The effect of municipality-level social media use on youth mental health

The effect of municipality-level social media use on youth mental health | 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 The effect of municipality-level social media use on youth mental health Olav Bertin Tveit, Guido Biele This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6647544/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 14 Oct, 2025 Read the published version in BMC Public Health → Version 1 posted 12 You are reading this latest preprint version Abstract Background Rising internalizing problems among youth, particularly among females, have raised concerns about potential societal causes. Social media use (SMU) has emerged as a key focus, given its widespread adoption since the early 2010s. While individual-level cross-sectional studies suggest small to moderate correlations between SMU and internalizing problems, a complementary community-based perspective allows for assessing the effects of living in environments with high or low social media use on youth mental health. Methods This study investigates the effect of SMU on internalizing symptoms among Norwegian youth in a longitudinal study at the municipality level. The study uses data from the nationwide Ungdata surveys (2014–2024), covering 528 cohorts across 181 municipalities. Anxiety and depressive symptoms were assessed using items adapted from the Hopkins Symptom Checklist and the Depressive Mood Inventory, respectively. We applied a Bayesian multilevel model using the R package brms , accounting for time-varying and time-constant confounders. Results An additional hour of average SMU corresponded to a 0.70 [0.26, 1.14] SD increase in anxiety scores for boys but showed no clear association for girls (-0.03 [-0.31, 0.24]). For depressive symptoms, a one-hour increase in average SMU corresponded to an increase of 0.25 [0.03, 0.46] for boys, with no clear effect for girls (0.02 [-0.15, 0.19]). The models accounted for a substantial proportion of variance in T2 outcomes (r² = .6 - .8). Conclusion Assuming that all relevant factors influencing both social media use and youth mental health were accounted for, the findings suggest that living in communities with high social media use may have a small effect on youth mental health symptoms. However, these effects are inconsistent and primarily observed among boys. Figures Figure 1 Figure 2 Introduction The prevalence of internalizing problems among youths has increased in recent decades, particularly among females ( 1 , 2 ). One of the most significant societal changes since the early 2010s has been the widespread adoption of smartphones and social media, with for instance more than a third of Norwegian youth reporting daily social media use (SMU) exceeding three hours ( 3 ). There is considerable debate in public and academic spheres regarding social media’s role in shaping current mental health trends ( 4 – 6 ). Associations and effects diverge considerably, depending on the type of analysis. Cross-sectional meta-analyses generally give a correlation in the r = .05-0.17 range between SMU and depression ( 6 ). Longitudinal studies typically show slightly smaller associations ( 7 ), and a recent meta-analysis of experimental studies concluded that there is no consistent evidence that SMU is harmful to youth ( 4 ). Most research on the link between SMU and youth mental health has focused on individual-level associations or effects. While this approach is essential for identifying dose-response relationships, it may fail to capture determinants of ill-being to which most of the population is exposed ( 8 ). Widespread use of a medium in a group can affect the social fabric, also for non-users ( 9 ). For instance, while youths as a group engage in less face-to-face socialization than in previous decades, social media use is positively correlated with in-person socializing at the individual level ( 10 ). Because SMU is ubiquitous it is an integral part of young people’s social lives. The cost of abstaining from SMU in a group where the majority are online may thus outweigh the benefits, even if some types of SMU are detrimental to mental health. Identifying all effects of social media on youth as a group requires inclusion of both individual-level and aggregate-level ecologic perspectives ( 8 ). Such studies remain relatively rare, despite a few notable examples ( 11 – 13 ). Aggregate-level studies face distinct methodological challenges, particularly when attempting to make inference about individual-level effects from ecologic data ( 14 , 15 ). Nonetheless, analyzing mental health trends at the regional level is valuable for public health policy, as many interventions—such as school-based restrictions or community-wide initiatives—are implemented at this scale ( 16 , 17 ). At the heart of the SMU-mental health debate lies a causal question. Identifying causal effects requires clarity about the direction of effects and assuming no unobserved confounders ( 18 ). This makes causal inference from observational data inherently challenging and a matter of subject matter debate, as reverse causality in cross-sectional studies and control of all common causes cannot be verified empirically ( 18 ). For example, Nilsen and colleagues ( 13 ) showed that rising SMU partially statistically explained the concurrent increase in physical health complaints among Norwegian youth. They used a large repeated cross-sectional municipality-level dataset and sophisticated multilevel models to control for time invariant confounders. However, as they investigated concurrent associations, their analysis left the directionality question unresolved. There is a need for large-scale, causally informed longitudinal studies that approach the issue with several methodologies and analytic perspectives ( 6 , 19 ). This study investigates the longitudinal effects of social media use on symptoms of depression and anxiety among Norwegian youth at the municipality level. Recognizing that unobserved confounding is an important challenge for causal inference from observational data, we chose a study design that implemented a longitudinal design and removed all time-varying unobserved confounding prior to the measurement of SMU by adjusting for the pre-treatment levels of the outcome, and we adjusted for time-constant confounders by estimating municipality level random effects. Methods The data were derived from the Ungdata surveys, a nation-wide study implemented annually at the municipality level for 8 th to 13 th graders. Participation in the Ungdata study was voluntary, and participating students gave their informed consent and could withdraw at any time. Parents were also informed and had the option to withdraw their children from the study. Participants completed an electronic questionnaire in class. The average response rate across municipalities ranged from 78% to 85% (3). Criteria for inclusion in the present study were (1) Ungdata had information about the exposure SMU and the outcomes depression or anxiety at two time point. (2) The number of participants from the cohort at the second measurement did not deviate by more than 15% from the sample size at the first measurement. This study used data from 528 cohorts from 181 municipalities, collected between 2014 and 2024. Measures The Ungdata study uses mental health items adapted from the Hopkins Symptom Checklist (20) and the Depressive Mood Inventory (21). Depressive symptoms during the previous week were assessed with six items (e.g., “Felt unhappy, sad or depressed”). Anxiety during the previous week was assessed with three items (e.g., “Felt constant fear or anxiety”). All items were rated on a 4-point Likert scale (1= Not been affected at all , 4 = Been affected a great deal ). A mean score was used in analyses. Social media use was assessed with the question “Think about what you do on a normal day. How much time do you spend on social media (Facebook, Instagram, etc.)?” with response options 1 = No time, 2 = Less than 30 minutes, 3 = 30 minutes - 1 hour, 4 = 1-2 hours, 5 = 2-3 hours, 6 = more than 3 hours. To facilitate analysis, response categories were recoded to reflect the midpoint of each time interval in hours or minutes (e.g., 1–2 hours was recoded as 1 hour 30 minutes), with the final category (more than 3 hours) recoded as 4 hours. See supplementary materials for a complete list of survey items used in the study. The unit of analysis was cohorts within municipalities that had participated in the survey twice, with intervals of one, two, or three years. From 2020 onward, anxiety items were removed from the core section of the survey, leading to a reduced number of municipalities with post-2020 anxiety data. As a result, the analytic sample included 262 cohorts across 113 municipalities for anxiety, and 528 cohorts across 181 municipalities for depression. See Table 1 for a description of the sample. Table 1. Sample Sizes of Municipalities and Cohorts. Years T1 to T2 School grade T1 N N unique municipalities N students Anxiety Depression Anxiety Depression Anxiety Depression 1 8 11 38 6 24 419 1502 1 9 17 43 10 29 906 1979 1 10 1 7 1 6 68 315 1 11 5 15 4 11 481 953 1 12 2 8 2 8 112 400 2 8 80 116 50 69 6861 8630 2 9 17 33 14 27 844 2672 2 10 15 23 13 20 818 1625 2 11 3 4 3 4 164 754 3 8 41 95 30 66 4847 8964 3 9 43 95 33 65 4600 8648 3 10 27 51 25 41 2122 3572 Table 1: The table shows the sample size of municipalities stratified by school year and time between baseline and follow-up assessments. N indicates the number of cohorts from unique schools. Analyses To estimate the municipality-level effects of social media use on self-reported symptoms of anxiety and depression, we employed