Social media use frequency and adolescent mental health: context-dependent associations by school connectedness and parental monitoring in the 2023 YRBS

Social media use frequency and adolescent mental health: context-dependent associations by school connectedness and parental monitoring in the 2023 YRBS | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Social media use frequency and adolescent mental health: context-dependent associations by school connectedness and parental monitoring in the 2023 YRBS Nikesh Lagun This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9007537/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Objective To examine associations between social media use frequency and poor mental health among U.S. adolescents and to assess whether these associations vary by school connectedness and parental monitoring. Methods We analyzed data from the 2023 National Youth Risk Behavior Survey, a nationally representative survey of U.S. students in grades 9–12 (n = 10,340). Social media use frequency was modeled as an ordered exposure, and poor mental health was defined as reporting mental health “not good” most of the time or always in the past 30 days. Survey-weighted logistic regression models estimated associations between social media use and poor mental health, adjusting for grade, sex, and race/ethnicity, with effect modification assessed using interaction terms. Results Overall, 28.5% (95% CI: 26.7–30.4) of adolescents reported poor mental health, and 77.0% reported high-frequency social media use. Higher social media use frequency was associated with higher adjusted odds and predicted probabilities of poor mental health. Associations differed by social context (interaction p < 0.01). At the highest levels of social media use, adolescents reporting low school connectedness had approximately 15 percentage points higher predicted probability of poor mental health than those reporting high connectedness, with similar differences observed by parental monitoring. Conclusions Associations between social media use frequency and adolescent mental health are context-dependent. Supportive school and family environments are associated with lower risk across levels of social media use, highlighting the importance of social context in adolescent mental health prevention research. Epidemiology Adolescents Social media use Mental health School connectedness Parental monitoring Youth Risk Behavior Survey Effect heterogeneity Survey-weighted analysis Figures Figure 1 Figure 2 1. Introduction Social media platforms are a central arena of social interaction for adolescents, shaping how young people communicate, seek support, and construct social identity. Alongside this ubiquity, concerns have grown regarding associations between social media use and adolescent mental health. A substantial body of research has examined links between digital media use and internalizing symptoms, psychological distress, and well-being; however, findings remain mixed and contested [ 1 – 5 ]. Recent reviews and meta-analyses consistently conclude that observed associations are typically small in magnitude, heterogeneous across individuals, and sensitive to measurement and analytic choices [ 1 , 3 , 6 , 7 ]. As a result, the field has increasingly moved away from generalized claims that social media use is uniformly harmful or beneficial toward interpretations that emphasize variability in adolescents’ experiences. One reason for this inconsistency is substantial variation across study designs, exposure definitions, and outcome measures. Some studies report positive associations between higher digital media use and depressive symptoms or psychological distress, particularly following the widespread adoption of smartphone-based social media after 2010 [ 2 , 5 , 8 ]. Other work, including longitudinal and diary-based studies, reports weak, null, or highly person-specific associations, challenging simple causal narratives [ 1 , 9 , 10 ]. Competing theoretical perspectives further complicate interpretation. Displacement models emphasize the possibility that excessive online engagement may crowd out sleep or face-to-face interaction [ 11 ], whereas social compensation and support perspectives highlight that online platforms may enhance social connection for some adolescents [ 6 , 12 ]. Together, this literature suggests that average associations may obscure meaningful heterogeneity in how adolescents experience and are affected by social media use. Developmental theory provides a strong basis for expecting such heterogeneity during adolescence. Adolescence is a sensitive period for socioemotional development, characterized by heightened responsiveness to peer evaluation, social belonging, and interpersonal feedback [ 13 , 14 ]. From an ecological perspective, adolescents’ experiences are shaped by multiple interacting social contexts, including schools, families, and peer networks [ 15 ]. Digital behaviors, therefore, do not operate in isolation; the same level of social media exposure may be experienced as supportive, neutral, or stressful depending on the offline relational environments in which adolescents are embedded. This framework implies that associations between social media use and mental health may be context-dependent rather than uniform across individuals. School connectedness represents one such contextual factor with well-established links to adolescent mental health and adjustment. Defined as students’ perceived closeness, belonging, and support within the school environment, school connectedness has been consistently associated with lower levels of emotional distress, substance use, and other risk behaviors [ 16 – 20 ]. Students who feel connected to their school also tend to report greater emotional support and stronger peer relationships, which are independently associated with better mental health outcomes [ 21 ]. Beyond its role as a protective correlate, ecological models suggest that school connectedness may shape how adolescents interpret and respond to social experiences, including those occurring online [ 22 ]. These considerations motivate the expectation that school connectedness may modify, rather than simply confound, associations between social media use frequency and mental health. Parental monitoring constitutes a second key social context relevant to adolescents’ digital lives. Conceptualized not merely as parental control but as parental knowledge of adolescents’ activities and social environments, parental monitoring reflects adolescents’ disclosure within supportive family relationships [ 23 ]. Higher levels of monitoring have been associated with lower psychological distress and reduced engagement in problem behaviors [ 23 , 24 ]. In the context of social media use, parental monitoring may influence adolescents’ coping strategies, boundary-setting, and norms around online engagement, shaping how online experiences are interpreted and managed. Consistent with stress-buffering models of social support, supportive parental involvement may therefore condition associations between frequent social media use and mental health outcomes [ 25 , 26 ]. Despite extensive theoretical and empirical work on adolescent development, social connectedness, and family processes, relatively few studies have directly tested whether associations between social media use frequency and adolescent mental health vary systematically across school and family contexts in nationally representative samples. Much existing population-based work, including analyses using large surveys, typically treats school and family characteristics as adjustment covariates rather than as potential moderators, limiting insight into whether and how these contexts alter exposure–outcome gradients [ 3 , 6 , 7 ]. Recent reviews explicitly call for research that moves beyond average effects and examines heterogeneity in digital media associations using large-scale population data [ 3 , 6 , 7 ]. To address this gap, the present study uses data from the 2023 National Youth Risk Behavior Survey (YRBS) [ 27 ] to examine associations between social media use frequency and poor mental health among U.S. students in grades 9–12, and to test whether these associations differ by levels of school connectedness and parental monitoring. Unlike prior nationally representative analyses that focus on adjustment alone, this study explicitly evaluates moderation using an interaction framework and emphasizes adjusted predicted probabilities to improve interpretability. Guided by ecological and stress-buffering perspectives [ 15 , 25 ], we pursue three aims: (1) to estimate the association between social media use frequency and poor mental health; (2) to test moderation by school connectedness; and (3) to test moderation by parental monitoring. We hypothesize that higher social media use frequency will be associated with a higher prevalence of poor mental health and that this association will be weaker among adolescents reporting greater school connectedness and higher parental monitoring. 2. Method 2.1. Data source and study design This study used data from the 2023 National Youth Risk Behavior Survey (YRBS) [ 27 ], a cross-sectional, school-based survey designed to generate nationally representative estimates of health behaviors and psychosocial indicators among U.S. students in grades 9–12. The national YRBS employs a three-stage cluster sampling design (schools, classes, students). All analyses incorporated survey weights, strata, and primary sampling unit (PSU) identifiers to account for the complex sampling design and to support population-level inference. Given the cross-sectional design, analyses were restricted to associational inference, and no claims regarding temporal ordering, directionality, or causality were made. 2.2. Sample and analytic population The 2023 national YRBS dataset contained 20,103 usable student questionnaires following standard CDC data processing and quality control procedures. The primary analytic sample was defined a priori by the availability of non-missing responses for (i) the primary exposure (Q80), (ii) the primary outcome (QN84), (iii) moderators (QN103, QN104), and (iv) pre-specified covariates (grade, sex, race/ethnicity). After applying these criteria, the final analytic sample comprised 10,340 respondents, corresponding to 51.4% of the original unweighted sample; 9,763 observations were excluded due to missingness on one or more required variables. To assess potential selection related to item nonresponse, we compared the weighted distributions of grade, sex, and race/ethnicity between included and excluded respondents using survey-adjusted descriptives. Differences were modest across these characteristics, suggesting that complete-case analyses are unlikely to substantially bias estimates, although reduced precision remains a limitation. 2.3. Measures Operational definitions, original response categories, and analytic codings for all study variables are provided in Table 1 . Table 1 Operationalization of study variables in the 2023 National YRBS (grades 9–12) Panel A. Primary exposure Construct YRBS variable Item label (YRBS) Original response categories (coded) Primary analytic specification Sensitivity specification Social media use frequency Q80 Social media 1. I do not use social media 2. A few times a month 3. About once a week 4. A few times a week 5. About once a day 6. Several times a day 7. About once an hour 8. More than once an hour Treated as ordinal (8-level ordered exposure) Re-coded to high-frequency user using derived dichotomy QN80 (see Panel D) Panel B. Primary outcome and secondary outcomes Construct YRBS variable Item label (YRBS) Original response categories (coded) Analytic coding Role Poor mental health (primary) Q84 Current mental health 1. Never 2. Rarely 3. Sometimes 4. Most of the time 5. Always Primary outcome operationalized using QN84 (derived binary indicator: “most of the time/always” vs other) Primary outcome Persistent sadness/hopelessness (optional secondary) Q26 Sad/hopeless for 2 + weeks (YRBS standard binary item; derived as QN26 in dataset) Used only as a robustness outcome (swap-out) Sensitivity outcome Sleep duration (optional secondary) Q85 Hours of sleep on a school night 1. ≤4h 2. 5h 3. 6h 4. 7h 5. 8h 6. 9h 7. ≥10h Modeled as ordered categories or dichotomized (e.g., < 8 vs ≥ 8) only if pre-specified Secondary (optional) Academic functioning (optional descriptive/secondary) Q87 Grades in school 1. Mostly A’s 2. Mostly B’s 3. Mostly C’s 4. Mostly D’s 5. Mostly F’s 6. None 7. Not sure Used descriptively or as a robustness correlate (not required) Secondary (optional) Panel C. Moderators (social context) Construct YRBS variable Item label (YRBS) Original response categories (coded) Analytic coding (pre-specified) Role in models School connectedness Q103 Feel close to people at their school 1. Strongly agree 2. Agree 3. Not sure 4. Disagree 5. Strongly disagree Dichotomized via derived QN103: agree/strongly agree vs other Moderator (interaction with Q80) Parental monitoring Q104 Parental monitoring 1. Never 2. Rarely 3. Sometimes 4. Most of the time 5. Always Dichotomized via derived QN104: most of the time/always vs less Moderator (interaction with Q80) Panel D. CDC-derived dichotomies used for primary inference and robustness Derived indicator Source variable(s) Definition used in this study Purpose QN80 Q80 High-frequency social media use: “several times/day or more” (corresponds to Q80 categories 6–8) vs less Sensitivity exposure QN84 Q84 Poor mental health: “most of the time/always” (Q84 categories 4–5) vs other Primary outcome QN103 Q103 School connectedness: agree/strongly agree (Q103 categories 1–2) vs other Moderator QN104 Q104 Parental monitoring: most of the time/always (Q104 categories 4–5) vs less Moderator QN86 Q86 Unstable housing (binary indicator in derived set) Sensitivity covariate QN26 Q26 Sad/hopeless for ≥ 2 weeks (binary indicator in derived set) Sensitivity outcome Panel E. Covariates and survey design variables Variable type YRBS variable Label Coding / categories Analytic use Sociodemographic covariate Q3 In what grade are you 1. 