used.
For dietary habits, single-item questions were used to measure the frequency of having breakfast in the past 7 days, which is phrased as “In the past 7 days, on how many days did you have breakfast?”, (1: 0 day–8: 7 days) and eating fast food, which is “In the past 7 days, how often did you eat fast food?” (1: 0 times a week; 2: 1–2 times a week; 3: 3–4 times a week; 4: 5–6 times a week; 5: once every day; 6: twice every day; and 7: ≥3 times every day).
For physical activities, frequency of high-intensity physical activities, moderate-intensity activities, muscle strengthening exercise, walking, and direct exercise during PE class were used. High-intensity physical activities were measured based on the question, “How many days in the past 7 days did you perform physical activities that made you feel out of breath or caused you to sweat?” Moderate-intensity physical activities were measured based on the question, “How many days in the past 7 days did you perform physical activities that made you feel slightly out of breath than usual?” Muscle strengthening exercise was measured based on the question, “How many days in the past 7 days did you perform muscle strengthening exercises such as push-ups, sit-ups, lifting weights, dumbbells, pull-ups, and/or parallel bars?” Each question was scored on an 8-point scale (1: not at all; 8: 7 days/week). The frequency of direct exercise during PE class was measured based on the question, “How many days in the past 7 days did you directly exercise on the field or gymnasium during PE class?” This question was scored on a 4-point scale (1: none; 2: once a week; 3: twice a week; 4: ≥3 times a week).
Violence victimization experience was measured based on the question, “Have you received medical treatment at a hospital due to violence (such as physical assault, threats, bullying, etc.) from friends, seniors, or adults in the past 12 months?”. This question was scored on a 7-point scale (1: 0 times; 7: ≥6 times).
Sexual intercourse experience was measured based on the question, “Have you ever had sexual intercourse?”, to which the respondents answered “Yes” or “No.”
Health risk behaviors were investigated based on smoking and drinking. Smoking was measured using a single-item question, “Have you ever smoked a cigarette?”, to which the respondents answered “Yes” or “No.” Drinking was measured using a single-item question, “Have you ever consumed even one serving of alcohol?”, to which the respondents answered “Yes” or “No.”
For the general characteristics, academic grades and household economic status were investigated. Academic grades were measured using a question, “ How were your grades in the past 12 months.” This question was scored on a 5-point scale (1: upper to 5: lower), with higher scores indicating poorer academic grades. Household economic status was measured based on the question, “What is the economic status of your household?” This question was also scored on a 5-point scale (1: upper to 5: lower), with higher scores indicating lower household economic status.
Data analysis
The present study used LCA to identify the suicidality among adolescents and examine the predictors for each class. Conventional variable-oriented approach has the limitation of not being able to account for subgroups as it neglects heterogeneity within individuals [19]. In contrast, LCA, which is based on a person-oriented approach, is an analytical method that offers the advantage of deriving subgroups based on the response patterns of individuals to the given questions and identifying the influencing factors and outcome variables for each class [20].
To find the optimal LCA, statistical indicators and interpretability were considered. Moreover, statistical indicators, such as information ratio, model comparison verification, and quality of classification, were used to determine the number of latent classes. For information ratio, Bayesian information criterion (BIC) and sample size-adjusted BIC (SABIC) were used, where smaller values indicated better fitness of the model [21,22]. For model comparison verification, the Lo–Mendell–Rubin likelihood ratio test (LMR-LRT) and the bootstrap likelihood ratio test (BLRT) were used, in which significant p-values indicated an statistically ideal profile model [23,24]. For quality of classification, entropy was examined, where values >0.70 indicated better quality of classification [25]. To conclude concerning whether the models can be theoretically explained, we consider the final number of acceptable models.
To classify the latent classes and identify the influencing factors for each latent class, a three-step approach was applied [26]. This approach was proposed as a method to prevent changes in the classification of latent classes owing to influencing factors, which has the advantage of resolving the problem of classification instability that may appear in analysis with the conventional one-step approach [26,27]. In the current study, to examine predictors, the R3STEP approach was utilized, which involves multinomial logistic regression. These analyses were conducted using Mplus Version 8.10 (Muthén & Muthén, Los Angeles, CA, USA).