a Bayesian multilevel model using the R package brms (22), which utilizes the Stan probabilistic programming language and a Hamiltonian Monte Carlo sampler for model estimation (23). The analysis was designed to approximate a pre-test–post-test design, adjusting for both pre-treatment time-varying and time-constant confounders. To control for potential confounding, we adjusted for key pre-test covariates, including baseline symptoms, family socioeconomic status, sex, school year, and the time interval between baseline and follow-up. Random effects were specified at multiple levels to account for clustering and time-constant confounding (e.g., urbanicity) at the municipality level (24). Specifically, we included municipality-level random effects to adjust for time-invariant contextual differences, and nested random effects by school year, time interval, and birth year to account for repeated observations within municipalities and variations across cohorts. Exposure random slopes were employed to allow the effects of SMU to vary across key covariates. To minimize parametric assumptions about the functional form of the exposure-outcome association, we used penalized spline functions to model continuous covariates, including socioeconomic status and social media use. Additionally, we used data from Statistics Norway to adjust for urbanicity of municipalities. We adjusted for potential bias from random effects components by including municipality-level averages of exposure at baseline (25). Given the potential impact of COVID-19-related disruptions in 2021, we included an indicator variable for this period. Model residual variance was allowed to vary by sex and birth year. To estimate average treatment effects (ATEs) of social media use on mental health outcomes, we applied G-estimation methods (26). This approach ensures a robust estimation of social media’s impact on adolescent mental health while addressing confounding at multiple levels through pre-test adjustment, hierarchical modeling, and flexible functional forms. Results See Fig. 1 for time trends in self-reported anxiety, depressive symptoms, and social media between split by gender. The data show a steady increase in all measures, with a stabilization of depressive symptoms in recent years. Girls consistently reported more symptoms, and more time spent on social media than boys. The estimated standardized effects of SMU at Time 1 on depressive and anxiety symptoms at Time 2 are shown in Fig. 2 . Results are shown separately for boys and girls. One standard deviation increase in social media use was not clearly associated with a change in anxiety scores for girls (-0.01 [-0.13, 0.10]) and was associated with an increase of 0.24 [0.09, 0.38] in anxiety scores for boys. Similarly for depressive symptoms, an increase of one standard deviation of social media use was not clearly associated with a change in depression scores for girls (0.01 [-0.06, 0.08]) and was weakly associated with an increase of 0.10 [0.01, 0.18] in depression scores for boys. Table 2 presents standardized and unstandardized effect estimates. The models explained a substantial amount of variation in the T2 outcomes, with r 2 = .6 − .8. Table 2 Unstandardized and Standardized Results Gender Standardized Anxiety estimate (95% CI) Depression estimate (95% CI) Boys - 0.09 (0.03, 0.15) 0.04 (0.01, 0.07) Girls - 0.00 (-0.05, 0.04) 0.00 (-0.03, 0.03) Boys xy 0.24 (0.09, 0.38) 0.10 (0.01, 0.18) Girls xy -0.01 (-0.13, 0.10) 0.01 (-0.06, 0.08) Boys y 0.70 (0.26, 1.14) 0.25 (0.03, 0.46) Girls y -0.03 (-0.31, 0.24) 0.02 (-0.15, 0.19) Table 2 : The table shows the estimated effect of social media use on anxious and depressive symptoms for boys and girls, respectively. Xy standardization refers to the expected increase in the outcome in standard deviations from a one standard deviation increase in social media use. Y standardization refers to the expected increase in standard deviations of the outcome from a one unit increase in social media use with respect to the survey scale. Discussion This large-scale study of Norwegian adolescents’ internalizing problems found no consistent effect of social media use at the municipality level. SMU had a small effect on subsequent anxiety and depression in boys, but not girls. Our findings largely parallel those from individual-level studies showing that SMU and mental health complaints are generally correlated within cohorts ( 6 ), but where a recent meta-analysis of experimental studies found no consistent evidence of a causal relationship ( 4 ). Confronted with deteriorating mental health among youths in large parts of the world, it is imperative that researchers make use of available large-scale data to identify its causes ( 1 , 2 ). Any claim to estimate causal effects rests on the assumption of no unmeasured confounding ( 18 ). We accounted for all time-varying unobserved confounding prior to the assessment of SMU by adjusting for the pre-treatment levels of internalizing problems, and we adjusted for time-constant confounders through a mixture of fixed and random effects ( 25 ). The large proportion of variance in internalizing symptoms explained by our models supports our assumption that we have not left out key explanatory variables. Still, a crucial insight from the causal inference literature is that sufficient confounder adjustment cannot be verified empirically but relies instead on subject matter knowledge and debate ( 18 ). One potential source of bias unique to aggregate-level research is migration between municipalities across assessment waves ( 14 ), but we know of no evidence of substantial migration patterns in Norway during the study period. Taken together, we the combination of a solid study design that allows adjusting for many unobserved confounders and the finding that the analysis model allows to explain a large portion of the variance make a causal interpretation of our results plausible. We found a small effect of SMU on anxiety and depression for boys, but we are hesitant to overinterpret this on two accounts. Firstly, the standardized effect size was quite small. Researchers in the social media field disagree widely over what to consider a meaningful effect ( 6 ). Our model suggests that a mean increase of 1 hour of SMU corresponds to a 0.09-point increase in anxiety symptoms and a 0.04-point increase in depressive symptoms on a scale from 1 to 4. We would argue that the effect on depression is negligible while the effect on anxiety is potentially meaningful if replicated. Secondly, the gendered pattern is opposite from what is commonly found, namely that there is a stronger association between SMU and internalizing problems in girls ( 12 , 13 , 27 ). Girls typically spend more time on social media and report more internalizing symptoms than their male peers ( 28 ), also in our data. It is not readily apparent why boys should be more susceptible to plausible group or individual-level mechanisms for mental health effects of SMU, such as social comparisons or contagion of negative emotions ( 28 – 30 ). For these reasons, Keyes & Platt ( 28 ) have argued that any plausible cause for the increase in youth internalizing problems should display a gendered effect which is stronger for girls than for boys. We are thus inclined to interpret the overall result of our study as an inconsistent effect of SMU on youth internalizing symptoms at the municipality level. A key contribution of our study is clearly estimating the direction of effects from social media to mental health, which has been a large caveat with much previous cross-sectional work ( 31 ). This also expands on more recent work by Nilsen and colleagues ( 13 ), who found that increases in a municipality’s average social media use (SMU) were associated with concurrent increases in physical health complaints, particularly among girls. To the extent that physical health complaints and internalizing symptoms are related ( 32 ), our contrasting findings could suggest that the concurrent associations reflect broader societal changes or reverse causality. The aggregated level of analysis in this study has both strengths and limitations. It is well suited to identify indirect network or cultural effects of living in a peer group with high social media use and can provide insight into the potential impact of municipality-wide initiatives aimed at reducing access to social media. However, drawing conclusions about individual-level effects from aggregated data requires stronger assumptions than we are willing to make ( 15 ). It remains possible that social media use affects youth mental health at the individual level within municipalities. For instance, if social media’s impact is highly non-linear—primarily affecting extreme users, as some studies suggest—individual-level effects may be obscured in aggregate data ( 33 – 35 ). It is also possible that SMU disproportionately affects more vulnerable youth ( 36 ), which at the level of municipalities can be balanced out by youth who gain something from social media ( 37 ). Therefore, our findings do not justify a broad recommendation for parents or professionals to disregard concerns about individual youth’s social media use. The results are congruent with a perspective that social media has both negative and positive effects on individuals, which may balance each other out at the population level ( 6 , 31 , 37 ). It is important to recognize that our analysis does not preclude other worrying group-level effects of SMU, such as polarization of gender roles or political opinions ( 38 ). It is also possible that the mere presence of social media in a community is enough to have downstream effects on mental health beyond the average time spent on social