9th 2. 10th 3. 11th 4. 12th 5. other/ungraded Adjusted covariate (pre-specified) Sociodemographic covariate Q2 What is your sex 1. Female 2. Male Adjusted covariate (pre-specified) Sociodemographic covariate RACEETH Race/Ethnicity 1. AI/AN 2. Asian 3. Black 4. NH/PI 5. White 6. Hispanic/Latino 7. Multiple–Hispanic 8. Multiple–Non-Hispanic Adjusted covariate (pre-specified) Structural context (optional) Q86 / QN86 Unstable housing Q86 has 7 residence categories; QN86 is derived binary Sensitivity covariate Survey design WEIGHT Overall analysis weight Continuous Required for national estimates Survey design STRATUM Sampling strata Categorical Required for correct SEs Survey design PSU Primary sampling unit Categorical Required for correct SEs Note. The table reports item labels and response categories as specified in the YRBS 2023 SPSS setup syntax. Primary models use Q80 as an ordered exposure and QN84 as the primary outcome. CDC-derived dichotomies (QN variables) are used for moderation and sensitivity analyses to align with established YRBS reporting conventions. 2.3.1. Exposure: social media use frequency Social media use frequency was assessed using Q80, an eight-category ordered item ranging from non-use to more than once an hour (Table 1 , Panel A). The primary analytic specification treated Q80 as an ordinal exposure to preserve graded differences in use frequency rather than imposing an a priori threshold. Treating Q80 as ordinal assumes a monotonic association between increasing use frequency and poor mental health. To evaluate this assumption, we estimated fully categorical models and visually inspected category-specific adjusted predicted probabilities. These assessments did not indicate meaningful departures from monotonicity, and substantive conclusions were unchanged across ordinal and categorical specifications. In pre-specified sensitivity analyses, we used the CDC-derived dichotomous indicator QN80, classifying respondents who reported using social media several times per day or more frequently (Q80 categories 6–8) versus all others (Table 1 , Panel D), to assess robustness to exposure specification. 2.3.2. Primary outcome: poor mental health The primary outcome was poor mental health, measured using QN84, which identifies respondents reporting that their mental health was not good most of the time or always during the past 30 days (Table 1 , Panel B/D). This measure captures recent self-reported psychological distress and is interpreted as an indicator of well-being rather than a clinical diagnosis. 2.3.3. Moderators: social context Two indicators of adolescents’ social context were examined as moderators (Table 1 , Panel C/D): School connectedness (QN103) : respondents who agreed or strongly agreed that they felt close to people at their school versus other responses. Parental monitoring (QN104) : respondents who reported that parents or other adults in their family knew where they were going or with whom most of the time or always, versus less frequent knowledge. Both moderators were operationalized as dichotomous variables to align with CDC reporting conventions and to prioritize interpretability of interaction effects, rather than to impose substantive thresholds. These measures were examined as conditioning factors consistent with ecological and stress-buffering perspectives, while remaining agnostic regarding causal mechanisms due to the cross-sectional design. 2.3.4. Covariates All primary models adjusted for a minimal, pre-specified set of sociodemographic covariates: grade, sex, and race/ethnicity (Table 1 , Panel E). This strategy was adopted to reduce the risk of overadjustment for variables plausibly downstream of social media use, such as sleep disruption or peer conflict. In sensitivity analyses, additional models included unstable housing (QN86) and bullying victimization to assess robustness to broader structural and psychosocial stressors. Inclusion of these covariates did not materially alter effect estimates or substantive conclusions. 2.4. Statistical analysis 2.4.1. Survey design handling All descriptive estimates and regression models incorporated YRBS sampling weights, strata, and PSUs using IBM SPSS Statistics Complex Samples procedures. Survey-weighted estimates and standard errors were computed using Taylor series linearization. Adjusted predicted probabilities were computed from fitted survey-weighted logistic regression models using marginal standardization. Ninety-five percent confidence intervals were derived using design-based variance estimation. 2.4.2. Primary model set We estimated a series of survey-weighted logistic regression models with poor mental health (QN84) as the outcome: Model 1 (primary association) : QN84 as a function of Q80 and covariates. Model 2 (school connectedness moderation) : Model 1 plus QN103 and an interaction term between Q80 and QN103. Model 3 (parental monitoring moderation) : Model 1 plus QN104 and an interaction term between Q80 and QN104. Interaction terms were specified a priori to test whether the association between social media use frequency and poor mental health varied across social contexts. 2.4.3. Interaction inference and reporting strategy Interaction terms were evaluated using design-based Wald tests. In both moderation models, interactions reflected differences in the slope of the association between social media use frequency and poor mental health, rather than parallel shifts in baseline risk. Primary inference emphasized adjusted predicted probabilities of poor mental health across Q80 categories, stratified by moderator status, with 95% confidence intervals. Odds ratios were reported to facilitate comparison with prior literature. Interaction results were summarized using predicted-probability contrasts and marginal effects. 2.5. Robustness analyses A limited set of pre-specified robustness checks evaluated the stability of conclusions: Exposure coding : ordinal Q80 versus binary QN80. Outcome substitution : persistent sadness or hopelessness (Q26/QN26). Non-linearity : comparison of ordinal and categorical exposure specifications. Stratified sensitivity : models stratified by sex and grade when survey-weighted denominators supported stable inference. No analyses beyond these checks were treated as confirmatory. 2.6. Ethics The YRBS is a publicly available, de-identified dataset. This study involved secondary analysis of existing data and did not require additional institutional review board approval. 3. Results 3.1. Sample characteristics (weighted) Weighted demographic and psychosocial characteristics are summarized in Table 2 . Grade was evenly distributed (9th: 26.4%; 10th: 25.8%; 11th: 24.2%; 12th: 23.3%), and the sample was balanced by sex (48.1% female; 51.9% male). By race/ethnicity, the largest group was White (non-Hispanic) (48.1%), followed by Black (non-Hispanic) (13.3%), Asian (4.3%), and other categories; the estimate corresponding to the confidence interval previously reported for Hispanic/Latino was 19.8% (95% CI: 16.2–23.9), correcting the earlier transcription error where the point estimate did not fall within its interval. High-frequency social media use (QN80) was common (77.0%). School connectedness (QN103) was reported by 55.3% of students, and high parental monitoring (QN104) by 84.0%. The weighted prevalence of poor mental health (QN84) was 28.5% (95% CI: 26.7–30.4). Table 2 Weighted demographic and psychosocial characteristics, YRBS 2023 (national) Characteristic Weighted % 95% CI Unweighted n Grade (Q3) 19,910 9th 26.4 [24.3, 28.6] 5,680 10th 25.8 [24.0, 27.6] 5,410 11th 24.2 [22.3, 26.3] 4,811 12th 23.3 [21.1, 25.7] 3,961 Sex (Q2) 19,945 Female 48.1 [46.0, 50.3] 9,884 Male 51.9 [49.7, 54.0] 10,061 Race/Ethnicity (RACEETH) 19,733 White (non-Hispanic) 48.1 [41.5, 54.9] 9,700 Black (non-Hispanic) 13.3 [9.2, 18.9] 1,791 Hispanic/Latino 19.8 [16.2, 23.9] 2,786 Asian 4.3 [2.9, 6.3] 995 Other / Multiple (combined) 14.5 — 4,461 High-frequency social media use (QN80) 77.0 [73.5, 80.1] 15,203 School connectedness (QN103) 55.3 [52.8, 57.8] 11,177 Parental monitoring (QN104) 84.0 [81.2, 86.5] 10,850 Poor mental health (QN84) 28.5 [26.7, 30.4] 15,705 Note. Percentages are survey-weighted. Confidence intervals use the YRBS complex design. Unweighted n reflects the number of respondents with non-missing data for each characteristic (available-case denominators), which can differ across rows. The regression models below use the pre-specified complete-case analytic sample (N = 10,340). 3.2. Primary association between social media use and poor mental health In survey-weighted models adjusted for grade, sex, and race/ethnicity, higher social media use frequency (Q80) was associated with a higher adjusted probability of poor mental health (QN84). Predicted probabilities increased across ordered categories of use, supporting the ordinal specification used in primary models. Because odds ratios are less intuitive for public health interpretation, results are presented primarily as adjusted predicted probabilities. 3.3. Moderation by school connectedness School connectedness (QN103) was evaluated as an effect modifier using a survey-weighted interaction model and summarized with adjusted predicted probabilities across Q80 categories (Fig. 1 ). Across all levels of social media use, adolescents reporting higher school connectedness had a lower adjusted probability of poor mental health than those reporting lower connectedness. As social media use increased, the adjusted probability of poor mental health rose in both groups, but the increase was steeper among adolescents reporting lower school connectedness. At the highest social media use category, the gap between groups was approximately 15 percentage points. The interaction was statistically significant by design-based Wald test (p < .01), indicating a difference in the exposure–outcome gradient by connectedness level. 3.4. Moderation by parental monitoring Parental monitoring (QN104) was examined in parallel survey-weighted interaction models (Fig. 2 ). At low levels of social media use, adolescents reporting higher parental monitoring had a lower adjusted probability of poor mental health than those reporting lower monitoring. With increasing social media use, predicted probabilities rose in both groups, but the rise was larger among adolescents reporting lower parental monitoring. At the highest use category, the difference between monitoring groups was again approximately 15 percentage points. The interaction was statistically significant by design-based Wald test (p < .01), consistent with effect modification by family context reflected in differences in the gradient across use categories. 3.5. Robustness and sensitivity analyses Robustness analyses substituting persistent sadness or hopelessness (Q26/QN26) for the primary outcome yielded qualitatively similar patterns to those observed for poor mental health (QN84). Descriptive, survey-weighted prevalence estimates by grade and sex are reported in Supplementary Table S1 using the available Q26 sample. These analyses were conducted to provide contextual comparison only and did not materially alter, nor were they used to draw, the primary inferences based on QN84. 4. Discussion 4.1. Principal findings In this nationally representative sample of U.S. adolescents, higher social media use frequency was associated with a higher prevalence of poor mental health. Importantly, this association was not uniform across adolescents. The strength of the association differed by social context: adolescents reporting higher school connectedness and greater parental monitoring exhibited consistently lower predicted probabilities of poor mental health across all levels of social media use, whereas those with lower connectedness or monitoring showed steeper exposure–outcome gradients. These findings indicate that the association between social media use frequency and adolescent mental health varies systematically by social context rather than operating as a uniform risk across the population. 