Ethics statement
Korea Youth Risk Behavior Survey (KYRBS) contained nationally approved statistical data and the KYRBS raw data used in this study were approved for research use through the KDCA website. The protocol of this secondary analysis study was approved by the Institutional Review Board of Korea University Medical Center, Ansan Hospital, Gyeonggi-do, Korea (No. 2023AS0277).
RESULTS
Descriptive statistics
Table 1 shows the descriptive statistics of the variables used in the present study. Among the participants, 50.9% were boys and 49.1% were girls. There were more middle-school students (54%) than high-school students (46%). Regarding feeling sadness/hopelessness, 71.2% of the students responded “No” and 28.8% responded “Yes.” Moreover, 54.4% and 45.6% responded that they did and did not feel loneliness, respectively. Meanwhile, 85.8% of the students indicated that they had thought about suicide before and 14.2% of the students indicated that they had not; 95.5% had never planned suicide and 4.5% had planned; and 97.3% had never attempted suicide and 2.7% had attempted.
Demographic and mental health descriptive statistics of participants (N=51,850)
Extraction of latent classes
To derive the latent classes of suicidality among adolescents, LCA was performed using five indicators (i.e., sadness/hopelessness, loneliness, suicidal thought, suicidal plan, and suicidal attempt). The analysis was performed with the number of latent classes increased in order from 1 to 4. Information ratio (BIC, SABIC), model comparison verification (LMR-LRT, BLRT), and quality of classification (entropy) were statistically examined. When the information ratio was checked, the results showed that the value decreased as the number of latent classes increased (Table 2). While lower information ratio value indicates that the model is fit, the information ratio value tended to decrease as the number latent classes increased. Therefore, the number of latent classes is determined by considering the point at which the margin of decrease flattens [20]. In the present study, the results showed that the margin of decrease was relatively larger when the number of latent classes increased from 2 to 3. A comparison of model validation results showed that they were all significant (p<0.05). Meanwhile, the quality of classification was all >0.7 when the number of models was 2–4. Therefore, the number of latent classes for the present study was determined as three by considering not just the statistical indicators but also the interpretability and parsimony of classes.
Latent class enumeration and model fit indices
Characteristics of each latent class
When the three latent classes were compared, Class 1 (n=2,486, 4.8%) showed high scores for sadness/hopelessness, loneliness, suicidal thought, suicidal plan, and suicidal attempt, and accordingly, this class was determined as a high-risk class and named “active suicidality class.” Class 2 (n=12,179, 23.5%) showed high scores for depression, loneliness, and moderately elevated levels of suicidal thought, but it did not actually show suicidal plan and suicidal attempt, and, thus, it was named “passive suicidality class.” Class 3 (n=37,185, 71.7%) showed low scores for all variables, and accordingly, it was named “non-suicidality class.” The characteristics of each latent class of suicidality among adolescents are presented in Table 3 and Figure 1.
Characteristics of each latent class of suicidality (response probability)
Line graph of conditional response probabilities comparing latent class profiles on sadness and hopeless, loneliness, suicidal thought, suicidal plan, and suicidal attempt (Probability of Response: 0=No, 1=Yes).
Predictors for the classification of latent classes
The effects of the predictors for classification of the final three latent class models were investigated. Table 4 and Figure 2 show the variables that had a statistically significant influence on the latent classes among adolescents’ dietary habits (frequency of having breakfast and frequency of eating fast foods), physical activities (number of days of high-intensity activities, number of days of moderate-intensity activities, number of days of muscle strengthen exercise, number of days of walking exercise, and frequency of direct exercise during PE class), violence victimization experience, sexual activity experience, health risk behavior (smoking and drinking), and general characteristics (academic grades, household economic status, sex, and year in school).