media, which could explain the discrepancy between our findings and the detrimental effects found from the introduction of Facebook to US colleges ( 12 ). However, the decline in Norwegian youth mental health began prior to the widespread adoption of smartphones and social media, indicating that these factors cannot be the sole drivers of the trend ( 1 ). One implication of our findings is that municipality-wide interventions to reduce the overall SMU among youth may not improve the population’s internalizing problems. This is in accordance with the mixed findings on the effects of school smartphone bans ( 16 , 17 , 39 ). One recent UK study showed no differences in mental well-being among pupils in schools with restrictive or permissive phone policies ( 17 ). However, the same study also found no differences in the overall time spent on social media between groups, suggesting that school phone bans in their current form are ineffective at reducing adolescent SMU ( 17 ). Interventions targeting vulnerable youth with heavy social media use have shown more promising effects ( 40 ), congruent with a perspective that SMU can be harmful to some even if no effect is found on the population level. Limitations The study’s strengths include a large, nation-wide dataset of Norwegian adolescents and a causally informed analysis. However, besides the difficulty in translating our results to individual-level effects, some limitations should be noted. We assessed the average time youth reported spending on social media, with the upper category being 3 hours or more. It is possible that this is too narrow a range, as it hinders us from differentiating youth using social media 3.5 hours a day from 5–6 hours a day ( 41 ). Additionally, more nuanced measures of how youth spend their time on social media (active versus passive use; type of platform) is also important in future research. Conclusions Given the careful attempts to adjust for observed and unobserved confounders, the present study suggests that at the level of municipalities increasing in social media use is not a key driver of increases in youth’s depression and anxiety. The study is congruent with a perspective that social media may have both detrimental and positive effects on youth mental health which outweigh each other at the population level. Declarations Ethics approval and consent to participate The study adhered to the guidelines of the Helsinki Declaration. The study presents analyses of previously collected data. Participation was voluntary, and both students and their parents were informed accordingly. All students had to consent to participate in the study and could withdraw at any time. Parents had the option to withdraw their children from the study. For grades 8 to 10 the survey was completely anonymous and did not require approval from data protection authorities. The data protection officer at OsloMet approved data collection from high school students. The National Committee for Research Ethics in the Social Sciences and the Humanities (NESH) approved the use of passive parental consent for the Ungdata study. NESH is an independent ethics committee appointed by the Ministry of Education and Research. Additional details can be found on the project’s official website: https://www.ungdata.no/english/. Consent for publication Not applicable. Availability of data and materials The full Ungdata survey used in the study can be downloaded from https://www.ungdata.no/wp-content/uploads/2020/09/Ungdata-Dokumentasjonsrapport-2010-2019-PDF-1.pdf. The dataset analyzed in the current study can be downloaded from https://surveybanken.sikt.no/en/study/query/ungdata/page/1. Codes for Oslo city districts are not in the file but can be requested from the Ungdata team. Competing interests The authors declare that they have no competing interests. Funding The authors were funded by the Norwegian Ministry of Health and Care Services and by the Norwegian Institute of Public Health. Authors' contributions GB contributed with the initial idea, study design and analyses. OT conducted the literature review and wrote the draft for the paper. Both authors read and approved the final manuscript. Acknowledgements Not applicable. References Potrebny T, Nilsen SA, Bakken A, von Soest T, Kvaløy K, Samdal O, et al. Secular trends in mental health problems among young people in Norway: a review and meta-analysis. European Child & Adolescent Psychiatry. 2024:1-13. McGorry PD, Mei C, Dalal N, Alvarez-Jimenez M, Blakemore S-J, Browne V, et al. The Lancet Psychiatry Commission on youth mental health. The Lancet Psychiatry. 2024;11(9):731-74. Bakken A. Ungdata 2024. Nasjonale resultater. 2024. Ferguson CJ, Kaye LK, Branley-Bell D, Markey P. There is no evidence that time spent on social media is correlated with adolescent mental health problems: Findings from a meta-analysis. Professional Psychology: Research and Practice. 2024. Twenge JM, Haidt J, Joiner TE, Campbell WK. Underestimating digital media harm. Nature Human Behaviour. 2020;4(4):346-8. Valkenburg PM, Meier A, Beyens I. Social media use and its impact on adolescent mental health: An umbrella review of the evidence. Current Opinion in Psychology. 2022;44:58-68. Tang S, Werner-Seidler A, Torok M, Mackinnon AJ, Christensen H. The relationship between screen time and mental health in young people: A systematic review of longitudinal studies. Clinical Psychology Review. 2021;86:102021. Rose G. Sick individuals and sick populations. International Journal of Epidemiology. 2001;30(3):427-32. Nesi J, Choukas-Bradley S, Prinstein MJ. Transformation of adolescent peer relations in the social media context: Part 1—A theoretical framework and application to dyadic peer relationships. Clinical Child and Family Psychology Review. 2018;21:267-94. Twenge JM, Spitzberg BH, Campbell WK. Less in-person social interaction with peers among US adolescents in the 21st century and links to loneliness. Journal of Social and Personal Relationships. 2019;36(6):1892-913. Arenas-Arroyo E, Fernández-Kranz D, Nollenberger N. High speed internet and the widening gender gap in adolescent mental health: Evidence from hospital records. IZA Discussion Papers; 2022. Braghieri L, Levy Re, Makarin A. Social media and mental health. American Economic Review. 2022;112(11):3660-93. Nilsen SA, Stormark KM, Heradstveit O, Breivik K. Trends in physical health complaints among adolescents from 2014–2019: Considering screen time, social media use, and physical activity. SSM-Population Health. 2023;22:101394. Greenland S. Ecologic versus individual-level sources of bias in ecologic estimates of contextual health effects. International Journal of Epidemiology. 2001;30(6):1343-50. Greenland S, Robins J. Invited commentary: ecologic studies—biases, misconceptions, and counterexamples. American Journal of Epidemiology. 1994;139(8):747-60. Beneito P, Vicente-Chirivella Ó. Banning mobile phones in schools: evidence from regional-level policies in Spain. Applied Economic Analysis. 2022;30(90):153-75. Goodyear VA, Randhawa A, Adab P, Al-Janabi H, Fenton S, Jones K, et al. School phone policies and their association with mental wellbeing, phone use, and social media use (SMART Schools): a cross-sectional observational study. The Lancet Regional Health–Europe. 2025. Hernán M, Robins J. Causal Inference: What if. Boca Raton: Chapman & Hall; 2020. Abbasi J. Surgeon general sounds the alarm on social media use and youth mental health crisis. JAMA. 2023;330(1):11-2. Derogatis LR, Lipman RS, Rickels K, Uhlenhuth EH, Covi L. The Hopkins Symptom Checklist (HSCL): A self‐report symptom inventory. Behavioral Science. 1974;19(1):1-15. Kandel DB, Davies M. Epidemiology of depressive mood in adolescents: An empirical study. Archives of General Psychiatry. 1982;39(10):1205-12. Bürkner P-C. brms: An R package for Bayesian multilevel models using Stan. Journal of Statistical Software. 2017;80:1-28. Carpenter B, Gelman A, Hoffman MD, Lee D, Goodrich B, Betancourt M, et al. Stan: A probabilistic programming language. Journal of statistical software. 2017;76:1-32. Gelman A, Hill J. Data analysis using regression and multilevel/hierarchical models: Cambridge university press; 2007. Neuhaus JM, Kalbfleisch JD. Between-and within-cluster covariate effects in the analysis of clustered data. Biometrics. 1998:638-45. Robins J. A new approach to causal inference in mortality studies with a sustained exposure period—application to control of the healthy worker survivor effect. Mathematical Modelling. 1986;7(9-12):1393-512. Twenge JM, Haidt J, Lozano J, Cummins KM. Specification curve analysis shows that social media use is linked to poor mental health, especially among girls. Acta Psychologica. 2022;224:103512. Keyes KM, Platt JM. Annual Research Review: Sex, gender, and internalizing conditions among adolescents in the 21st century–trends, causes, consequences. Journal of Child Psychology and Psychiatry. 2024;65(4):384-407. Orben A, Meier A, Dalgleish T, Blakemore S-J. Mechanisms linking social media use to adolescent mental health vulnerability. Nature Reviews Psychology. 2024:1-17. Kramer AD, Guillory JE, Hancock JT. Experimental evidence of massive-scale emotional contagion through social networks. Proceedings of the National Academy of Sciences. 2014;111(24):8788-90. Orben A. Teenagers, screens and social media: a narrative review of reviews and key studies. Social Psychiatry and Psychiatric Epidemiology. 2020;55(4):407-14. Amendola S, Hengartner MP, Spensieri V, Grillo L, Cerutti R. Patterns of internalizing symptoms and disability functioning in children and adolescents. European Child & Adolescent Psychiatry. 2022;31(9):1455-64. Greenland S, Morgenstern H. Ecological bias, confounding, and effect modification. International Journal of Epidemiology. 