4.2. Interpretation and relation to existing literature The present findings align with a growing body of work emphasizing heterogeneity in associations between digital media use and adolescent well-being [ 1 , 3 , 6 , 7 ]. Rather than supporting a simple harmful-use narrative, the results suggest that broader social environments shape how adolescents experience and are associated with frequent social media engagement. This interpretation is consistent with ecological models of development, which posit that individual behaviors and outcomes are embedded within interacting social contexts such as schools and families [ 15 ]. School connectedness and parental monitoring can be understood as indicators of relational support and social integration. Prior research has consistently shown that school connectedness is associated with lower emotional distress and more favorable mental health outcomes [ 16 – 20 ], while parental monitoring reflects supportive family processes linked to adolescent adjustment [ 23 , 24 ]. Within stress-related frameworks, these forms of social support may correspond to a flatter increase in mental health risk across increasing levels of online exposure, including experiences such as social comparison, peer conflict, or cyberbullying [ 25 , 28 , 29 ]. Several complementary mechanisms may underlie these patterns. Supportive school and family environments may facilitate emotional regulation, enabling adolescents to manage negative affect arising from online interactions. They may also contribute to norm-setting around time use, online conduct, and responses to peer feedback. In addition, such environments may support interpretive buffering, whereby adolescents embedded in supportive relationships are less likely to internalize negative online experiences or social comparisons as personally salient. Together, these mechanisms provide a theoretically grounded explanation for why similar levels of social media exposure may be associated with different mental health outcomes across adolescents. These findings also help clarify why prior empirical evidence on social media use and mental health has been inconsistent. Reviews routinely note that average associations often mask substantial between-person variation [ 1 , 3 , 5 ]. By directly testing moderation in a large, population-based dataset, the present study provides evidence that social context is one dimension along which this variation occurs. This context-dependent pattern is consistent with recent calls to move beyond global effect estimates toward analytic approaches that explicitly examine when and for whom digital media use is associated with poorer mental health [ 6 , 7 ]. Importantly, the results distinguish between two related but conceptually distinct phenomena. Differences in baseline levels of poor mental health across social contexts reflect risk stratification, whereas differences in the strength of the association between social media use frequency and poor mental health reflect effect heterogeneity. The present analyses provide evidence of both: adolescents embedded in more supportive contexts exhibit lower baseline risk, and the increase in predicted probability across social media use categories is flatter in these groups. Recognizing the coexistence of these patterns is essential for interpreting heterogeneous associations in population-based studies and for avoiding oversimplified conclusions about digital media use. 4.3. Strengths This study has several strengths. First, it draws on a large, nationally representative sample of U.S. adolescents, supporting generalizability to the national population. Second, all analyses incorporated the complex survey design of the YRBS, enabling valid population-level inference. Third, the analytic strategy was pre-specified, with a focused set of outcomes, moderators, and covariates, reducing the risk of selective reporting. Finally, moderation results were presented using adjusted predicted probabilities rather than relying solely on odds ratios, enhancing interpretability for applied and public health audiences. 4.4. Limitations Several limitations should be considered. The cross-sectional design precludes conclusions about temporal ordering, and reverse causation remains possible, such that adolescents experiencing poorer mental health may engage more frequently with social media. Selection into supportive school or family contexts may also contribute to observed moderation patterns; adolescents with better baseline mental health may be more likely to report higher connectedness or monitoring. All measures relied on self-report and may be subject to recall or reporting differences. Adolescents experiencing poorer mental health may perceive or report lower levels of social support independent of objective differences. Social media use was measured as frequency rather than duration, content, or platform-specific engagement, and the data do not capture qualitative differences between positive and negative online experiences. Finally, although school connectedness and parental monitoring capture important aspects of social context, they do not directly measure peer relationship quality or family dynamics. 4.5. Implications These findings underscore the importance of considering social context when interpreting associations between social media use frequency and adolescent mental health. Rather than framing frequent social media use as inherently harmful, the results indicate that associations differ meaningfully across social environments. At the highest levels of use, differences in predicted probability of poor mental health on the order of approximately 15 percentage points were observed between adolescents reporting higher versus lower levels of school connectedness or parental monitoring, highlighting the population-level relevance of contextual variation. At the same time, these findings do not imply that increased social media use is benign or that supportive contexts eliminate risk. Instead, they suggest that exposure operates differently across adolescents depending on broader social environments. This perspective aligns with broader evidence emphasizing that adolescent well-being reflects the interplay of individual behaviors and relational contexts [ 15 , 25 ]. While the present study does not support specific policy or clinical interventions, it reinforces the value of contextualized approaches to understanding adolescents’ digital lives and cautions against one-size-fits-all interpretations of social media effects. 5. Conclusion In a nationally representative sample of U.S. adolescents, higher social media use frequency was associated with a higher prevalence of poor mental health, with substantial variation by school connectedness and parental monitoring. These findings indicate that associations between social media use and adolescent mental health are context-dependent rather than uniform across the population. Adolescents embedded in more supportive school and family environments exhibited both lower baseline risk and a flatter increase in predicted probability across increasing levels of social media use. These results support the conclusion that social media use frequency is associated with adolescent mental health in ways that depend on broader social environments. At the same time, the cross-sectional design precludes conclusions about directionality, causality, or underlying mechanisms. Together, the findings underscore the importance of situating adolescents’ digital behaviors within the social contexts in which they occur and caution against one-size-fits-all interpretations of social media effects in population-based research. Declarations Conflict of interest The author declares no conflicts of interest. Ethical approval Ethical approval was not required for this study, as it involved secondary analysis of publicly available, de-identified data from the Youth Risk Behavior Surveillance System. The dataset contains no personally identifiable information and involves no direct interaction with human participants. Funding statement The author received no external funding for the preparation of this manuscript. Acknowledgments None. Data availability The data analyzed in this study are publicly available from the Centers for Disease Control and Prevention through the Youth Risk Behavior Surveillance System (YRBS). The specific dataset used was the 2023 National YRBS, accessible at https://www.cdc.gov/yrbs/data/index.html. All analyses were conducted using IBM SPSS Statistics (Complex Samples module) to account for the survey’s stratified cluster design. The analytic syntax and code used for data preparation and statistical modeling are available from the author upon reasonable request. ORCID iD Nikesh Lagun: https://orcid.org/0009-0005-6372-4852 CRediT author statement Nikesh Lagun: Conceptualization; Methodology; Formal analysis; Data curation; Software; Validation; Investigation; Visualization; Writing – Original Draft; Writing – Review & Editing; Project administration. Declaration of generative AI and AI-assisted technologies in the manuscript preparation process During the preparation of this manuscript, the authors used generative AI tools solely for language editing and improvement of clarity and writing quality. Following the use of these tools, the authors carefully reviewed, edited, and revised all content and take full responsibility for the accuracy, integrity, and originality of the published article. References Orben A, Przybylski AK (2019) The association between adolescent well-being and digital technology use. Nat Hum Behav 3(2):173–182. https://doi.org/10.1038/s41562-018-0506-1 Keles B, McCrae N, Grealish A (2020) A systematic review: the influence of social media on depression, anxiety and psychological distress in adolescents. Int J Adolesc Youth 25(1):79–93. https://doi.org/10.1080/02673843.2019.1590851 Odgers CL, Jensen MR (2020) Annual research review: adolescent mental health in the digital age: facts, fears, and future directions. J Child Psychol Psychiatry 61(3):336–348. https://doi.org/10.1111/jcpp.13190 Shannon H, Bush K, Villeneuve PJ, Hellemans KG, Guimond S (2022) Problematic social media use in adolescents and young adults: systematic review and meta-analysis, JMIR Ment. Health 9(4):e33450. https://doi.org/10.2196/33450 Fassi L, Thomas K, Parry DA, Leyland-Craggs A, Ford TJ, Orben A (2024) Social media use and internalizing symptoms in clinical and community adolescent samples: a systematic review and meta-analysis. JAMA Pediatr 178(8):814–822. https://doi.org/10.1001/jamapediatrics.2024.2078 Prinstein MJ, Nesi J, Telzer EH, Commentary: an updated agenda for the study of digital media use and adolescent development—future directions following Odgers, Jensen (2020) J. Child Psychol. Psychiatry 61 (3) (2020) 349–352. https://doi.org/10.1111/jcpp.13219 Valkenburg P, Beyens I, Keijsers L (2024) Investigating heterogeneity in (social) media effects: experience-based recommendations. Meta-Psychology 8. https://doi.org/10.15626/MP.2022.3649 Twenge JM, Joiner TE, Rogers ML, Martin GN (2018) Increases in depressive symptoms, suicide-related outcomes, and suicide rates among U.S. adolescents after 2010 and links to increased new media screen time. Clin Psychol Sci 6(1):3–17. https://doi.org/10.1177/2167702617723376 Orben A, Przybylski AK (2019) Screens, teens, and psychological well-being: evidence from three time-use-diary studies. Psychol Sci 30(5):682–696. https://doi.org/10.1177/0956797619830329 Coyne SM, Rogers AA, Zurcher JD, Stockdale L, Booth M (2020) Does time spent using social media impact mental health? an eight year longitudinal study. Comput Hum Behav 104:106160. https://doi.org/10.1016/j.chb.2019.106160 Twenge JM, Joiner TE, Martin G, Rogers ML (2018) Amount of time online is problematic if it displaces face-to-face social interaction and sleep. Clin Psychol Sci 6(4):456–457. https://doi.org/10.1177/2167702618778562 Weinstein E (2018) The social media see-saw: positive and negative influences on adolescents’ affective well-being. New Media Soc 20(10):3597–3623. https://doi.org/10.1177/1461444818755634 Steinberg L (2008) A social neuroscience perspective on adolescent risk-taking. Dev Rev 28(1):78–106. https://doi.org/10.1016/j.dr.2007.08.002 Blakemore SJ, Mills KL (2014) Is adolescence a sensitive period for sociocultural processing? Annu Rev Psychol 65:187–207. https://doi.org/10.1146/annurev-psych-010213-115202 Bronfenbrenner U (1979) The Ecology of Human Development: Experiments by Nature and Design. Harvard University Press, Cambridge, MA. https://doi.org/10.2307/j.ctv26071r6 Bonny AE, Britto MT, Klostermann BK, Hornung RW, Slap GB (2000) School disconnectedness: identifying adolescents at risk. Pediatrics 106(5):1017–1021. https://doi.org/10.1542/peds.106.5.1017 McNeely CA, Nonnemaker JM, Blum RW (2002) Promoting school connectedness: evidence from the National Longitudinal Study of Adolescent Health. J School Health 72(4):138–146. https://doi.org/10.1111/j.1746-1561.2002.tb06533.x Catalano RF, Haggerty KP, Oesterle S, Fleming CB, Hawkins JD (2004) The importance of bonding to school for healthy development: findings from the Social Development Research Group. J School Health 74(7):252–261. https://doi.org/10.1111/j.1746-1561.2004.tb08281.x Bond L, Butler H, Thomas L, Carlin J, Glover S, Bowes G, Patton G (2007) Social and school connectedness in early secondary school as predictors of late teenage substance use, mental health, and academic outcomes. J Adolesc Health 40(4):357e9–357e18. https://doi.org/10.1016/j.jadohealth.2006.10.013 Shochet IM, Dadds MR, Ham D, Montague R (2006) School connectedness is an underemphasized parameter in adolescent mental health: results of a community prediction study. J Clin Child Adolesc Psychol 35(2):170–179. https://doi.org/10.1207/s15374424jccp3502_1 Resnick MD, Bearman PS, Blum RW, Bauman KE, Harris KM, Jones J, Tabor J, Beuhring T, Sieving RE, Shew M, Ireland M, Bearinger LH, Udry JR (1997) Protecting adolescents from harm: findings from the National Longitudinal Study on Adolescent Health. JAMA 278(10):823–832. https://doi.org/10.1001/jama.278.10.823 Putnam RD (2000) Bowling Alone: The Collapse and Revival of American Community. Touchstone Books/Simon & Schuster, New York. https://doi.org/10.1145/358916.361990 Stattin H, Kerr M (2000) Parental monitoring: A reinterpretation. Child Dev 71(4):1072–1085. https://doi.org/10.1111/1467-8624.00210 Waizenhofer RN, Buchanan CM, Jackson-Newsom J (2004) Mothers' and fathers' knowledge of adolescents' daily activities: its sources and its links with adolescent adjustment. J Fam Psychol 18(2):348–360. https://doi.org/10.1037/0893-3200.18.2.348 Cohen S, Wills TA (1985) Stress, social support, and the buffering hypothesis. Psychol Bull 98(2):310–357. https://doi.org/10.1037/0033-2909.98.2.310 Foster H, Brooks-Gunn J (2009) Toward a stress process model of children's exposure to physical family and community violence. Clin Child Fam Psychol Rev 12(2):71–94. https://doi.org/10.1007/s10567-009-0049-0 Centers for Disease Control and Prevention (2023) Youth Risk Behavior Surveillance System (YRBSS) [dataset]. https://www.cdc.gov/yrbs/data/index.html , (accessed 28 January 2026) Hinduja S, Patchin JW (2010) Bullying, cyberbullying, and suicide, Arch. Suicide Res 14(3):206–221. https://doi.org/10.1080/13811118.2010.494133 Patchin JW, Hinduja S (2010) Cyberbullying and self-esteem. J School Health 80(12):614–624. https://doi.org/10.1111/j.1746-1561.2010.00548.x Additional Declarations The authors declare no competing interests. Supplementary Files supplement.pdf Weighted prevalence of persistent sadness or hopelessness by grade and sex, YRBS 2023 (national). Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9007537","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":599124379,"identity":"8bbff24d-384c-494b-95aa-00b908eb0c6b","order_by":0,"name":"Nikesh Lagun","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABFklEQVRIiWNgGAWjYBACPgglYcDAwMP4IKGCTQ7EPfAAjxY2JC3MBh/O8BmDtSQQ1sIA0sImObNFLrEBxMWrhf34NenCHRbG/PxnD0jzNpilzw87/BBoi52cbgMOLTw5ZdIzz0iYSc7ISzDm3ZGWu/F2mgFQS7Kx2QFcDstJk+Ztk7AxuMFjkMx75ljuxtkJIC0HErfh0sL/BqLF/vwZg8O8bf/TDWenf8CvRSL9GEiLmQFDjmHjzDa2BHnpHAK2SLxhtgZqMZa4kWPM8OEMm+EG6ZyCAwkGuP3Cz5/+8DZvW51hf/8Z8x/AqJSXn52++cOHCjs5XFqA0WGAyjcAqzTAohIO2B+g8uUb8KkeBaNgFIyCkQgA2p5cQBVdBDgAAAAASUVORK5CYII=","orcid":"https://orcid.org/0009-0005-6372-4852","institution":"Independent Researcher","correspondingAuthor":true,"prefix":"","firstName":"Nikesh","middleName":"","lastName":"Lagun","suffix":""}],"badges":[],"createdAt":"2026-03-02 08:24:50","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9007537/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9007537/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104400769,"identity":"ea6176be-55d5-4a56-96b3-edd410be1acb","added_by":"auto","created_at":"2026-03-11 12:10:58","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":79903,"visible":true,"origin":"","legend":"\u003cp\u003eAdjusted predicted probability of poor mental health (QN84) across social media use frequency categories (Q80), stratified by school connectedness (QN103). Shaded areas indicate 95% confidence intervals.\u003c/p\u003e","description":"","filename":"fig1.png","url":"https://assets-eu.researchsquare.com/files/rs-9007537/v1/bfc54d229a227a19c5af40bf.png"},{"id":103808623,"identity":"2fdeed93-8385-4d85-b74d-7726740b9bd6","added_by":"auto","created_at":"2026-03-03 07:48:23","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":72490,"visible":true,"origin":"","legend":"\u003cp\u003eAdjusted predicted probability of poor mental health (QN84) across social media use frequency categories (Q80), stratified by parental monitoring (QN104). Shaded areas indicate 95% confidence intervals.\u003c/p\u003e","description":"","filename":"fig2.png","url":"https://assets-eu.researchsquare.com/files/rs-9007537/v1/a3c3140131cecc61cfc507d2.png"},{"id":104412862,"identity":"78c8e28f-1d60-4206-9136-e509b99454e4","added_by":"auto","created_at":"2026-03-11 13:01:24","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1532459,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9007537/v1/61bc4549-1065-41ba-9f4b-45ec4f1e77af.pdf"},{"id":103808625,"identity":"3b3625ec-a0c5-41b1-b3b9-086e275a9efc","added_by":"auto","created_at":"2026-03-03 07:48:23","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":104199,"visible":true,"origin":"","legend":"\u003cp\u003eWeighted prevalence of persistent sadness or hopelessness by grade and sex, YRBS 2023 (national).\u003c/p\u003e","description":"","filename":"supplement.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9007537/v1/c0fa2e6831273fc250c34c55.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eSocial media use frequency and adolescent mental health: context-dependent associations by school connectedness and parental monitoring in the 2023 YRBS\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eSocial media platforms are a central arena of social interaction for adolescents, shaping how young people communicate, seek support, and construct social identity. Alongside this ubiquity, concerns have grown regarding associations between social media use and adolescent mental health. A substantial body of research has examined links between digital media use and internalizing symptoms, psychological distress, and well-being; however, findings remain mixed and contested [\u003cspan additionalcitationids=\"CR2 CR3 CR4\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Recent reviews and meta-analyses consistently conclude that observed associations are typically small in magnitude, heterogeneous across individuals, and sensitive to measurement and analytic choices [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. As a result, the field has increasingly moved away from generalized claims that social media use is uniformly harmful or beneficial toward interpretations that emphasize variability in adolescents\u0026rsquo; experiences.\u003c/p\u003e \u003cp\u003eOne reason for this inconsistency is substantial variation across study designs, exposure definitions, and outcome measures. Some studies report positive associations between higher digital media use and depressive symptoms or psychological distress, particularly following the widespread adoption of smartphone-based social media after 2010 [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Other work, including longitudinal and diary-based studies, reports weak, null, or highly person-specific associations, challenging simple causal narratives [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Competing theoretical perspectives further complicate interpretation. Displacement models emphasize the possibility that excessive online engagement may crowd out sleep or face-to-face interaction [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], whereas social compensation and support perspectives highlight that online platforms may enhance social connection for some adolescents [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Together, this literature suggests that average associations may obscure meaningful heterogeneity in how adolescents experience and are affected by social media use.\u003c/p\u003e \u003cp\u003eDevelopmental theory provides a strong basis for expecting such heterogeneity during adolescence. Adolescence is a sensitive period for socioemotional development, characterized by heightened responsiveness to peer evaluation, social belonging, and interpersonal feedback [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. From an ecological perspective, adolescents\u0026rsquo; experiences are shaped by multiple interacting social contexts, including schools, families, and peer networks [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Digital behaviors, therefore, do not operate in isolation; the same level of social media exposure may be experienced as supportive, neutral, or stressful depending on the offline relational environments in which adolescents are embedded. This framework implies that associations between social media use and mental health may be context-dependent rather than uniform across individuals.\u003c/p\u003e \u003cp\u003eSchool connectedness represents one such contextual factor with well-established links to adolescent mental health and adjustment. Defined as students\u0026rsquo; perceived closeness, belonging, and support within the school environment, school connectedness has been consistently associated with lower levels of emotional distress, substance use, and other risk behaviors [\u003cspan additionalcitationids=\"CR17 CR18 CR19\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. Students who feel connected to their school also tend to report greater emotional support and stronger peer relationships, which are independently associated with better mental health outcomes [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Beyond its role as a protective correlate, ecological models suggest that school connectedness may shape how adolescents interpret and respond to social experiences, including those occurring online [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. These considerations motivate the expectation that school connectedness may modify, rather than simply confound, associations between social media use frequency and mental health.\u003c/p\u003e \u003cp\u003eParental monitoring constitutes a second key social context relevant to adolescents\u0026rsquo; digital lives. Conceptualized not merely as parental control but as parental knowledge of adolescents\u0026rsquo; activities and social environments, parental monitoring reflects adolescents\u0026rsquo; disclosure within supportive family relationships [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Higher levels of monitoring have been associated with lower psychological distress and reduced engagement in problem behaviors [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. In the context of social media use, parental monitoring may influence adolescents\u0026rsquo; coping strategies, boundary-setting, and norms around online engagement, shaping how online experiences are interpreted and managed. Consistent with stress-buffering models of social support, supportive parental involvement may therefore condition associations between frequent social media use and mental health outcomes [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eDespite extensive theoretical and empirical work on adolescent development, social connectedness, and family processes, relatively few studies have directly tested whether associations between social media use frequency and adolescent mental health vary systematically across school and family contexts in nationally representative samples. Much existing population-based work, including analyses using large surveys, typically treats school and family characteristics as adjustment covariates rather than as potential moderators, limiting insight into whether and how these contexts alter exposure\u0026ndash;outcome gradients [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Recent reviews explicitly call for research that moves beyond average effects and examines heterogeneity in digital media associations using large-scale population data [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eTo address this gap, the present study uses data from the 2023 National Youth Risk Behavior Survey (YRBS) [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] to examine associations between social media use frequency and poor mental health among U.S. students in grades 9\u0026ndash;12, and to test whether these associations differ by levels of school connectedness and parental monitoring. Unlike prior nationally representative analyses that focus on adjustment alone, this study explicitly evaluates moderation using an interaction framework and emphasizes adjusted predicted probabilities to improve interpretability. Guided by ecological and stress-buffering perspectives [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e], we pursue three aims: (1) to estimate the association between social media use frequency and poor mental health; (2) to test moderation by school connectedness; and (3) to test moderation by parental monitoring. We hypothesize that higher social media use frequency will be associated with a higher prevalence of poor mental health and that this association will be weaker among adolescents reporting greater school connectedness and higher parental monitoring.\u003c/p\u003e"},{"header":"2. Method","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Data source and study design\u003c/h2\u003e \u003cp\u003eThis study used data from the 2023 National Youth Risk Behavior Survey (YRBS) [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e], a cross-sectional, school-based survey designed to generate nationally representative estimates of health behaviors and psychosocial indicators among U.S. students in grades 9\u0026ndash;12. The national YRBS employs a three-stage cluster sampling design (schools, classes, students).\u003c/p\u003e \u003cp\u003eAll analyses incorporated survey weights, strata, and primary sampling unit (PSU) identifiers to account for the complex sampling design and to support population-level inference. Given the cross-sectional design, analyses were restricted to associational inference, and no claims regarding temporal ordering, directionality, or causality were made.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Sample and analytic population\u003c/h2\u003e \u003cp\u003eThe 2023 national YRBS dataset contained 20,103 usable student questionnaires following standard CDC data processing and quality control procedures. The primary analytic sample was defined a priori by the availability of non-missing responses for (i) the primary exposure (Q80), (ii) the primary outcome (QN84), (iii) moderators (QN103, QN104), and (iv) pre-specified covariates (grade, sex, race/ethnicity).\u003c/p\u003e \u003cp\u003eAfter applying these criteria, the final analytic sample comprised 10,340 respondents, corresponding to 51.4% of the original unweighted sample; 9,763 observations were excluded due to missingness on one or more required variables.