This study used the Korean Youth Risk Behavior Survey (KYRBS), an anonymous, self-administered web-based survey that assesses the health behaviors among adolescents in South Korea. KYRBS is a nationally representative survey conducted annually by the Korea Centers for Disease Control and Prevention since 2005 to generate official statistics on adolescent health behaviors. For the present analysis, we used data collected between 2020 and 2024, comprising responses from 231,292 students aged 12-18 years enrolled in middle and high schools. Participation in the survey was voluntary, and the mean response rate over the study period was 93.6% [12]. The sampling design followed a structured three-stage process: population stratification, sample allocation, and sample selection. First, schools were stratified according to school characteristics and geographical region. Second, proportional allocation was applied to determine the sample size for each stratum. Finally, stratified cluster sampling was conducted, with schools and classes serving as the first and second sampling units, respectively. KYRBS data were collected anonymously. The study protocol was approved by the Institutional Review Board of the Korea Disease Control and Prevention Agency (2014-06EXP-02-P-A) and complied with the Population Health Promotion Act 19 (117058) of the Korean government. Approval for this secondary analysis was exempted by the Institutional Review Board of Kyung Hee University (KHSIRB-25-125). All participants provided written informed consent. For participants younger than 16 years, written informed consent was also obtained from a parent or legal guardian. The study was conducted in accordance with the principles of the Declaration of Helsinki.

To assess smartphone screen time, participants were asked, “On average, how many hours per day did you use your smartphone over the last 30 days?” The original responses, recorded in minutes, were converted into hours for analysis. Daily smartphone screen time was categorized into three groups: <2 hours, 2-4 hours, and >4 hours per day. These cut-off points were selected based on prior epidemiologic studies and empirical usage patterns observed in large-scale youth media surveys [13,14].

Loneliness was assessed with the question: “During the past 12 months, how often have you felt lonely [15]?” Depressive mood was measured using the question: “During the past 12 months, have you ever experienced sadness or despair severe enough to interfere with your daily activities for at least two weeks [16]?” Perceived stress was evaluated with the question: “How much stress do you usually experience [17]?”

This study included a range of sociodemographic, behavioral, and health-related covariates [16]. Sociodemographic characteristics comprised grade level (7-9th and 10-12th grade), sex (boys and girls), region of residence (urban and rural) [18], type of residence (living with family and with friends, alone, in a dormitory, or in a facility), household economic status (low, middle-low, middle, middle-high, and high), school performance (low, middle-low, middle, middle-high, high), parental educational level (high school or less and university or higher), and multicultural family status (yes and no). Behavioral lifestyle factors included vigorous physical activity (0-1, 2-3, and 4-7 days per week), sleep duration (short, moderate, and adequate), smoking status during the past month (nonsmoker and smoker), and alcohol consumption during the past month (non-drinker and drinker). Body mass index (BMI) classification was classified according to the 2017 Korean Children and Adolescent Growth Charts, because standard adult BMI criteria are not applicable to adolescents. BMI categories were defined as underweight (<5th percentile), normal weight (5th-84th percentile), overweight (85th-94th percentile), and obesity (≥95th percentile) [17]. Based on the literature and expert knowledge [19], we illustrated the hypothesized causal relationship using a directed acyclic graph representing the effect of all covariates in relation to screen time and the mental health outcomes (Supplementary Figure 1).

Using data from the KYRBS, differences in the distribution of sociodemographic, behavioral, and health-related variables across screen time groups (<2, 2-4, and >4 hours per day) were assessed using survey-weighted chi-square tests that accounted for the complex sampling design [20]. We then conducted survey-weighted analyses to examine the association between smartphone screen time and adolescent mental health (loneliness, depressive mood, and perceived stress). To evaluate these associations, we fitted multivariable logistic regression models adjusting for sociodemographic, behavioral, and health-related covariates. Weighted odds ratios (wORs) and 95% confidence intervals (CIs) were calculated.