Three-step analysis results for antecedents (R3STEP)
Multinomial logistic regression models of ORs with 95% CIs for Class 1 vs. Class 3 (ref), Class 2 vs. Class 3 (ref), and Class 1 vs. Class 2 (ref), based on multinomial logistic regression. Each point represents the OR for predictors, with horizontal lines indicating the 95% CIs. Classes labeled with (Ref) serve as reference categories for comparisons. A vertical line at OR=1 denotes no effect. PE, physical exercises; OR, odds ratio; CI, confidence interval.
The analysis results showed that adolescents with the following characteristics were more likely to belong to the active rather than to the non-suicidal thought class: having a dietary habit of skipping breakfast or frequently eating fast foods; performing more moderate-intensity physical activities and muscle strengthening exercise and less exercise during PE class with respect to physical activities; having violence victimization and sexual activity experience; smoking and drinking; individuals of female sex; middle-school students; having lower academic grades; and having lower household economic status.
Moreover, adolescents skipping breakfast or frequently eating fast foods; performing moderate-intensity physical activities, muscle-strengthening exercise, and walking exercise for more days; having violence victimization and sexual activity experience; having more experience of smoking and drinking; having lower academic grades; having lower household economic status; individuals of female sex; and middle-school students were more likely to belong to the passive suicidal thought class rather than the non-suicidal ideation class.
Finally, adolescents performing less walking exercise and exercise during PE class; having violence victimization and sexual activity experience; smoking and drinking; individuals of female sex; and those having lower household economic status were more likely to belong to the active suicidal ideation class rather than the passive suicidal ideation class.
DISCUSSION
The present study derived the latent classes of suicidality among adolescents and examined the influence of adolescents’ health behaviors (dietary habit, physical activities, violence victimization experience, sexual activity experience, and deviant behavior), academic grades, household economic status, sex and year in school on the latent classes.
Suicidality among adolescents was classified into three latent classes: active suicidality (n=2,486; 4.8%), passive suicidality (n=12,179; 23.5%), and non-suicidality (n=37,185; 71.7%) classes. The active suicidality characteristically showed high scores for sadness/hopelessness, loneliness, suicidal thought, suicidal plan, and suicidal attempt. The passive suicidality showed similar levels of sadness/hopelessness and loneliness and relatively lower suicidal thought as compared to those of the active suicidal ideation class and low levels of suicidal plan and suicidal attempt as compared to those of the non-suicidal ideation class. Interestingly, two classes vulnerable to suicidality were identified, and there were differences in the level of risk between the two classes. The participants belonging to the active suicidal ideation class had a higher suicidality as they showed the highest likelihood of suicidal plan and attempt. Accordingly, active intervention for suicide prevention is needed for adolescents who belong to this class. In contrast, the participants belonging to the passive suicidal ideation class had low likelihood of suicidal plan and suicidal attempt but still showed high levels of sadness/hopelessness, loneliness, and moderately elevated levels of suicidal thought. Therefore, emotional and psychological intervention and assessment and continued monitoring of the possibility of worsening suicidality are needed for adolescents belonging to this class.
Second, examination of factors that influence the classification of latent classes of suicidality showed that violence victimization experience, sexual activity experience, deviant behavior (smoking and drinking), household income, sex, and year in school were all identified as factors with significant influence on the classification of latent classes. With respect to specific factors, dietary habit was found to have an influence in differentiating between non-suicidal and passive suicidality and between non-suicidal and active suicidality. The results also showed that adolescents who skipped breakfast or ate fast foods more often were relatively likely to be vulnerable. Such findings were consistent with those of previous studies reporting that unhealthy dietary habit is highly associated with neurosis [28,29]. Therefore, applying specific and systematic programs for improving dietary habits should be helpful for these adolescents. Such programs should include education that emphasizes the importance of regular meals and provision of well-balanced diet, which can be expected to contribute to improving mental health and reducing suicidality among adolescents.