1989;18(1):269-74. Przybylski AK, Weinstein N. A large-scale test of the goldilocks hypothesis: quantifying the relations between digital-screen use and the mental well-being of adolescents. Psychological Science. 2017;28(2):204-15. Riehm KE, Feder KA, Tormohlen KN, Crum RM, Young AS, Green KM, et al. Associations between time spent using social media and internalizing and externalizing problems among US youth. JAMA Psychiatry. 2019;76(12):1266-73. Elmer T, Fernández A, Stadel M, Kas MJ, Langener AM. Bidirectional associations between smartphone usage and momentary well-being in young adults: Tackling methodological challenges by combining experience sampling methods with passive smartphone data. Emotion. 2025. Beyens I, Pouwels JL, van Driel II, Keijsers L, Valkenburg PM. Social media use and adolescents’ well-being: Developing a typology of person-specific effect patterns. Communication Research. 2024;51(6):691-716. Kubin E, Von Sikorski C. The role of (social) media in political polarization: a systematic review. Annals of the International Communication Association. 2021;45(3):188-206. Kessel D, Hardardottir HL, Tyrefors B. The impact of banning mobile phones in Swedish secondary schools. Economics of Education Review. 2020;77:102009. Davis CG, Goldfield GS. Limiting social media use decreases depression, anxiety, and fear of missing out in youth with emotional distress: A randomized controlled trial. Psychology of Popular Media. 2024. Kelly Y, Zilanawala A, Booker C, Sacker A. Social media use and adolescent mental health: Findings from the UK Millennium Cohort Study. EClinicalMedicine. 2018;6:59-68. Additional Declarations No competing interests reported. Supplementary Files Surveyitems.docx Cite Share Download PDF Status: Published Journal Publication published 14 Oct, 2025 Read the published version in BMC Public Health → Version 1 posted Editorial decision: Revision requested 18 Jul, 2025 Reviews received at journal 10 Jul, 2025 Reviewers agreed at journal 09 Jul, 2025 Reviews received at journal 08 Jul, 2025 Reviewers agreed at journal 07 Jul, 2025 Reviewers agreed at journal 25 Jun, 2025 Reviewers agreed at journal 24 Jun, 2025 Reviewers invited by journal 04 Jun, 2025 Editor assigned by journal 04 Jun, 2025 Editor invited by journal 26 May, 2025 Submission checks completed at journal 21 May, 2025 First submitted to journal 21 May, 2025 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-6647544","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":467042643,"identity":"95ce53bd-d1b6-496a-9871-7110491ac55a","order_by":0,"name":"Olav Bertin Tveit","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAjElEQVRIiWNgGAWjYDACdiD+YHOAhwQtzMwMjDPSSNXCzJN2gAQd/Mz8x6RtEu7IMEg3E6lPspmZTTon4RkPg8yxBOK0GBxmZjbO/XGYh0Eix4AELRYJJGphfMxAkhagXwwf9gC1sEmkJRCnhZ+98cGBHwmH7fklkg8QpwUO2EhUPwpGwSgYBaMAHwAAChkiCeezCIgAAAAASUVORK5CYII=","orcid":"","institution":"Norwegian Institute of Public Health","correspondingAuthor":true,"prefix":"","firstName":"Olav","middleName":"Bertin","lastName":"Tveit","suffix":""},{"id":467042644,"identity":"d66d9b5b-3dbe-478b-b648-a0b1b533aa80","order_by":1,"name":"Guido Biele","email":"","orcid":"","institution":"Norwegian Institute of Public Health","correspondingAuthor":false,"prefix":"","firstName":"Guido","middleName":"","lastName":"Biele","suffix":""}],"badges":[],"createdAt":"2025-05-12 14:23:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6647544/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6647544/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12889-025-24727-4","type":"published","date":"2025-10-14T15:57:19+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":84094867,"identity":"6fb2377e-4e5a-4d1a-b568-04b47f6e9910","added_by":"auto","created_at":"2025-06-06 17:21:28","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":107640,"visible":true,"origin":"","legend":"\u003cp\u003eTime Trends in Outcomes and Exposure.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eTime trends for scaled sum scores and potential causes split by gender. Values for outcomes are scaled to allow for easy comparisons of the trends. Scaling is performed using data from 2013-2015. For these years, the mean for boys is 1, the standard deviation of sum scores is 1, and the mean score for girls is offset by the raw mean difference divided by the pooled standard deviation. This scaling facilitates a comparison of the time trends of the outcomes while maintaining the difference between boys and girls. Points indicate average values for a year, whereas the size of the points indicates the sample size. Lines are trend-lines estimated from the data.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6647544/v1/90cfe6b0566c4a42a4a5604a.png"},{"id":84095582,"identity":"248e2646-f208-4530-bdf1-4cfd24664996","added_by":"auto","created_at":"2025-06-06 17:29:28","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":39228,"visible":true,"origin":"","legend":"\u003cp\u003eEffect of SMU on Anxiety and Depression.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eEstimated standardized effect of exposures on outcomes. The effects are standardized to enable comparison across different results and indicate how many standard deviations the outcome increases when the exposure increases by one standard deviation. Each column represents specific outcomes for boys (blue) and girls (red), respectively. The height of the bars indicates the strength and direction of the effect estimate, while the vertical lines denote 95% credibility intervals. Effect estimates where the credibility interval includes 0 are shown in a lighter color.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6647544/v1/6be96087db13b5d37c273a97.png"},{"id":93955939,"identity":"860e6897-d802-469f-ba42-40dffa9659d2","added_by":"auto","created_at":"2025-10-20 16:07:53","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":646091,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6647544/v1/42adbe36-567d-4f3d-a690-158c3467d901.pdf"},{"id":84094864,"identity":"c726ca5b-2a7c-4fd6-821f-a0e6cc6e57f8","added_by":"auto","created_at":"2025-06-06 17:21:28","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":18528,"visible":true,"origin":"","legend":"","description":"","filename":"Surveyitems.docx","url":"https://assets-eu.researchsquare.com/files/rs-6647544/v1/38bd76bf6c04a385fc2ef40e.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"The effect of municipality-level social media use on youth mental health ","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe prevalence of internalizing problems among youths has increased in recent decades, particularly among females (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). One of the most significant societal changes since the early 2010s has been the widespread adoption of smartphones and social media, with for instance more than a third of Norwegian youth reporting daily social media use (SMU) exceeding three hours (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). There is considerable debate in public and academic spheres regarding social media\u0026rsquo;s role in shaping current mental health trends (\u003cspan additionalcitationids=\"CR5\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). Associations and effects diverge considerably, depending on the type of analysis. Cross-sectional meta-analyses generally give a correlation in the r\u0026thinsp;=\u0026thinsp;.05-0.17 range between SMU and depression (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). Longitudinal studies typically show slightly smaller associations (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e), and a recent meta-analysis of experimental studies concluded that there is no consistent evidence that SMU is harmful to youth (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eMost research on the link between SMU and youth mental health has focused on individual-level associations or effects. While this approach is essential for identifying dose-response relationships, it may fail to capture determinants of ill-being to which most of the population is exposed (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). Widespread use of a medium in a group can affect the social fabric, also for non-users (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). For instance, while youths as a group engage in less face-to-face socialization than in previous decades, social media use is positively correlated with in-person socializing at the individual level (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). Because SMU is ubiquitous it is an integral part of young people\u0026rsquo;s social lives. The cost of abstaining from SMU in a group where the majority are online may thus outweigh the benefits, even if some types of SMU are detrimental to mental health.