Scrolling through adolescence: a systematic review of the impact of TikTok on adolescent mental health

The review addresses whether and how TikTok use is associated with adolescent mental health outcomes, considering TikTok’s unique features and adolescents’ developmental vulnerabilities. The context is a documented rise in youth mental health problems since the early 2010s, potentially linked in part to widespread smartphone and social media adoption. TikTok is highly popular among teens and characterized by an algorithmically curated, highly visual, rapid, and personalized content stream that may influence engagement, social comparison, body image, and exposure to mental health-related content. The study aims to synthesize evidence specific to TikTok (rather than aggregating across platforms) to clarify potential benefits and harms, and to explore how usage type and user characteristics relate to mental health.

Prior research on social media and youth mental health has been mixed, with umbrella and systematic reviews noting weak or inconsistent associations across platforms. Some studies suggest links between social media use and anxiety, depression, sleep problems, disordered eating, and reduced wellbeing, while others find null or mixed results. Few studies isolate platform-specific effects or consider platform affordances. TikTok’s distinct design (endless short-form video feed, algorithmic personalization, high visuality, and user-generated content dynamics) has drawn attention for potential impacts on body image, self-esteem, and possible social contagion of psychiatric symptoms (e.g., tic-like behaviors). The literature also emphasizes considering developmental context, individual differences (e.g., personality, emotion regulation, social comparison), and usage patterns (active vs. passive) to explain heterogeneity in outcomes.

Design: Systematic review conducted under PRISMA guidelines. Databases and dates: PubMed, Scopus, APA PsycINFO, and Web of Science (also termed Web of Knowledge) searched from inception to April 30, 2024, without year limits. Search strategy: ("adolescen*" OR "child*" OR "pediatric" OR "paediatric") AND ("tik tok" OR "tik-tok" OR "tiktok") AND ("mental health" OR "mental illness*" OR "psychopatholog*" OR "psychiatr*" OR "conversion" OR "tic disorder" OR "Tourette" OR "dissociative identity" OR "neurodevelopmental disorder" OR "ADHD" OR "attention deficit hyperactivity disorder" OR "suicide" OR "suicide attempt*" OR "self-injury" OR "self-harm" OR "self harm" OR "non suicidal self-injury" OR "body image" OR "eating disorder"). Douyin and other regional variants excluded due to differences. Eligibility criteria (inclusion): focus on TikTok; participants ≤19 years with data specifically on them; mental health-related outcomes (including neurodevelopmental/psychopathological conditions). Exclusion: non-English; reviews, case reports, commentaries, protocols, etc.; animal studies; content-only analyses without participant outcomes; studies combining platforms where TikTok impact cannot be isolated. Screening and selection: Duplicates removed; titles/abstracts screened, then full texts assessed by two independent reviewers with consensus procedures (third reviewer as needed). Reference lists were hand-checked for additional studies. PRISMA flow: Records identified n=1370 (databases) plus n=24 other sources; after duplicates n=1250; screened n=1250; excluded n=983; full texts assessed n=267; excluded n=247 for reasons (e.g., no specific ≤19 data n=116; no TikTok focus n=41; review n=26; etc.); included n=20. Data extraction: PICOS-guided extraction of authors/year, design, sample, controls (if any), social media use measures, assessments, outcomes, findings, and quality. Quality assessment: Adapted Newcastle-Ottawa Scale (9-star model); details in Supplementary Table S1.

Study corpus: 20 studies from 10 countries, totaling 17,336 adolescent participants (per abstract). Designs: 15 cross-sectional, 2 longitudinal, and 3 non-quantitative (2 case series; 1 narrative inquiry). No control groups. Quality indices ranged 4–6/9 (mean 4.8; median 5: fair quality) with substantial methodological heterogeneity. Usage characterization: Few studies reported total screen time or TikTok-specific time; one Italian clinical sample indicated TikTok as the preferred platform for 62.8% with mean 1.4 ± 1.0 h/day. Several studies distinguished active (posting/engaging) vs. passive (scrolling) use. Overall mental health associations:

• Passive TikTok time predicted lower life satisfaction and fewer positive emotions; active posting predicted higher life satisfaction (Wu et al.).
• Adolescents using TikTok showed higher mean anger than non-TikTok social media users, though usage duration was not associated with loneliness/anger (Sarman et al.).
• TikTok use was more common among internet-addicted youth and independently predicted higher depressive symptoms (Ilić-Živojinović et al.).
• Less extraverted teens and those prone to negative affective reactions to social media reported elevated depressive symptoms with higher use, particularly for Instagram and TikTok (Gentzler et al.).
• Two case series during COVID-19 linked viewing TikTok influencers with tics to rapid-onset functional tic-like behaviors (Hull et al.; Nagy et al.). Problematic TikTok use (PTU)/addictive use:
• Smartphone time and TikTok use predicted social media addiction risk; use of TikTok alongside other visual platforms increased risk (Marengo et al.).
• Flow-related factors (concentration, time distortion) were associated with PTU; parental active mediation attenuated the concentration–PTU link (Qin et al. 2022, 2023).
• TikTok use disorder correlated with memory loss, depression, anxiety, and stress (Sha & Dong).
• Latent class analysis suggested that differing time on TikTok did not map neatly onto psychosocial risk markers (social comparison, emotion dysregulation, distress) (Fortunato et al.).
• Offline isolation combined with greater TikTok use predicted higher exposure to online risks (Muñoz-Rodríguez et al.).
• Motivations: Passive consumption predominated; entertainment/affect drive use; heavy contributory users (especially pre-adolescent girls) exhibited fame-seeking motives (Bucknell Bossen et al.). Body image and self-esteem:
• TikTok users reported higher likelihood of body image issues (Sagrera et al.).
• Exposure to influencer advertising on TikTok/other platforms was linked to lower body satisfaction and heightened perceived importance of physical appearance to others (Feijoo et al.).
• Greater daily TikTok use associated with lower self-esteem; passive use and searches for food/diet/ED content related to lower self-esteem (Pruccoli et al.).
• Longitudinally, TikTok time did not predict changes in internalization of beauty ideals or body image self-discrepancy among girls (Maes et al.).
• No significant direct correlation between TikTok and eating disorders (SCOFF) in bivariate analyses, though high social media addiction increased ED likelihood (López-Gil et al.).
• Hypersexualized self-representations on TikTok may function as perceived empowerment but also invite criticism/peer pressure, potentially harming self-esteem (Soriano-Ayala et al.). Platform features/mechanisms:
• Algorithmic content curation, high visuality, and auto-looped short videos promote immersion ("flow"), concentration, and time distortion, linked to PTU.
• Active vs. passive use and user traits (introversion, negative affectivity, social comparison, emotion regulation) moderated outcomes. Geography/timeframe: Most studies from Europe; several conducted during COVID-19, potentially influencing usage patterns and outcomes.

The findings suggest that TikTok’s impact on adolescent mental health is multifaceted and likely contingent on usage type (active vs. passive), individual vulnerabilities (e.g., introversion, negative affectivity, social comparison, emotion regulation), and platform affordances (algorithmic short-form video feeds fostering immersive “flow”). Passive consumption relates to lower life satisfaction, while active creation can be associated with higher life satisfaction, indicating that engagement mode matters. Flow-related concentration and time distortion are plausible mechanisms for problematic use, with parental mediation emerging as a protective factor. Body-centered content and hypersexualization intersect with developmental sensitivities to peer feedback and self-presentation, potentially amplifying body image concerns and lowering self-esteem for some teens, though longitudinal evidence for causal change remains limited. The documented spread of functional tic-like behaviors during COVID-19 underscores potential social contagion via algorithmically amplified illness narratives, particularly among vulnerable youth. Overall, TikTok may confer both opportunities (creativity, connection, digital play) and risks (depressed mood, anger, PTU, body concerns, symptom contagion), with small-to-moderate associations and substantial heterogeneity, emphasizing the need for developmentally informed, individualized interpretations rather than time-based or platform-wide generalizations.

Evidence specific to TikTok indicates potential risks for adolescent mental health—lower life satisfaction during passive use, links to depressive symptoms and anger, problematic/addictive patterns mediated by flow experience, and increased body image/self-esteem concerns—alongside potential benefits related to creativity, self-expression, and connection. Causality remains unclear due to predominantly cross-sectional designs and methodological variability. A subset of adolescents appears particularly vulnerable, especially those with pre-existing psychopathology or traits (introversion, negative affectivity, social comparison). Future research should: employ longitudinal and experimental designs; integrate objective usage metrics with validated mental health measures; delineate active vs. passive use; examine algorithmic exposure and content categories; test protective strategies (e.g., active parental mediation, digital literacy); and consider developmental windows of sensitivity and sex differences. Policymakers and clinicians should adopt balanced, developmentally grounded approaches that mitigate risks while leveraging potential benefits.

Most included studies are cross-sectional and correlational, limiting causal inference. Selection bias is common (convenience/purposive samples), reducing generalizability. Reliance on self-report for both TikTok use and mental health outcomes introduces reporting and recall biases; many studies used non-validated, self-developed instruments, creating measurement bias and hindering cross-study comparability. Platform evolution and software updates risk instrument obsolescence. Publication bias may overrepresent adverse findings. Few studies quantified TikTok usage precisely; limited control for confounders (e.g., socioeconomic status, pre-existing conditions). No studies included control groups. Many were conducted during COVID-19, potentially inflating or distorting associations due to unique social contexts.