Among physical activities, walking exercise and frequency of exercise during PE class were factors that should be examined with greatest importance among the findings in the present study. This is because these were factors that influenced the differentiation of the two classes that were identified as suicide risk classes. The findings showed that adolescents with less walking exercise and lower frequency of exercise during PE class were more likely to belong to the active than the passive suicidality class. Adolescents performing more walking workouts and having higher frequency of performing exercise during PE class belonged to the passive suicidality class; such individuals had suicidal thoughts but did not reach the stage of suicidal plan and attempt. Accordingly, it can be inferred that walking exercise and exercise during PE class can act as protective factors to prevent suicidal attempts. Such conclusion is consistent with those reported in previous studies, which stated that physical activities had a positive effect on psychological outcomes, such as depression, hopelessness, anxiety, and mental well-being [30,31]. Particularly, considering that exercise had the effect of reducing suicidal impulse [32], walking exercise and active PE during PE class in school can help reduce suicide risk among adolescents. However, unlike previous studies reporting that physical activities had a positive effect on mental health [30], our findings showed that there were no significant differences in high-intensity physical activities between all classes. In contrast, adolescents with higher frequency of moderate-intensity physical activities and level of muscle strengthening exercise were more likely to belong to the active or passive suicidality class than the non-suicidality class. Such findings should be interpreted with caution. Although the Youth Health Behavior Survey includes assessment of the intensity of physical activities, it does not include what type of physical activities were performed under what circumstances. There is a possibility that the findings in the present study may have been influenced by the inclusion of adolescents who were engaged in physical activities not as a leisure activity, but rather as part of part-time work or under the pressure of academic excellence, as well as young athletes and those exercising for weight loss, may have influenced the results [33]. Exercise during PE class tended to be promoted when adolescents maintain a stable relationship with their peers, and, thus, it is believed to have produced results that are different from those of moderate-intensity physical activities. Based on these findings, it is necessary to gain deeper understanding of and identify factors that can negatively influence suicidality.
In the present study, adolescents with more violence victimization and sexual activity experiences were more likely to belong to active and passive suicidality classes. Violence victimization experience was found to be a predictor of suicidality in all classes. In particular, considering that studies have reported that violence victimization experience has a negative effect on the psychological state of the victim, including self-worth, hopelessness, and loneliness, which can lead to suicidal thought and behavior [16], there is a need for intervention to treat psychological difficulties that students may face after violence victimization experience. Moreover, previous studies have identified that sexual activity, smoking, and drinking were associated with increased depression, suicidal thought, and suicidal attempt in adolescents, while another study reported that tests for suicidality should be seriously considered when treating adolescents who report sexual activity, drinking, and smoking [18]. Considering these reports, screening for mental health issues of adolescents associated with such behaviors should help establish appropriate management plan for suicide prevention.
Furthermore, adolescents with health risk behaviors, such as smoking and drinking, were identified as having greater likelihood of belonging to suicidality group. Such findings have consistently appeared in previous studies [34]. As adolescents’ smoke and drink by their own choice and action, such behaviors can be managed and modified. Accordingly, education and intervention to prevent adolescents from engaging in such risky behaviors are needed. Recently, the smoking and drinking rates among Korean adolescents have been reduced, but since the COVID-19 pandemic, the rates are being maintained or increasing slightly, which indicates the need for anti-smoking and anti-drinking education in schools. Considering a previous study that the demonstrated the effectiveness of media literacy education for preventing drinking during adolescence [35], active use of school-based smoking and drinking media literacy education could help cessation of smoking and drinking, while continued education on cessation of smoking and drinking may have an effect on suicide prevention.
Academic grades were not a significant factor for differentiating between the suicide risk classes, especially between active and passive suicidality. However, it was a significant factor in the comparisons between non-suicidality and active suicidality and between non-suicidal and passive suicidality, with lower academic grades being associated with a higher suicidality. Academic grades of Korean adolescents can lead to academic stress, and, thus, that factor may be associated with suicidality. Therefore, although it cannot be claimed that all students with low academic grades have high suicidality, it is necessary to intervene to make sure that students with low academic grades do not exhibit academic stress and psychosocial problems.