\u003c/p\u003e \u003cp\u003eIdentifying all effects of social media on youth as a group requires inclusion of both individual-level and aggregate-level ecologic perspectives (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). Such studies remain relatively rare, despite a few notable examples (\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e). Aggregate-level studies face distinct methodological challenges, particularly when attempting to make inference about individual-level effects from ecologic data (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). Nonetheless, analyzing mental health trends at the regional level is valuable for public health policy, as many interventions\u0026mdash;such as school-based restrictions or community-wide initiatives\u0026mdash;are implemented at this scale (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAt the heart of the SMU-mental health debate lies a causal question. Identifying causal effects requires clarity about the direction of effects and assuming no unobserved confounders (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). This makes causal inference from observational data inherently challenging and a matter of subject matter debate, as reverse causality in cross-sectional studies and control of all common causes cannot be verified empirically (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). For example, Nilsen and colleagues (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e) showed that rising SMU partially statistically explained the concurrent increase in physical health complaints among Norwegian youth. They used a large repeated cross-sectional municipality-level dataset and sophisticated multilevel models to control for time invariant confounders. However, as they investigated concurrent associations, their analysis left the directionality question unresolved. There is a need for large-scale, causally informed longitudinal studies that approach the issue with several methodologies and analytic perspectives (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThis study investigates the longitudinal effects of social media use on symptoms of depression and anxiety among Norwegian youth at the municipality level. Recognizing that unobserved confounding is an important challenge for causal inference from observational data, we chose a study design that implemented a longitudinal design and removed all time-varying unobserved confounding prior to the measurement of SMU by adjusting for the pre-treatment levels of the outcome, and we adjusted for time-constant confounders by estimating municipality level random effects.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eThe data were derived from the Ungdata surveys, a nation-wide study implemented annually at the municipality level for 8\u003csup\u003eth\u003c/sup\u003e to 13\u003csup\u003eth\u003c/sup\u003e graders. Participation in the Ungdata study was voluntary, and participating students gave their informed consent and could withdraw at any time. Parents were also informed and had the option to withdraw their children from the study. Participants completed an electronic questionnaire in class. The average response rate across municipalities ranged from 78% to 85% (3). Criteria for inclusion in the present study were (1) Ungdata had information about the exposure SMU and the outcomes depression or anxiety at two time point. (2) The number of participants from the cohort at the second measurement did not deviate by more than 15% from the sample size at the first measurement. This study used data from 528 cohorts from 181 municipalities, collected between 2014 and 2024.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eMeasures\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe Ungdata study uses mental health items adapted from the Hopkins Symptom Checklist (20) and the Depressive Mood Inventory (21).\u003cem\u003e\u0026nbsp;\u003c/em\u003eDepressive symptoms during the previous week were assessed with six items (e.g., \u0026ldquo;Felt unhappy, sad or depressed\u0026rdquo;). Anxiety during the previous week was assessed with three items (e.g., \u0026ldquo;Felt constant fear or anxiety\u0026rdquo;). All items were rated on a 4-point Likert scale (1= \u003cem\u003eNot been affected at all\u003c/em\u003e, 4 = \u003cem\u003eBeen affected a great deal\u003c/em\u003e). A mean score was used in analyses.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSocial media use was assessed with the question \u0026ldquo;Think about what you do on a normal day. How much time do you spend on social media (Facebook, Instagram, etc.)?\u0026rdquo; with response options 1 = No time, 2 = Less than 30 minutes, 3 = 30 minutes - 1 hour, 4 = 1-2 hours, 5 = 2-3 hours, 6 = more than 3 hours. To facilitate analysis, response categories were recoded to reflect the midpoint of each time interval in hours or minutes (e.g., 1\u0026ndash;2 hours was recoded as 1 hour 30 minutes), with the final category (more than 3 hours) recoded as 4 hours. See supplementary materials for a complete list of survey items used in the study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe unit of analysis was cohorts within municipalities that had participated in the survey twice, with intervals of one, two, or three years. From 2020 onward, anxiety items were removed from the core section of the survey, leading to a reduced number of municipalities with post-2020 anxiety data. As a result, the analytic sample included 262 cohorts across 113 municipalities for anxiety, and 528 cohorts across 181 municipalities for depression. See Table 1 for a description of the sample.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 1. Sample Sizes of Municipalities and Cohorts.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eYears T1 to T2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSchool grade T1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 147px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eN\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 186px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eN\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;unique municipalities\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 153px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eN\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;students\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAnxiety\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDepression\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAnxiety\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDepression\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAnxiety\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDepression\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e38\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e419\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e1502\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e906\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e1979\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e68\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e481\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e953\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e112\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e400\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e116\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e6861\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e8630\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e844\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e2672\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e818\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e1625\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e164\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e754\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e4847\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e8964\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e4600\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e8648\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003e2122\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e3572\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTable 1: The table shows the sample size of municipalities stratified by school year and time between baseline and follow-up assessments. \u003cem\u003eN\u003c/em\u003e indicates the number of cohorts from unique schools. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAnalyses\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo estimate the municipality-level effects of social media use on self-reported symptoms of anxiety and depression, we employed a Bayesian multilevel model using the R package \u003cem\u003ebrms\u003c/em\u003e (22), which utilizes the Stan probabilistic programming language and a Hamiltonian Monte Carlo sampler for model estimation (23). The analysis was designed to approximate a pre-test\u0026ndash;post-test design, adjusting for both pre-treatment time-varying and time-constant confounders.\u003c/p\u003e\n\u003cp\u003eTo control for potential confounding, we adjusted for key pre-test covariates, including baseline symptoms, family socioeconomic status, sex, school year, and the time interval between baseline and follow-up.\u003c/p\u003e\n\u003cp\u003eRandom effects were specified at multiple levels to account for clustering and time-constant confounding (e.g., urbanicity) at the municipality level (24). Specifically, we included municipality-level random effects to adjust for time-invariant contextual differences, and nested random effects by school year, time interval, and birth year to account for repeated observations within municipalities and variations across cohorts. Exposure random slopes were employed to allow the effects of SMU to vary across key covariates.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo minimize parametric assumptions about the functional form of the exposure-outcome association, we used penalized spline functions to model continuous covariates, including socioeconomic status and social media use. Additionally, we used data from Statistics Norway to adjust for urbanicity of municipalities. We adjusted for potential bias from random effects components by including municipality-level averages of exposure at baseline (25). Given the potential impact of COVID-19-related disruptions in 2021, we included an indicator variable for this period. Model residual variance was allowed to vary by sex and birth year.