To further explore the functional form of the association between screen time and each mental health outcome, we applied generalized additive models with a continuous specification of screen time. To identify particularly vulnerable groups, we conducted stratified analyses by sex, residential area, grade level, residence type, multicultural family status, BMI group, school performance, parental education level, household economic status, vigorous physical activity, sleep duration, smoking status, and alcohol consumption. In addition, we examined whether these associations differed across subgroups by including interaction terms between smartphone screen time categories and each subgroup variable in separate models. In addition, we performed a causal mediation analysis using sleep duration as a hypothesized mediator. All analyses incorporated the sampling weights, stratification, and clustering of the KYRBS design and were performed using SAS software (version 9.4; SAS Institute Inc.). Two-sided tests were used, and a p-value<0.05 was considered statistically significant [21].

This study showed that higher levels of smartphone screen time were associated with increased risks of loneliness and depressive mood, with the highest use (>4 hours per day) consistently showing elevated odds of these symptoms. By contrast, perceived stress followed a nonlinear pattern: adolescents reporting 2-4 hours per day of smartphone use had slightly lower odds of high stress than those with <2 hours per day, whereas stress increased again among those with >4 hours per day. Subgroup analyses showed that the associations between heavy smartphone use and both loneliness and depressive mood were more pronounced among middle school-aged adolescents, girls, and underweight adolescents. The associations were also higher among those with higher academic performance, higher socioeconomic status, and longer sleep dura-tion, indicating that these factors may amplify vulnerability to the adverse mental health effects of excessive smartphone use. For perceived stress, stronger associations with longer smartphone use were observed among middle school students, underweight adolescents, and those reporting moderate or adequate sleep duration. Overall, these findings suggest that heavy smartphone use is linked to adverse mental health outcomes among adolescents, and that these associations are amplified in specific sociodemographic and behavioral subgroups.

Several psychological and physiological pathways may help explain the observed associations between heavy smartphone use and poor mental health in adolescents. First, prolonged screen exposure, particularly in the evening, disrupts biological processes that regulate sleep and emotion. Continuous bluelight exposure suppresses melatonin secretion and delays circadian rhythm [22], which results in shorter sleep duration and poorer sleep quality. These disturbances in sleep patterns impair mood regulation, heighten physiological stress reactivity, and increase vulnerability to depressive symptoms and emotional instability [23,24]. Second, intensive use of social networking platforms promotes frequent social comparison, reduced self-esteem, and increased anxiety. Adolescents are particularly sensitive to peer evaluation and online exclusion, which may intensify the emotional impact of these interactions [25]. Third, excessive smartphone use may displace in-person social contact, contributing to social isolation and loneliness [26]. These adverse processes appear particularly pronounced when daily smartphone use exceeds approximately 4 hours, a level that previous studies have suggested may be associated with increased stress, suicidal ideation, and risk behaviors such as substance use [9]. Therefore, these mechanisms underscore the complex and interconnected pathways through which smartphone overuse may adversely affect adolescent mental health.

A growing body of international research supports the association between prolonged screen exposure and poorer mental health in adolescents. In the United States, adolescents who spend long hours on non-academic screens are more likely to report irregular sleep, lower physical activity, weight concerns, and more frequent depressive and anxiety symptoms [27]. Similarly, studies from Spain, South Korea, and China show that exceeding recommended screen time, particularly smartphone use over 4 hours per day, is associated with poorer health-related quality of life, higher stress, poorer sleep, and more severe depressive and suicide-related symptoms, often partly mediated by reduced sleep duration [17,24,28]. Additional population-based studies from the United States and Europe further indicate that prolonged smartphone use is linked to increased psychological distress, with stronger effects among girls and younger adolescents [29,30]. Our findings are broadly consistent with previous studies, showing that higher levels of smartphone use were linked to greater risks of depressive mood and loneliness among adolescents, and that very high levels of use were unfavorable for perceived stress [31].