Finally, middle school students, females, and students with lower household income were relatively more vulnerable to suicidality than their respective counterparts. Therefore, early intervention programs for middle school students should be designed for early screening and management of mental health issues among adolescents. Moreover, considering that female individuals and students with lower household income were relatively more vulnerable to suicidality, customized approach that considers sex-specific mental health differences and stress factors is also needed. In particular, availability of support systems with consideration for socioeconomic background should be checked and improvement of accessibility to mental health services should also be included.
Our findings emphasized the need for customized response strategies according to the level of suicidality when developing effective interventional strategies of suicide prevention for adolescents. The study focused on identifying vulnerable populations and identifying the characteristics of such populations. Influencing factors of suicide did not consist of a single factor, but rather, the influence was diverse and complex. Considering this point, accurately screening and selecting the target of intervention still remains a challenge. Therefore, the present study applied a holistic approach for identifying and measuring the factors that influence classification of suicidality classes from their daily life. Through this approach, we provided more specific and practical intervention measures that can be applied in clinical and educational settings. Our findings suggest that clinicians in clinical settings might find it beneficial to consider adolescents’ engagement in physical activities, such as walking and PE classes, as potential indicators to gauge suicide risk. This approach could help in identifying those who might be at higher risk, particularly if they are less active and have known risk factors. By incorporating this perspective into their assessment, clinicians may be able to tailor interventions in a more nuanced way, potentially improving early support for at-risk youth.
The present study had some limitations. First, cross-sectional data were used. Therefore, unlike data from panel surveys, continuous characteristics of individuals could not be identified and the causal relationship could not be accurately determined. Risky sexual activity, drinking, smoking, and low academic grades, which were assessed in the present study, could have led to interesting outcomes after the onset of a mental disorder. If long-term data can be analyzed, how suicidality of individuals changes and which factors influence such change can be analyzed longitudinally. Secondly, the Youth Health Behavior Survey is a self-reporting survey and the survey may include adolescents who provided insincere responses. To minimize this, the KDCA provided the data after data cleansing by imputing data with logical errors or outliers. Third, the definition of physical activity is unclear in the data provided by KYRBS. Depending on the type of physical activity and the individual’s attitude toward physical activity, the impact on emotions may be variable. Future research should confirm the impact of the type of physical activity and individual perception of physical activity on mental health. Additionally, research is needed on various daily activities that can affect the mental health of adolescents other than physical activity.
Despite these limitations, our study is important as it not only helped categorize the participants into groups of adolescents with and without having a suicidality but also categorized the suicidality groups into those with active and passive suicidality. Such differentiation enables preemptive intervention for adolescents who exhibit suicidality, while identification of major factors that influence the differentiation of suicidality classes enables classification of the suicidality of an individual according to his/her behavior. Moreover, considering that these factors are modifiable behaviors in daily life, adolescents who exhibit such risk can be screened more specifically in clinical and educational settings. Our findings emphasized the importance of multi-dimensional approach. Moreover, our study contributed to the literature as it highlighted the need for customized education focusing on the lifestyle habits of adolescents and provided the direction for future intervention.
Notes
Availability of Data and Material
The datasets generated or analyzed during the current study are available in the Korea Disease Control and Prevention Agency repository https://www.kdca.go.kr/yhs/home.jsp.
Conflicts of Interest
The authors have no potential conflicts of interest to disclose.
Author Contributions
Conceptualization: Seojung Kim, Jongha Lee. Data curation: Seojung Kim, SuHyuk Chi, Boram Chae. Formal statistical analysis: Seojung Kim. Funding acquisition: Jongha Lee. Methodology: Seojung Kim, SuHyuk Chi, Boram Chae. Writing—original draft: Seojung Kim, Jongha Lee. Writing—review & editing: all authors.
Funding Statement
This study was supported by the Korea Disease Control and Prevention Agency (Q2315581) and a Korea University Ansan Hospital Grant (K2316001). The funders had no role in the design and conduct of the study or the decision to submit the article for publication.
Acknowledgements
None