\u003c/p\u003e\n\u003cp\u003eTo estimate average treatment effects (ATEs) of social media use on mental health outcomes, we applied G-estimation methods (26). This approach ensures a robust estimation of social media\u0026rsquo;s impact on adolescent mental health while addressing confounding at multiple levels through pre-test adjustment, hierarchical modeling, and flexible functional forms.\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eSee Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e for time trends in self-reported anxiety, depressive symptoms, and social media between split by gender. The data show a steady increase in all measures, with a stabilization of depressive symptoms in recent years. Girls consistently reported more symptoms, and more time spent on social media than boys.\u003c/p\u003e\n\u003cp\u003eThe estimated standardized effects of SMU at Time 1 on depressive and anxiety symptoms at Time 2 are shown in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e. Results are shown separately for boys and girls. One standard deviation increase in social media use was not clearly associated with a change in anxiety scores for girls (-0.01 [-0.13, 0.10]) and was associated with an increase of 0.24 [0.09, 0.38] in anxiety scores for boys. Similarly for depressive symptoms, an increase of one standard deviation of social media use was not clearly associated with a change in depression scores for girls (0.01 [-0.06, 0.08]) and was weakly associated with an increase of 0.10 [0.01, 0.18] in depression scores for boys. Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e presents standardized and unstandardized effect estimates. The models explained a substantial amount of variation in the T2 outcomes, with \u003cem\u003er\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;.6 \u0026minus;\u0026thinsp;.8.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\u0026nbsp;\u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eUnstandardized and Standardized Results\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003ccolgroup cols=\"4\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eGender\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eStandardized\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eAnxiety estimate (95% CI)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eDepression estimate (95% CI)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBoys\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.09 (0.03, 0.15)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.04 (0.01, 0.07)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGirls\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.00 (-0.05, 0.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.00 (-0.03, 0.03)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBoys\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003exy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.24 (0.09, 0.38)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.10 (0.01, 0.18)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGirls\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003exy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e-0.01 (-0.13, 0.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.01 (-0.06, 0.08)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBoys\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ey\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.70 (0.26, 1.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.25 (0.03, 0.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGirls\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ey\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e-0.03 (-0.31, 0.24)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.02 (-0.15, 0.19)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eTable \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e: \u003cem\u003eThe table shows the estimated effect of social media use on anxious and depressive symptoms for boys and girls, respectively. Xy standardization refers to the expected increase in the outcome in standard deviations from a one standard deviation increase in social media use. Y standardization refers to the expected increase in standard deviations of the outcome from a one unit increase in social media use with respect to the survey scale.\u003c/em\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis large-scale study of Norwegian adolescents\u0026rsquo; internalizing problems found no consistent effect of social media use at the municipality level. SMU had a small effect on subsequent anxiety and depression in boys, but not girls. Our findings largely parallel those from individual-level studies showing that SMU and mental health complaints are generally correlated within cohorts (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e), but where a recent meta-analysis of experimental studies found no consistent evidence of a causal relationship (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eConfronted with deteriorating mental health among youths in large parts of the world, it is imperative that researchers make use of available large-scale data to identify its causes (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). Any claim to estimate causal effects rests on the assumption of no unmeasured confounding (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). We accounted for all time-varying unobserved confounding prior to the assessment of SMU by adjusting for the pre-treatment levels of internalizing problems, and we adjusted for time-constant confounders through a mixture of fixed and random effects (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e). The large proportion of variance in internalizing symptoms explained by our models supports our assumption that we have not left out key explanatory variables. Still, a crucial insight from the causal inference literature is that sufficient confounder adjustment cannot be verified empirically but relies instead on subject matter knowledge and debate (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). One potential source of bias unique to aggregate-level research is migration between municipalities across assessment waves (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e), but we know of no evidence of substantial migration patterns in Norway during the study period. Taken together, we the combination of a solid study design that allows adjusting for many unobserved confounders and the finding that the analysis model allows to explain a large portion of the variance make a causal interpretation of our results plausible.\u003c/p\u003e \u003cp\u003eWe found a small effect of SMU on anxiety and depression for boys, but we are hesitant to overinterpret this on two accounts. Firstly, the standardized effect size was quite small. Researchers in the social media field disagree widely over what to consider a meaningful effect (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). Our model suggests that a mean increase of 1 hour of SMU corresponds to a 0.09-point increase in anxiety symptoms and a 0.04-point increase in depressive symptoms on a scale from 1 to 4. We would argue that the effect on depression is negligible while the effect on anxiety is potentially meaningful if replicated. Secondly, the gendered pattern is opposite from what is commonly found, namely that there is a stronger association between SMU and internalizing problems in girls (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e). Girls typically spend more time on social media and report more internalizing symptoms than their male peers (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e), also in our data. It is not readily apparent why boys should be more susceptible to plausible group or individual-level mechanisms for mental health effects of SMU, such as social comparisons or contagion of negative emotions (\u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). For these reasons, Keyes \u0026amp; Platt (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e) have argued that any plausible cause for the increase in youth internalizing problems should display a gendered effect which is stronger for girls than for boys. We are thus inclined to interpret the overall result of our study as an inconsistent effect of SMU on youth internalizing symptoms at the municipality level.\u003c/p\u003e \u003cp\u003eA key contribution of our study is clearly estimating the direction of effects from social media to mental health, which has been a large caveat with much previous cross-sectional work (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). This also expands on more recent work by Nilsen and colleagues (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e), who found that increases in a municipality\u0026rsquo;s average social media use (SMU) were associated with concurrent increases in physical health complaints, particularly among girls. To the extent that physical health complaints and internalizing symptoms are related (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e), our contrasting findings could suggest that the concurrent associations reflect broader societal changes or reverse causality.