\u003c/p\u003e \u003cp\u003eTo assess potential selection related to item nonresponse, we compared the weighted distributions of grade, sex, and race/ethnicity between included and excluded respondents using survey-adjusted descriptives. Differences were modest across these characteristics, suggesting that complete-case analyses are unlikely to substantially bias estimates, although reduced precision remains a limitation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Measures\u003c/h2\u003e \u003cp\u003eOperational definitions, original response categories, and analytic codings for all study variables are provided in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eOperationalization of study variables in the 2023 National YRBS (grades 9\u0026ndash;12)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003ePanel A. Primary exposure\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eConstruct\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eYRBS variable\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eItem label (YRBS)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eOriginal response categories (coded)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003ePrimary analytic specification\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003eSensitivity specification\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial media use frequency\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSocial media\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1. I do not use social media\u003c/p\u003e \u003cp\u003e2. A few times a month\u003c/p\u003e \u003cp\u003e3. About once a week\u003c/p\u003e \u003cp\u003e4. A few times a week\u003c/p\u003e \u003cp\u003e5. About once a day\u003c/p\u003e \u003cp\u003e6. Several times a day\u003c/p\u003e \u003cp\u003e7. About once an hour\u003c/p\u003e \u003cp\u003e8. More than once an hour\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTreated as ordinal (8-level ordered exposure)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRe-coded to high-frequency user using derived dichotomy QN80 (see Panel D)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003ePanel B. Primary outcome and secondary outcomes\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eConstruct\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eYRBS variable\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eItem label (YRBS)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eOriginal response categories (coded)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003eAnalytic coding\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003eRole\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePoor mental health (primary)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCurrent mental health\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1. Never\u003c/p\u003e \u003cp\u003e2. Rarely\u003c/p\u003e \u003cp\u003e3. Sometimes\u003c/p\u003e \u003cp\u003e4. Most of the time\u003c/p\u003e \u003cp\u003e5. Always\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ePrimary outcome operationalized using QN84 (derived binary indicator: \u0026ldquo;most of the time/always\u0026rdquo; vs other)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003ePrimary outcome\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePersistent sadness/hopelessness (optional secondary)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSad/hopeless for 2\u0026thinsp;+\u0026thinsp;weeks\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(YRBS standard binary item; derived as QN26 in dataset)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUsed only as a robustness outcome (swap-out)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSensitivity outcome\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSleep duration (optional secondary)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHours of sleep on a school night\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1. \u0026le;4h\u003c/p\u003e \u003cp\u003e2. 5h\u003c/p\u003e \u003cp\u003e3. 6h\u003c/p\u003e \u003cp\u003e4. 7h\u003c/p\u003e \u003cp\u003e5. 8h\u003c/p\u003e \u003cp\u003e6. 9h\u003c/p\u003e \u003cp\u003e7. \u0026ge;10h\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eModeled as ordered categories or dichotomized (e.g., \u0026lt;\u0026thinsp;8 vs\u0026thinsp;\u0026ge;\u0026thinsp;8) only if pre-specified\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSecondary (optional)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAcademic functioning (optional descriptive/secondary)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGrades in school\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1. Mostly A\u0026rsquo;s\u003c/p\u003e \u003cp\u003e2. Mostly B\u0026rsquo;s\u003c/p\u003e \u003cp\u003e3. Mostly C\u0026rsquo;s\u003c/p\u003e \u003cp\u003e4. Mostly D\u0026rsquo;s\u003c/p\u003e \u003cp\u003e5. Mostly F\u0026rsquo;s\u003c/p\u003e \u003cp\u003e6. None\u003c/p\u003e \u003cp\u003e7. Not sure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUsed descriptively or as a robustness correlate (not required)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSecondary (optional)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabb\" border=\"1\"\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003ePanel C. Moderators (social context)\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eConstruct\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eYRBS variable\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eItem label (YRBS)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eOriginal response categories (coded)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003eAnalytic coding (pre-specified)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003eRole in models\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSchool connectedness\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFeel close to people at their school\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1. Strongly agree\u003c/p\u003e \u003cp\u003e2. Agree\u003c/p\u003e \u003cp\u003e3. Not sure\u003c/p\u003e \u003cp\u003e4. Disagree\u003c/p\u003e \u003cp\u003e5. Strongly disagree\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eDichotomized via derived QN103: agree/strongly agree vs other\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eModerator (interaction with Q80)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eParental monitoring\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ104\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eParental monitoring\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1. Never\u003c/p\u003e \u003cp\u003e2. Rarely\u003c/p\u003e \u003cp\u003e3. Sometimes\u003c/p\u003e \u003cp\u003e4. Most of the time\u003c/p\u003e \u003cp\u003e5. Always\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eDichotomized via derived QN104: most of the time/always vs less\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eModerator (interaction with Q80)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabc\" border=\"1\"\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003ePanel D. CDC-derived dichotomies used for primary inference and robustness\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDerived indicator\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eSource variable(s)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eDefinition used in this study\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003ePurpose\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQN80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHigh-frequency social media use: \u0026ldquo;several times/day or more\u0026rdquo; (corresponds to Q80 categories 6\u0026ndash;8) vs less\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSensitivity exposure\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQN84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePoor mental health: \u0026ldquo;most of the time/always\u0026rdquo; (Q84 categories 4\u0026ndash;5) vs other\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePrimary outcome\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQN103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSchool connectedness: agree/strongly agree (Q103 categories 1\u0026ndash;2) vs other\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eModerator\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQN104\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ104\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eParental monitoring: most of the time/always (Q104 categories 4\u0026ndash;5) vs less\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eModerator\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQN86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eUnstable housing (binary indicator in derived set)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSensitivity covariate\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQN26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSad/hopeless for \u0026ge;\u0026thinsp;2 weeks (binary indicator in derived set)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSensitivity outcome\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabd\" border=\"1\"\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003ePanel E. Covariates and survey design variables\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eVariable type\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eYRBS variable\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eLabel\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eCoding / categories\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003eAnalytic use\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSociodemographic covariate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIn what grade are you\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1. 9th\u003c/p\u003e \u003cp\u003e2. 10th\u003c/p\u003e \u003cp\u003e3. 11th\u003c/p\u003e \u003cp\u003e4. 12th\u003c/p\u003e \u003cp\u003e5. other/ungraded\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAdjusted covariate (pre-specified)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSociodemographic covariate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWhat is your sex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1. Female\u003c/p\u003e \u003cp\u003e2. Male\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAdjusted covariate (pre-specified)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSociodemographic covariate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRACEETH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRace/Ethnicity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1. AI/AN\u003c/p\u003e \u003cp\u003e2. Asian\u003c/p\u003e \u003cp\u003e3. Black\u003c/p\u003e \u003cp\u003e4. NH/PI\u003c/p\u003e \u003cp\u003e5. White\u003c/p\u003e \u003cp\u003e6. Hispanic/Latino\u003c/p\u003e \u003cp\u003e7. Multiple\u0026ndash;Hispanic\u003c/p\u003e \u003cp\u003e8. Multiple\u0026ndash;Non-Hispanic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAdjusted covariate (pre-specified)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStructural context (optional)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQ86 / QN86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eUnstable housing\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eQ86 has 7 residence categories; QN86 is derived binary\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSensitivity covariate\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSurvey design\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWEIGHT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOverall analysis weight\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eContinuous\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRequired for national estimates\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSurvey design\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSTRATUM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSampling strata\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCategorical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRequired for correct SEs\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSurvey design\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePSU\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePrimary sampling unit\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCategorical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRequired for correct SEs\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003e\u003cem\u003eNote. The table reports item labels and response categories as specified in the YRBS 2023 SPSS setup syntax. Primary models use Q80 as an ordered exposure and QN84 as the primary outcome. CDC-derived dichotomies (QN variables) are used for moderation and sensitivity analyses to align with established YRBS reporting conventions.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e2.3.1. Exposure: social media use frequency\u003c/h2\u003e \u003cp\u003eSocial media use frequency was assessed using Q80, an eight-category ordered item ranging from non-use to more than once an hour (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Panel A). The primary analytic specification treated Q80 as an ordinal exposure to preserve graded differences in use frequency rather than imposing an a priori threshold.