However, relatively few studies have examined these associations using nationally representative data collected over multiple years [21,32]. Most existing literatures have been constrained by small or localized samples [32], and inconsistent definitions of screen time and mental health indicators [17], which can contribute to mixed and sometimes conflicting results. Although prior studies have suggested relationships between smartphone overuse and disrupted sleep, elevated stress, and depressive symptoms, they have rarely incorporated temporal trends or detailed subgroup comparisons. In this context, the present study extends previous research by using 5 years of nationally representative data focused exclusively on adolescents, a population undergoing critical psychological and behavioral development.

The findings of this study indicate a consistent association between heavy smartphone use and poor mental health among adolescents, particularly middle school students and girls. These results suggest that routine assessment of digital media use in clinical and school settings may help identify adolescents at risk for psychological difficulties, including depressive mood, loneliness, and related social withdrawal [21]. Clinicians could incorporate questions about smartphone use into standard psychosocial assessments and consider early intervention strategies, such as psychoeducation, behavioral counselling, and family-centered programs on healthy digital habits, especially for high-risk groups such as girls, younger adolescents, or socially isolated youth [33]. At the policy level, coordinated efforts by schools, families, and government agencies are needed to mitigate mental health risks potentially associated with smartphone overuse. Several countries, including Canada and Australia, recommend that children and adolescents limit noneducational screen time to <2 hours per day [34,35]. The WHO similarly advises that screen time should not displace essential activities such as sleep, physical activity, and in-person social interaction, particularly in children and adolescents [2]. Despite these international recommendations, South Korea, one of the countries with the highest rates of smartphone ownership among young people, has not yet adopted national guidelines on adolescent screen time, even as national data document both high levels of smartphone use and increasing concern about youth mental health. Our findings support consideration of national guidelines that encourage adolescents to limit recreational smartphone use, particularly beyond the level of >4 hours per day, at which we observed elevated risks of adverse mental health outcomes. Integrating such recommendations into school-based digital literacy curricula and parent education programs may help promote healthier patterns of digital engagement and reduce mental health risks in vulnerable groups.

This study has several limitations. First, its cross-sectional design precludes causal inference, and it remains uncertain whether excessive smartphone use contributes to poor mental health or whether adolescents with existing psychological difficulties are more likely to engage in heavy smartphone use. However, by analyzing 5 consecutive years of nationally representative data and observing consistent patterns across multiple mental health indicators, we provide a more stable characterization of these associations than single-wave surveys. Second, both smartphone screen time and mental health indicators were assessed using self-report questionnaires, and the mental health measures relied on single items rather than validated multi-item scales or clinical diagnoses. This approach may have introduced reporting or misclassification bias and may not fully capture the severity or spectrum of underlying conditions, although the KYRBS employs standardized, repeatedly field-tested survey protocols in a very large sample, which helps to enhance reliability. Third, we lacked information on the purpose and content of smartphone use and on important contextual and psychosocial factors, such as family relationships, parental monitoring, and peer interactions, limiting our ability to distinguish between beneficial and harmful digital behaviors or to fully adjust for environmental influences. Fourth, the KYRBS includes only adolescents enrolled in school; out-of-school adolescents, who may have different patterns of smartphone use and mental health risk, are therefore likely underrepresented, and caution is needed when generalizing these findings to all adolescents. Finally, although we used meaningful screen-time categories to reflect potential public health thresholds, any categorization involves some loss of information, and residual non-linear associations may persist despite robustness analyses treating screen time as a continuous variable. Despite these limitations, the study’s large, nationally representative sample, use of a weighted complex sampling design, extensive multivariable adjustment, and sys-tematic subgroup analyses provide robust and generalizable evidence on the associations between smartphone use and adolescent mental health.

This study found that heavy smartphone use was significantly associated with adverse mental health outcomes, including loneliness, depressive mood, and perceived stress, among Korean adolescents. Adolescents who used smartphones for >4 hours per day had higher risks of these outcomes (loneliness, depressive mood, and perceived stress), with particularly strong associations observed among middle school students and girls. These findings highlight the need for interventions that address not only the duration of smartphone use but also its purpose and social context. Early identification of high-risk groups, provision of appropriate emotional and behavioral support, and promotion of healthier smartphone use may help protect adolescent mental health and guide effective public health policies.