\u003c/p\u003e \u003cp\u003eThe aggregated level of analysis in this study has both strengths and limitations. It is well suited to identify indirect network or cultural effects of living in a peer group with high social media use and can provide insight into the potential impact of municipality-wide initiatives aimed at reducing access to social media. However, drawing conclusions about individual-level effects from aggregated data requires stronger assumptions than we are willing to make (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). It remains possible that social media use affects youth mental health at the individual level within municipalities. For instance, if social media\u0026rsquo;s impact is highly non-linear\u0026mdash;primarily affecting extreme users, as some studies suggest\u0026mdash;individual-level effects may be obscured in aggregate data (\u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e). It is also possible that SMU disproportionately affects more vulnerable youth (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e), which at the level of municipalities can be balanced out by youth who gain something from social media (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e). Therefore, our findings do not justify a broad recommendation for parents or professionals to disregard concerns about individual youth\u0026rsquo;s social media use. The results are congruent with a perspective that social media has both negative and positive effects on individuals, which may balance each other out at the population level (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIt is important to recognize that our analysis does not preclude other worrying group-level effects of SMU, such as polarization of gender roles or political opinions (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e). It is also possible that the mere presence of social media in a community is enough to have downstream effects on mental health beyond the average time spent on social media, which could explain the discrepancy between our findings and the detrimental effects found from the introduction of Facebook to US colleges (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). However, the decline in Norwegian youth mental health began prior to the widespread adoption of smartphones and social media, indicating that these factors cannot be the sole drivers of the trend (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOne implication of our findings is that municipality-wide interventions to reduce the overall SMU among youth may not improve the population\u0026rsquo;s internalizing problems. This is in accordance with the mixed findings on the effects of school smartphone bans (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e). One recent UK study showed no differences in mental well-being among pupils in schools with restrictive or permissive phone policies (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e). However, the same study also found no differences in the overall time spent on social media between groups, suggesting that school phone bans in their current form are ineffective at reducing adolescent SMU (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e). Interventions targeting vulnerable youth with heavy social media use have shown more promising effects (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e), congruent with a perspective that SMU can be harmful to some even if no effect is found on the population level.\u003c/p\u003e\n\u003ch3\u003eLimitations\u003c/h3\u003e\n\u003cp\u003eThe study\u0026rsquo;s strengths include a large, nation-wide dataset of Norwegian adolescents and a causally informed analysis. However, besides the difficulty in translating our results to individual-level effects, some limitations should be noted. We assessed the average time youth reported spending on social media, with the upper category being 3 hours or more. It is possible that this is too narrow a range, as it hinders us from differentiating youth using social media 3.5 hours a day from 5\u0026ndash;6 hours a day (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e). Additionally, more nuanced measures of how youth spend their time on social media (active versus passive use; type of platform) is also important in future research.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eGiven the careful attempts to adjust for observed and unobserved confounders, the present study suggests that at the level of municipalities increasing in social media use is not a key driver of increases in youth\u0026rsquo;s depression and anxiety. The study is congruent with a perspective that social media may have both detrimental and positive effects on youth mental health which outweigh each other at the population level.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cem\u003eEthics approval and consent to participate\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe study adhered to the guidelines of the Helsinki Declaration. The study presents analyses of previously collected data. Participation was voluntary, and both students and their parents were informed accordingly. All students had to consent to participate in the study and could withdraw at any time. Parents had the option to withdraw their children from the study. For grades 8 to 10 the survey was completely anonymous and did not require approval from data protection authorities. The data protection officer at OsloMet approved data collection from high school students. The National Committee for Research Ethics in the Social Sciences and the Humanities (NESH) approved the use of passive parental consent for the Ungdata study. NESH is an independent ethics committee appointed by the Ministry of Education and Research. Additional details can be found on the project\u0026rsquo;s official website: https://www.ungdata.no/english/.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eConsent for publication\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAvailability of data and materials\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe full Ungdata survey used in the study can be downloaded from https://www.ungdata.no/wp-content/uploads/2020/09/Ungdata-Dokumentasjonsrapport-2010-2019-PDF-1.pdf. The dataset analyzed in the current study can be downloaded from https://surveybanken.sikt.no/en/study/query/ungdata/page/1. Codes for Oslo city districts are not in the file but can be requested from the Ungdata team.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCompeting interests\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eFunding\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe authors were funded by the Norwegian Ministry of Health and Care Services and by the Norwegian Institute of Public Health.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAuthors\u0026apos; contributions\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eGB contributed with the initial idea, study design and analyses. OT conducted the literature review and wrote the draft for the paper. Both authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAcknowledgements\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003ePotrebny T, Nilsen SA, Bakken A, von Soest T, Kval\u0026oslash;y K, Samdal O, et al. Secular trends in mental health problems among young people in Norway: a review and meta-analysis. European Child \u0026amp; Adolescent Psychiatry. 2024:1-13.\u003c/li\u003e\n\u003cli\u003eMcGorry PD, Mei C, Dalal N, Alvarez-Jimenez M, Blakemore S-J, Browne V, et al. The Lancet Psychiatry Commission on youth mental health. The Lancet Psychiatry. 2024;11(9):731-74.\u003c/li\u003e\n\u003cli\u003eBakken A. Ungdata 2024. Nasjonale resultater. 2024.\u003c/li\u003e\n\u003cli\u003eFerguson CJ, Kaye LK, Branley-Bell D, Markey P. There is no evidence that time spent on social media is correlated with adolescent mental health problems: Findings from a meta-analysis. Professional Psychology: Research and Practice. 2024.\u003c/li\u003e\n\u003cli\u003eTwenge JM, Haidt J, Joiner TE, Campbell WK. Underestimating digital media harm. Nature Human Behaviour. 2020;4(4):346-8.\u003c/li\u003e\n\u003cli\u003eValkenburg PM, Meier A, Beyens I. Social media use and its impact on adolescent mental health: An umbrella review of the evidence. Current Opinion in Psychology. 2022;44:58-68.\u003c/li\u003e\n\u003cli\u003eTang S, Werner-Seidler A, Torok M, Mackinnon AJ, Christensen H. The relationship between screen time and mental health in young people: A systematic review of longitudinal studies. Clinical Psychology Review. 2021;86:102021.\u003c/li\u003e\n\u003cli\u003eRose G. Sick individuals and sick populations. International Journal of Epidemiology. 2001;30(3):427-32.\u003c/li\u003e\n\u003cli\u003eNesi J, Choukas-Bradley S, Prinstein MJ. Transformation of adolescent peer relations in the social media context: Part 1\u0026mdash;A theoretical framework and application to dyadic peer relationships. Clinical Child and Family Psychology Review. 2018;21:267-94.\u003c/li\u003e\n\u003cli\u003eTwenge JM, Spitzberg BH, Campbell WK. Less in-person social interaction with peers among US adolescents in the 21st century and links to loneliness. Journal of Social and Personal Relationships. 2019;36(6):1892-913.\u003c/li\u003e\n\u003cli\u003eArenas-Arroyo E, Fern\u0026aacute;ndez-Kranz D, Nollenberger N. High speed internet and the widening gender gap in adolescent mental health: Evidence from hospital records. IZA Discussion Papers; 2022.\u003c/li\u003e\n\u003cli\u003eBraghieri L, Levy Re, Makarin A. Social media and mental health. American Economic Review. 2022;112(11):3660-93.\u003c/li\u003e\n\u003cli\u003eNilsen SA, Stormark KM, Heradstveit O, Breivik K. Trends in physical health complaints among adolescents from 2014\u0026ndash;2019: Considering screen time, social media use, and physical activity. SSM-Population Health. 2023;22:101394.\u003c/li\u003e\n\u003cli\u003eGreenland S. Ecologic versus individual-level sources of bias in ecologic estimates of contextual health effects. International Journal of Epidemiology. 2001;30(6):1343-50.\u003c/li\u003e\n\u003cli\u003eGreenland S, Robins J. Invited commentary: ecologic studies\u0026mdash;biases, misconceptions, and counterexamples. American Journal of Epidemiology. 1994;139(8):747-60.