\u003c/p\u003e \u003cp\u003eTreating Q80 as ordinal assumes a monotonic association between increasing use frequency and poor mental health. To evaluate this assumption, we estimated fully categorical models and visually inspected category-specific adjusted predicted probabilities. These assessments did not indicate meaningful departures from monotonicity, and substantive conclusions were unchanged across ordinal and categorical specifications.\u003c/p\u003e \u003cp\u003eIn pre-specified sensitivity analyses, we used the CDC-derived dichotomous indicator QN80, classifying respondents who reported using social media several times per day or more frequently (Q80 categories 6\u0026ndash;8) versus all others (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Panel D), to assess robustness to exposure specification.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e \u003ch2\u003e2.3.2. Primary outcome: poor mental health\u003c/h2\u003e \u003cp\u003eThe primary outcome was poor mental health, measured using QN84, which identifies respondents reporting that their mental health was not good most of the time or always during the past 30 days (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Panel B/D). This measure captures recent self-reported psychological distress and is interpreted as an indicator of well-being rather than a clinical diagnosis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e2.3.3. Moderators: social context\u003c/h2\u003e \u003cp\u003eTwo indicators of adolescents\u0026rsquo; social context were examined as moderators (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Panel C/D):\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eSchool connectedness (QN103)\u003c/b\u003e: respondents who agreed or strongly agreed that they felt close to people at their school versus other responses.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eParental monitoring (QN104)\u003c/b\u003e: respondents who reported that parents or other adults in their family knew where they were going or with whom most of the time or always, versus less frequent knowledge.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eBoth moderators were operationalized as dichotomous variables to align with CDC reporting conventions and to prioritize interpretability of interaction effects, rather than to impose substantive thresholds. These measures were examined as conditioning factors consistent with ecological and stress-buffering perspectives, while remaining agnostic regarding causal mechanisms due to the cross-sectional design.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e2.3.4. Covariates\u003c/h2\u003e \u003cp\u003eAll primary models adjusted for a minimal, pre-specified set of sociodemographic covariates: grade, sex, and race/ethnicity (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Panel E). This strategy was adopted to reduce the risk of overadjustment for variables plausibly downstream of social media use, such as sleep disruption or peer conflict.\u003c/p\u003e \u003cp\u003eIn sensitivity analyses, additional models included unstable housing (QN86) and bullying victimization to assess robustness to broader structural and psychosocial stressors. Inclusion of these covariates did not materially alter effect estimates or substantive conclusions.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Statistical analysis\u003c/h2\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e2.4.1. Survey design handling\u003c/h2\u003e \u003cp\u003eAll descriptive estimates and regression models incorporated YRBS sampling weights, strata, and PSUs using IBM SPSS Statistics Complex Samples procedures. Survey-weighted estimates and standard errors were computed using Taylor series linearization.\u003c/p\u003e \u003cp\u003eAdjusted predicted probabilities were computed from fitted survey-weighted logistic regression models using marginal standardization. Ninety-five percent confidence intervals were derived using design-based variance estimation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e2.4.2. Primary model set\u003c/h2\u003e \u003cp\u003eWe estimated a series of survey-weighted logistic regression models with poor mental health (QN84) as the outcome:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eModel 1 (primary association)\u003c/b\u003e: QN84 as a function of Q80 and covariates.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eModel 2 (school connectedness moderation)\u003c/b\u003e: Model 1 plus QN103 and an interaction term between Q80 and QN103.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eModel 3 (parental monitoring moderation)\u003c/b\u003e: Model 1 plus QN104 and an interaction term between Q80 and QN104.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eInteraction terms were specified a priori to test whether the association between social media use frequency and poor mental health varied across social contexts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e2.4.3. Interaction inference and reporting strategy\u003c/h2\u003e \u003cp\u003eInteraction terms were evaluated using design-based Wald tests. In both moderation models, interactions reflected differences in the slope of the association between social media use frequency and poor mental health, rather than parallel shifts in baseline risk.\u003c/p\u003e \u003cp\u003ePrimary inference emphasized adjusted predicted probabilities of poor mental health across Q80 categories, stratified by moderator status, with 95% confidence intervals. Odds ratios were reported to facilitate comparison with prior literature. Interaction results were summarized using predicted-probability contrasts and marginal effects.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Robustness analyses\u003c/h2\u003e \u003cp\u003eA limited set of pre-specified robustness checks evaluated the stability of conclusions:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eExposure coding\u003c/b\u003e: ordinal Q80 versus binary QN80.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eOutcome substitution\u003c/b\u003e: persistent sadness or hopelessness (Q26/QN26).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eNon-linearity\u003c/b\u003e: comparison of ordinal and categorical exposure specifications.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eStratified sensitivity\u003c/b\u003e: models stratified by sex and grade when survey-weighted denominators supported stable inference.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eNo analyses beyond these checks were treated as confirmatory.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e2.6. Ethics\u003c/h2\u003e \u003cp\u003eThe YRBS is a publicly available, de-identified dataset. This study involved secondary analysis of existing data and did not require additional institutional review board approval.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Sample characteristics (weighted)\u003c/h2\u003e \u003cp\u003eWeighted demographic and psychosocial characteristics are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Grade was evenly distributed (9th: 26.4%; 10th: 25.8%; 11th: 24.2%; 12th: 23.3%), and the sample was balanced by sex (48.1% female; 51.9% male). By race/ethnicity, the largest group was White (non-Hispanic) (48.1%), followed by Black (non-Hispanic) (13.3%), Asian (4.3%), and other categories; the estimate corresponding to the confidence interval previously reported for Hispanic/Latino was 19.8% (95% CI: 16.2\u0026ndash;23.9), correcting the earlier transcription error where the point estimate did not fall within its interval.\u003c/p\u003e \u003cp\u003eHigh-frequency social media use (QN80) was common (77.0%). School connectedness (QN103) was reported by 55.3% of students, and high parental monitoring (QN104) by 84.0%. The weighted prevalence of poor mental health (QN84) was 28.5% (95% CI: 26.7\u0026ndash;30.4).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eWeighted demographic and psychosocial characteristics, YRBS 2023 (national)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWeighted %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e95% CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eUnweighted n\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eGrade (Q3)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e19,910\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9th\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e26.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[24.3, 28.6]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5,680\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10th\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e25.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[24.0, 27.6]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5,410\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e11th\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e24.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[22.3, 26.3]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4,811\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e12th\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e23.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[21.1, 25.7]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3,961\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSex (Q2)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e19,945\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e48.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[46.0, 50.3]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9,884\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e51.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[49.7, 54.0]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10,061\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRace/Ethnicity (RACEETH)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e19,733\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWhite (non-Hispanic)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e48.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[41.5, 54.9]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9,700\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBlack (non-Hispanic)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[9.2, 18.9]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1,791\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHispanic/Latino\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e19.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[16.2, 23.9]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2,786\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAsian\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[2.9, 6.3]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e995\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOther / Multiple (combined)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4,461\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHigh-frequency social media use (QN80)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e77.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[73.5, 80.1]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e15,203\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSchool connectedness (QN103)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e55.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[52.8, 57.8]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e11,177\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eParental monitoring (QN104)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e84.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[81.2, 86.5]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10,850\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePoor mental health (QN84)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e28.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e[26.7, 30.4]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e15,705\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cem\u003eNote. Percentages are survey-weighted. Confidence intervals use the YRBS complex design. Unweighted n reflects the number of respondents with non-missing data for each characteristic (available-case denominators), which can differ across rows. The regression models below use the pre-specified complete-case analytic sample (N\u0026thinsp;=\u0026thinsp;10,340).\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Primary association between social media use and poor mental health\u003c/h2\u003e \u003cp\u003eIn survey-weighted models adjusted for grade, sex, and race/ethnicity, higher social media use frequency (Q80) was associated with a higher adjusted probability of poor mental health (QN84). Predicted probabilities increased across ordered categories of use, supporting the ordinal specification used in primary models. Because odds ratios are less intuitive for public health interpretation, results are presented primarily as adjusted predicted probabilities.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e3.3. Moderation by school connectedness\u003c/h2\u003e \u003cp\u003eSchool connectedness (QN103) was evaluated as an effect modifier using a survey-weighted interaction model and summarized with adjusted predicted probabilities across Q80 categories (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Across all levels of social media use, adolescents reporting higher school connectedness had a lower adjusted probability of poor mental health than those reporting lower connectedness.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs social media use increased, the adjusted probability of poor mental health rose in both groups, but the increase was steeper among adolescents reporting lower school connectedness. At the highest social media use category, the gap between groups was approximately 15 percentage points. The interaction was statistically significant by design-based Wald test (p \u0026lt; .01), indicating a difference in the exposure\u0026ndash;outcome gradient by connectedness level.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e3.4. Moderation by parental monitoring\u003c/h2\u003e \u003cp\u003eParental monitoring (QN104) was examined in parallel survey-weighted interaction models (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). At low levels of social media use, adolescents reporting higher parental monitoring had a lower adjusted probability of poor mental health than those reporting lower monitoring.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWith increasing social media use, predicted probabilities rose in both groups, but the rise was larger among adolescents reporting lower parental monitoring. At the highest use category, the difference between monitoring groups was again approximately 15 percentage points. The interaction was statistically significant by design-based Wald test (p \u0026lt; .01), consistent with effect modification by family context reflected in differences in the gradient across use categories.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e3.5. Robustness and sensitivity analyses\u003c/h2\u003e \u003cp\u003eRobustness analyses substituting persistent sadness or hopelessness (Q26/QN26) for the primary outcome yielded qualitatively similar patterns to those observed for poor mental health (QN84). Descriptive, survey-weighted prevalence estimates by grade and sex are reported in Supplementary Table S1 using the available Q26 sample. These analyses were conducted to provide contextual comparison only and did not materially alter, nor were they used to draw, the primary inferences based on QN84.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Principal findings\u003c/h2\u003e \u003cp\u003eIn this nationally representative sample of U.S. adolescents, higher social media use frequency was associated with a higher prevalence of poor mental health. Importantly, this association was not uniform across adolescents. The strength of the association differed by social context: adolescents reporting higher school connectedness and greater parental monitoring exhibited consistently lower predicted probabilities of poor mental health across all levels of social media use, whereas those with lower connectedness or monitoring showed steeper exposure\u0026ndash;outcome gradients. These findings indicate that the association between social media use frequency and adolescent mental health varies systematically by social context rather than operating as a uniform risk across the population.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Interpretation and relation to existing literature\u003c/h2\u003e \u003cp\u003eThe present findings align with a growing body of work emphasizing heterogeneity in associations between digital media use and adolescent well-being [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Rather than supporting a simple harmful-use narrative, the results suggest that broader social environments shape how adolescents experience and are associated with frequent social media engagement. This interpretation is consistent with ecological models of development, which posit that individual behaviors and outcomes are embedded within interacting social contexts such as schools and families [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eSchool connectedness and parental monitoring can be understood as indicators of relational support and social integration. Prior research has consistently shown that school connectedness is associated with lower emotional distress and more favorable mental health outcomes [\u003cspan additionalcitationids=\"CR17 CR18 CR19\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e], while parental monitoring reflects supportive family processes linked to adolescent adjustment [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Within stress-related frameworks, these forms of social support may correspond to a flatter increase in mental health risk across increasing levels of online exposure, including experiences such as social comparison, peer conflict, or cyberbullying [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eSeveral complementary mechanisms may underlie these patterns. Supportive school and family environments may facilitate emotional regulation, enabling adolescents to manage negative affect arising from online interactions. They may also contribute to norm-setting around time use, online conduct, and responses to peer feedback. In addition, such environments may support interpretive buffering, whereby adolescents embedded in supportive relationships are less likely to internalize negative online experiences or social comparisons as personally salient. Together, these mechanisms provide a theoretically grounded explanation for why similar levels of social media exposure may be associated with different mental health outcomes across adolescents.\u003c/p\u003e \u003cp\u003eThese findings also help clarify why prior empirical evidence on social media use and mental health has been inconsistent. Reviews routinely note that average associations often mask substantial between-person variation [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. By directly testing moderation in a large, population-based dataset, the present study provides evidence that social context is one dimension along which this variation occurs. This context-dependent pattern is consistent with recent calls to move beyond global effect estimates toward analytic approaches that explicitly examine when and for whom digital media use is associated with poorer mental health [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eImportantly, the results distinguish between two related but conceptually distinct phenomena. Differences in baseline levels of poor mental health across social contexts reflect risk stratification, whereas differences in the strength of the association between social media use frequency and poor mental health reflect effect heterogeneity. The present analyses provide evidence of both: adolescents embedded in more supportive contexts exhibit lower baseline risk, and the increase in predicted probability across social media use categories is flatter in these groups. Recognizing the coexistence of these patterns is essential for interpreting heterogeneous associations in population-based studies and for avoiding oversimplified conclusions about digital media use.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Strengths\u003c/h2\u003e \u003cp\u003eThis study has several strengths. First, it draws on a large, nationally representative sample of U.S. adolescents, supporting generalizability to the national population. Second, all analyses incorporated the complex survey design of the YRBS, enabling valid population-level inference. Third, the analytic strategy was pre-specified, with a focused set of outcomes, moderators, and covariates, reducing the risk of selective reporting. Finally, moderation results were presented using adjusted predicted probabilities rather than relying solely on odds ratios, enhancing interpretability for applied and public health audiences.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section2\"\u003e \u003ch2\u003e4.4. Limitations\u003c/h2\u003e \u003cp\u003eSeveral limitations should be considered. The cross-sectional design precludes conclusions about temporal ordering, and reverse causation remains possible, such that adolescents experiencing poorer mental health may engage more frequently with social media. Selection into supportive school or family contexts may also contribute to observed moderation patterns; adolescents with better baseline mental health may be more likely to report higher connectedness or monitoring.\u003c/p\u003e \u003cp\u003eAll measures relied on self-report and may be subject to recall or reporting differences. Adolescents experiencing poorer mental health may perceive or report lower levels of social support independent of objective differences. Social media use was measured as frequency rather than duration, content, or platform-specific engagement, and the data do not capture qualitative differences between positive and negative online experiences. Finally, although school connectedness and parental monitoring capture important aspects of social context, they do not directly measure peer relationship quality or family dynamics.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec27\" class=\"Section2\"\u003e \u003ch2\u003e4.5. Implications\u003c/h2\u003e \u003cp\u003eThese findings underscore the importance of considering social context when interpreting associations between social media use frequency and adolescent mental health. Rather than framing frequent social media use as inherently harmful, the results indicate that associations differ meaningfully across social environments. At the highest levels of use, differences in predicted probability of poor mental health on the order of approximately 15 percentage points were observed between adolescents reporting higher versus lower levels of school connectedness or parental monitoring, highlighting the population-level relevance of contextual variation.\u003c/p\u003e \u003cp\u003eAt the same time, these findings do not imply that increased social media use is benign or that supportive contexts eliminate risk. Instead, they suggest that exposure operates differently across adolescents depending on broader social environments. This perspective aligns with broader evidence emphasizing that adolescent well-being reflects the interplay of individual behaviors and relational contexts [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. While the present study does not support specific policy or clinical interventions, it reinforces the value of contextualized approaches to understanding adolescents\u0026rsquo; digital lives and cautions against one-size-fits-all interpretations of social media effects.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eIn a nationally representative sample of U.S. adolescents, higher social media use frequency was associated with a higher prevalence of poor mental health, with substantial variation by school connectedness and parental monitoring. These findings indicate that associations between social media use and adolescent mental health are context-dependent rather than uniform across the population. Adolescents embedded in more supportive school and family environments exhibited both lower baseline risk and a flatter increase in predicted probability across increasing levels of social media use.\u003c/p\u003e \u003cp\u003eThese results support the conclusion that social media use frequency is associated with adolescent mental health in ways that depend on broader social environments. At the same time, the cross-sectional design precludes conclusions about directionality, causality, or underlying mechanisms. Together, the findings underscore the importance of situating adolescents\u0026rsquo; digital behaviors within the social contexts in which they occur and caution against one-size-fits-all interpretations of social media effects in population-based research.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eConflict of interest\u003c/p\u003e\n\u003cp\u003eThe author declares no conflicts of interest.\u003c/p\u003e\n\u003cp\u003eEthical approval\u003c/p\u003e\n\u003cp\u003eEthical approval was not required for this study, as it involved secondary analysis of publicly available, de-identified data from the Youth Risk Behavior Surveillance System. The dataset contains no personally identifiable information and involves no direct interaction with human participants.\u003c/p\u003e\n\u003cp\u003eFunding statement\u003c/p\u003e\n\u003cp\u003eThe author received no external funding for the preparation of this manuscript.\u003c/p\u003e\n\u003cp\u003eAcknowledgments\u003c/p\u003e\n\u003cp\u003eNone.\u003c/p\u003e\n\u003cp\u003eData availability\u003c/p\u003e\n\u003cp\u003eThe data analyzed in this study are publicly available from the Centers for Disease Control and Prevention through the Youth Risk Behavior Surveillance System (YRBS). The specific dataset used was the 2023 National YRBS, accessible at https://www.cdc.gov/yrbs/data/index.html. All analyses were conducted using IBM SPSS Statistics (Complex Samples module) to account for the survey\u0026rsquo;s stratified cluster design. The analytic syntax and code used for data preparation and statistical modeling are available from the author upon reasonable request.\u003c/p\u003e\n\u003cp\u003eORCID iD\u003c/p\u003e\n\u003cp\u003eNikesh Lagun: https://orcid.org/0009-0005-6372-4852\u003c/p\u003e\n\u003cp\u003eCRediT author statement\u003c/p\u003e\n\u003cp\u003eNikesh Lagun: Conceptualization; Methodology; Formal analysis; Data curation; Software; Validation; Investigation; Visualization; Writing \u0026ndash; Original Draft; Writing \u0026ndash; Review \u0026amp; Editing; Project administration.\u003c/p\u003e\n\u003cp\u003eDeclaration of generative AI and AI-assisted technologies in the manuscript preparation process\u003c/p\u003e\n\u003cp\u003eDuring the preparation of this manuscript, the authors used generative AI tools solely for language editing and improvement of clarity and writing quality. Following the use of these tools, the authors carefully reviewed, edited, and revised all content and take full responsibility for the accuracy, integrity, and originality of the published article.