\u003c/li\u003e\n\u003cli\u003eBeneito P, Vicente-Chirivella \u0026Oacute;. Banning mobile phones in schools: evidence from regional-level policies in Spain. Applied Economic Analysis. 2022;30(90):153-75.\u003c/li\u003e\n\u003cli\u003eGoodyear VA, Randhawa A, Adab P, Al-Janabi H, Fenton S, Jones K, et al. School phone policies and their association with mental wellbeing, phone use, and social media use (SMART Schools): a cross-sectional observational study. The Lancet Regional Health\u0026ndash;Europe. 2025.\u003c/li\u003e\n\u003cli\u003eHern\u0026aacute;n M, Robins J. Causal Inference: What if. Boca Raton: Chapman \u0026amp; Hall; 2020.\u003c/li\u003e\n\u003cli\u003eAbbasi J. Surgeon general sounds the alarm on social media use and youth mental health crisis. JAMA. 2023;330(1):11-2.\u003c/li\u003e\n\u003cli\u003eDerogatis LR, Lipman RS, Rickels K, Uhlenhuth EH, Covi L. The Hopkins Symptom Checklist (HSCL): A self‐report symptom inventory. Behavioral Science. 1974;19(1):1-15.\u003c/li\u003e\n\u003cli\u003eKandel DB, Davies M. Epidemiology of depressive mood in adolescents: An empirical study. Archives of General Psychiatry. 1982;39(10):1205-12.\u003c/li\u003e\n\u003cli\u003eB\u0026uuml;rkner P-C. brms: An R package for Bayesian multilevel models using Stan. Journal of Statistical Software. 2017;80:1-28.\u003c/li\u003e\n\u003cli\u003eCarpenter B, Gelman A, Hoffman MD, Lee D, Goodrich B, Betancourt M, et al. Stan: A probabilistic programming language. Journal of statistical software. 2017;76:1-32.\u003c/li\u003e\n\u003cli\u003eGelman A, Hill J. Data analysis using regression and multilevel/hierarchical models: Cambridge university press; 2007.\u003c/li\u003e\n\u003cli\u003eNeuhaus JM, Kalbfleisch JD. Between-and within-cluster covariate effects in the analysis of clustered data. Biometrics. 1998:638-45.\u003c/li\u003e\n\u003cli\u003eRobins J. A new approach to causal inference in mortality studies with a sustained exposure period\u0026mdash;application to control of the healthy worker survivor effect. Mathematical Modelling. 1986;7(9-12):1393-512.\u003c/li\u003e\n\u003cli\u003eTwenge JM, Haidt J, Lozano J, Cummins KM. Specification curve analysis shows that social media use is linked to poor mental health, especially among girls. Acta Psychologica. 2022;224:103512.\u003c/li\u003e\n\u003cli\u003eKeyes KM, Platt JM. Annual Research Review: Sex, gender, and internalizing conditions among adolescents in the 21st century\u0026ndash;trends, causes, consequences. Journal of Child Psychology and Psychiatry. 2024;65(4):384-407.\u003c/li\u003e\n\u003cli\u003eOrben A, Meier A, Dalgleish T, Blakemore S-J. Mechanisms linking social media use to adolescent mental health vulnerability. Nature Reviews Psychology. 2024:1-17.\u003c/li\u003e\n\u003cli\u003eKramer AD, Guillory JE, Hancock JT. Experimental evidence of massive-scale emotional contagion through social networks. Proceedings of the National Academy of Sciences. 2014;111(24):8788-90.\u003c/li\u003e\n\u003cli\u003eOrben A. Teenagers, screens and social media: a narrative review of reviews and key studies. Social Psychiatry and Psychiatric Epidemiology. 2020;55(4):407-14.\u003c/li\u003e\n\u003cli\u003eAmendola S, Hengartner MP, Spensieri V, Grillo L, Cerutti R. Patterns of internalizing symptoms and disability functioning in children and adolescents. European Child \u0026amp; Adolescent Psychiatry. 2022;31(9):1455-64.\u003c/li\u003e\n\u003cli\u003eGreenland S, Morgenstern H. Ecological bias, confounding, and effect modification. International Journal of Epidemiology. 1989;18(1):269-74.\u003c/li\u003e\n\u003cli\u003ePrzybylski AK, Weinstein N. A large-scale test of the goldilocks hypothesis: quantifying the relations between digital-screen use and the mental well-being of adolescents. Psychological Science. 2017;28(2):204-15.\u003c/li\u003e\n\u003cli\u003eRiehm KE, Feder KA, Tormohlen KN, Crum RM, Young AS, Green KM, et al. Associations between time spent using social media and internalizing and externalizing problems among US youth. JAMA Psychiatry. 2019;76(12):1266-73.\u003c/li\u003e\n\u003cli\u003eElmer T, Fern\u0026aacute;ndez A, Stadel M, Kas MJ, Langener AM. Bidirectional associations between smartphone usage and momentary well-being in young adults: Tackling methodological challenges by combining experience sampling methods with passive smartphone data. Emotion. 2025.\u003c/li\u003e\n\u003cli\u003eBeyens I, Pouwels JL, van Driel II, Keijsers L, Valkenburg PM. Social media use and adolescents\u0026rsquo; well-being: Developing a typology of person-specific effect patterns. Communication Research. 2024;51(6):691-716.\u003c/li\u003e\n\u003cli\u003eKubin E, Von Sikorski C. The role of (social) media in political polarization: a systematic review. Annals of the International Communication Association. 2021;45(3):188-206.\u003c/li\u003e\n\u003cli\u003eKessel D, Hardardottir HL, Tyrefors B. The impact of banning mobile phones in Swedish secondary schools. Economics of Education Review. 2020;77:102009.\u003c/li\u003e\n\u003cli\u003eDavis CG, Goldfield GS. Limiting social media use decreases depression, anxiety, and fear of missing out in youth with emotional distress: A randomized controlled trial. Psychology of Popular Media. 2024.\u003c/li\u003e\n\u003cli\u003eKelly Y, Zilanawala A, Booker C, Sacker A. Social media use and adolescent mental health: Findings from the UK Millennium Cohort Study. EClinicalMedicine. 2018;6:59-68.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6647544/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6647544/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e\u003cbr\u003e\nRising internalizing problems among youth, particularly among females, have raised concerns about potential societal causes. Social media use (SMU) has emerged as a key focus, given its widespread adoption since the early 2010s. While individual-level cross-sectional studies suggest small to moderate correlations between SMU and internalizing problems, a complementary community-based perspective allows for assessing the effects of living in environments with high or low social media use on youth mental health.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e\u003cbr\u003e\nThis study investigates the effect of SMU on internalizing symptoms among Norwegian youth in a longitudinal study at the municipality level. The study uses data from the nationwide Ungdata surveys (2014–2024), covering 528 cohorts across 181 municipalities. Anxiety and depressive symptoms were assessed using items adapted from the Hopkins Symptom Checklist and the Depressive Mood Inventory, respectively. We applied a Bayesian multilevel model using the R package\u003cem\u003e brms\u003c/em\u003e, accounting for time-varying and time-constant confounders.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003cbr\u003e\n \u003c/strong\u003eAn additional hour of average SMU corresponded to a 0.70 [0.26, 1.14] SD increase in anxiety scores for boys but showed no clear association for girls (-0.03 [-0.31, 0.24]). For depressive symptoms, a one-hour increase in average SMU corresponded to an increase of 0.25 [0.03, 0.46] for boys, with no clear effect for girls (0.02 [-0.15, 0.19]). The models accounted for a substantial proportion of variance in T2 outcomes (r² = .6 - .8).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion\u003c/strong\u003e\u003cbr\u003e\nAssuming that all relevant factors influencing both social media use and youth mental health were accounted for, the findings suggest that living in communities with high social media use may have a small effect on youth mental health symptoms. However, these effects are inconsistent and primarily observed among boys.\u003c/p\u003e","manuscriptTitle":"The effect of municipality-level social media use on youth mental health","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-06 17:21:24","doi":"10.21203/rs.3.rs-6647544/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-18T13:57:21+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-10T11:00:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"82569609724194700296261974965964489368","date":"2025-07-09T22:48:44+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-08T05:07:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"304600646259036285489423738595186985264","date":"2025-07-07T07:28:17+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"191980830863164147571158963597988930606","date":"2025-06-25T22:16:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"32499361882588934065659752646012912394","date":"2025-06-24T12:43:44+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-06-04T16:11:51+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-04T15:58:52+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-05-26T11:04:44+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-05-21T15:41:09+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Public Health","date":"2025-05-21T15:40:01+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"912796e7-e9eb-4cdf-8965-51ea01e1c7a9","owner":[],"postedDate":"June 6th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-10-20T16:00:31+00:00","versionOfRecord":{"articleIdentity":"rs-6647544","link":"https://doi.org/10.1186/s12889-025-24727-4","journal":{"identity":"bmc-public-health","isVorOnly":false,"title":"BMC Public Health"},"publishedOn":"2025-10-14 15:57:19","publishedOnDateReadable":"October 14th, 2025"},"versionCreatedAt":"2025-06-06 17:21:24","video":"","vorDoi":"10.1186/s12889-025-24727-4","vorDoiUrl":"https://doi.org/10.1186/s12889-025-24727-4","workflowStages":[]},"version":"v1","identity":"rs-6647544","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6647544","identity":"rs-6647544","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}