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eOrben A, Przybylski AK (2019) The association between adolescent well-being and digital technology use. Nat Hum Behav 3(2):173\u0026ndash;182. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41562-018-0506-1\u003c/span\u003e\u003cspan address=\"10.1038/s41562-018-0506-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeles B, McCrae N, Grealish A (2020) A systematic review: the influence of social media on depression, anxiety and psychological distress in adolescents. Int J Adolesc Youth 25(1):79\u0026ndash;93. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/02673843.2019.1590851\u003c/span\u003e\u003cspan address=\"10.1080/02673843.2019.1590851\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOdgers CL, Jensen MR (2020) Annual research review: adolescent mental health in the digital age: facts, fears, and future directions. J Child Psychol Psychiatry 61(3):336\u0026ndash;348. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jcpp.13190\u003c/span\u003e\u003cspan address=\"10.1111/jcpp.13190\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShannon H, Bush K, Villeneuve PJ, Hellemans KG, Guimond S (2022) Problematic social media use in adolescents and young adults: systematic review and meta-analysis, JMIR Ment. Health 9(4):e33450. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2196/33450\u003c/span\u003e\u003cspan address=\"10.2196/33450\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFassi L, Thomas K, Parry DA, Leyland-Craggs A, Ford TJ, Orben A (2024) Social media use and internalizing symptoms in clinical and community adolescent samples: a systematic review and meta-analysis. JAMA Pediatr 178(8):814\u0026ndash;822. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1001/jamapediatrics.2024.2078\u003c/span\u003e\u003cspan address=\"10.1001/jamapediatrics.2024.2078\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrinstein MJ, Nesi J, Telzer EH, Commentary: an updated agenda for the study of digital media use and adolescent development\u0026mdash;future directions following Odgers, Jensen (2020) J. Child Psychol. Psychiatry 61 (3) (2020) 349\u0026ndash;352. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jcpp.13219\u003c/span\u003e\u003cspan address=\"10.1111/jcpp.13219\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eValkenburg P, Beyens I, Keijsers L (2024) Investigating heterogeneity in (social) media effects: experience-based recommendations. Meta-Psychology 8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.15626/MP.2022.3649\u003c/span\u003e\u003cspan address=\"10.15626/MP.2022.3649\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTwenge JM, Joiner TE, Rogers ML, Martin GN (2018) Increases in depressive symptoms, suicide-related outcomes, and suicide rates among U.S. adolescents after 2010 and links to increased new media screen time. Clin Psychol Sci 6(1):3\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/2167702617723376\u003c/span\u003e\u003cspan address=\"10.1177/2167702617723376\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrben A, Przybylski AK (2019) Screens, teens, and psychological well-being: evidence from three time-use-diary studies. Psychol Sci 30(5):682\u0026ndash;696. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/0956797619830329\u003c/span\u003e\u003cspan address=\"10.1177/0956797619830329\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCoyne SM, Rogers AA, Zurcher JD, Stockdale L, Booth M (2020) Does time spent using social media impact mental health? an eight year longitudinal study. Comput Hum Behav 104:106160. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.chb.2019.106160\u003c/span\u003e\u003cspan address=\"10.1016/j.chb.2019.106160\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTwenge JM, Joiner TE, Martin G, Rogers ML (2018) Amount of time online is problematic if it displaces face-to-face social interaction and sleep. Clin Psychol Sci 6(4):456\u0026ndash;457. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/2167702618778562\u003c/span\u003e\u003cspan address=\"10.1177/2167702618778562\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeinstein E (2018) The social media see-saw: positive and negative influences on adolescents\u0026rsquo; affective well-being. New Media Soc 20(10):3597\u0026ndash;3623. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/1461444818755634\u003c/span\u003e\u003cspan address=\"10.1177/1461444818755634\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSteinberg L (2008) A social neuroscience perspective on adolescent risk-taking. Dev Rev 28(1):78\u0026ndash;106. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.dr.2007.08.002\u003c/span\u003e\u003cspan address=\"10.1016/j.dr.2007.08.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlakemore SJ, Mills KL (2014) Is adolescence a sensitive period for sociocultural processing? Annu Rev Psychol 65:187\u0026ndash;207. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1146/annurev-psych-010213-115202\u003c/span\u003e\u003cspan address=\"10.1146/annurev-psych-010213-115202\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBronfenbrenner U (1979) The Ecology of Human Development: Experiments by Nature and Design. Harvard University Press, Cambridge, MA. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2307/j.ctv26071r6\u003c/span\u003e\u003cspan address=\"10.2307/j.ctv26071r6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBonny AE, Britto MT, Klostermann BK, Hornung RW, Slap GB (2000) School disconnectedness: identifying adolescents at risk. Pediatrics 106(5):1017\u0026ndash;1021. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1542/peds.106.5.1017\u003c/span\u003e\u003cspan address=\"10.1542/peds.106.5.1017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcNeely CA, Nonnemaker JM, Blum RW (2002) Promoting school connectedness: evidence from the National Longitudinal Study of Adolescent Health. J School Health 72(4):138\u0026ndash;146. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1746-1561.2002.tb06533.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1746-1561.2002.tb06533.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCatalano RF, Haggerty KP, Oesterle S, Fleming CB, Hawkins JD (2004) The importance of bonding to school for healthy development: findings from the Social Development Research Group. J School Health 74(7):252\u0026ndash;261. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1746-1561.2004.tb08281.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1746-1561.2004.tb08281.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBond L, Butler H, Thomas L, Carlin J, Glover S, Bowes G, Patton G (2007) Social and school connectedness in early secondary school as predictors of late teenage substance use, mental health, and academic outcomes. J Adolesc Health 40(4):357e9\u0026ndash;357e18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jadohealth.2006.10.013\u003c/span\u003e\u003cspan address=\"10.1016/j.jadohealth.2006.10.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShochet IM, Dadds MR, Ham D, Montague R (2006) School connectedness is an underemphasized parameter in adolescent mental health: results of a community prediction study. J Clin Child Adolesc Psychol 35(2):170\u0026ndash;179. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1207/s15374424jccp3502_1\u003c/span\u003e\u003cspan address=\"10.1207/s15374424jccp3502_1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eResnick MD, Bearman PS, Blum RW, Bauman KE, Harris KM, Jones J, Tabor J, Beuhring T, Sieving RE, Shew M, Ireland M, Bearinger LH, Udry JR (1997) Protecting adolescents from harm: findings from the National Longitudinal Study on Adolescent Health. JAMA 278(10):823\u0026ndash;832. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1001/jama.278.10.823\u003c/span\u003e\u003cspan address=\"10.1001/jama.278.10.823\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePutnam RD (2000) Bowling Alone: The Collapse and Revival of American Community. Touchstone Books/Simon \u0026amp; Schuster, New York. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/358916.361990\u003c/span\u003e\u003cspan address=\"10.1145/358916.361990\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStattin H, Kerr M (2000) Parental monitoring: A reinterpretation. Child Dev 71(4):1072\u0026ndash;1085. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1467-8624.00210\u003c/span\u003e\u003cspan address=\"10.1111/1467-8624.00210\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWaizenhofer RN, Buchanan CM, Jackson-Newsom J (2004) Mothers' and fathers' knowledge of adolescents' daily activities: its sources and its links with adolescent adjustment. J Fam Psychol 18(2):348\u0026ndash;360. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1037/0893-3200.18.2.348\u003c/span\u003e\u003cspan address=\"10.1037/0893-3200.18.2.348\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCohen S, Wills TA (1985) Stress, social support, and the buffering hypothesis. Psychol Bull 98(2):310\u0026ndash;357. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1037/0033-2909.98.2.310\u003c/span\u003e\u003cspan address=\"10.1037/0033-2909.98.2.310\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFoster H, Brooks-Gunn J (2009) Toward a stress process model of children's exposure to physical family and community violence. Clin Child Fam Psychol Rev 12(2):71\u0026ndash;94. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10567-009-0049-0\u003c/span\u003e\u003cspan address=\"10.1007/s10567-009-0049-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCenters for Disease Control and Prevention (2023) Youth Risk Behavior Surveillance System (YRBSS) [dataset]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cdc.gov/yrbs/data/index.html\u003c/span\u003e\u003cspan address=\"https://www.cdc.gov/yrbs/data/index.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, (accessed 28 January 2026)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHinduja S, Patchin JW (2010) Bullying, cyberbullying, and suicide, Arch. Suicide Res 14(3):206\u0026ndash;221. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/13811118.2010.494133\u003c/span\u003e\u003cspan address=\"10.1080/13811118.2010.494133\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatchin JW, Hinduja S (2010) Cyberbullying and self-esteem. J School Health 80(12):614\u0026ndash;624. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1746-1561.2010.00548.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1746-1561.2010.00548.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Adolescents, Social media use, Mental health, School connectedness, Parental monitoring, Youth Risk Behavior Survey, Effect heterogeneity, Survey-weighted analysis","lastPublishedDoi":"10.21203/rs.3.rs-9007537/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9007537/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eTo examine associations between social media use frequency and poor mental health among U.S. adolescents and to assess whether these associations vary by school connectedness and parental monitoring.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe analyzed data from the 2023 National Youth Risk Behavior Survey, a nationally representative survey of U.S. students in grades 9\u0026ndash;12 (n\u0026thinsp;=\u0026thinsp;10,340). Social media use frequency was modeled as an ordered exposure, and poor mental health was defined as reporting mental health \u0026ldquo;not good\u0026rdquo; most of the time or always in the past 30 days. Survey-weighted logistic regression models estimated associations between social media use and poor mental health, adjusting for grade, sex, and race/ethnicity, with effect modification assessed using interaction terms.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eOverall, 28.5% (95% CI: 26.7\u0026ndash;30.4) of adolescents reported poor mental health, and 77.0% reported high-frequency social media use. Higher social media use frequency was associated with higher adjusted odds and predicted probabilities of poor mental health. Associations differed by social context (interaction p\u0026thinsp;\u0026lt;\u0026thinsp;0.01). At the highest levels of social media use, adolescents reporting low school connectedness had approximately 15 percentage points higher predicted probability of poor mental health than those reporting high connectedness, with similar differences observed by parental monitoring.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eAssociations between social media use frequency and adolescent mental health are context-dependent. Supportive school and family environments are associated with lower risk across levels of social media use, highlighting the importance of social context in adolescent mental health prevention research.\u003c/p\u003e","manuscriptTitle":"Social media use frequency and adolescent mental health: context-dependent associations by school connectedness and parental monitoring in the 2023 YRBS","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-03 07:48:14","doi":"10.21203/rs.3.rs-9007537/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"310a3f9c-f6cb-4797-8872-c378427048bd","owner":[],"postedDate":"March 3rd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":63751192,"name":"Epidemiology"}],"tags":[],"updatedAt":"2026-03-03T07:48:14+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-03 07:48:14","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9007537","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9007537","identity":"rs-9007537","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}