Exploring Problematic TikTok Use and Mental Health Issues: A Systematic Review of Empirical Studies

TikTok’s rapid global adoption since 2017, its high engagement driven by short-form, algorithm-personalized videos, and rising daily usage have raised concerns about excessive and problematic use and associated mental health outcomes. Prior work suggests TikTok’s design (For You Page, endless scroll, variable reinforcement) fosters sustained engagement and may heighten addiction risks, particularly among adolescents and young adults. Despite growing empirical studies, a comprehensive synthesis focused specifically on problematic TikTok use and mental health has been lacking. The review’s objective is to synthesize empirical evidence on prevalence, risk factors, usage patterns, and mental health correlates of problematic TikTok use, differentiating problematic use from more severe TikTok Use Disorder, and to inform clinical identification, prevention, and intervention.

Emerging studies highlight TikTok’s unique addictive potential relative to other social media due to its large adolescent user base and advanced recommender algorithm. Research using Stimulus–Organism–Response and flow frameworks links enjoyment, concentration, and time distortion to addiction-like behaviors, with concentration as a key factor. Narrative reviews and empirical works describe design features (personalized streams, endless scroll, refresh mechanics akin to slot machine pulls) that increase screen time and complicate disengagement. Comparative work indicates higher engagement and distinctive cognitive-behavioral patterns on TikTok versus platforms like Facebook and Instagram, with emotional investment in likes and followers tied to problematic use and depressive symptoms. Prior content analyses show both risks (misinformation; glamorization of substances; competitive appearance comparisons negatively affecting body image) and potential benefits (recovery-oriented messages). Despite national security and privacy debates leading to bans in some countries, policy actions rarely address problematic use specifically, underscoring the need for focused evidence on mental health impacts.

Design: Systematic review adhering to PRISMA 2020 guidelines. Protocol: Registered on OSF (https://osf.io/cjf97). Databases and timeframe: PubMed, Embase, Scopus, Web of Science, and PsycINFO searched from inception to July 10, 2024; no language restrictions. Search strategy: Iteratively refined Boolean queries combining terms including TikTok, TikTok addiction, problematic TikTok use, social media addiction, short video addiction, TikTok use disorder, mental health, psychological traits, usage patterns, demographic factors, technology use, compulsive use, and behavioral addiction. Supplementary manual reference mining. Study selection: Records managed in EndNote X9; duplicates removed. Two independent reviewers screened titles/abstracts against predefined criteria, retrieved full texts, and independently assessed eligibility; disagreements resolved by discussion and a third reviewer when needed. Inclusion: Empirical studies focused explicitly on TikTok addiction/problematic use; peer-reviewed; cross-sectional, longitudinal, qualitative, mixed-methods, or experimental; with clear objectives, methodology, and outcomes related to TikTok use. Exclusion: Non-empirical (reviews, commentaries, editorials), studies on general social media without specific TikTok focus or only content-analytical approaches, grey literature, conference abstracts, and non–peer-reviewed articles. Data extraction: Conducted by two reviewers using a standardized form capturing authors/year, study title, domain, geography, sample characteristics, objectives, design, interventions (if any), risk factors, prevalence, usage patterns/trends, statistical measures, and key findings. Quality assessment: Applied JBI checklist for cross-sectional studies; CASP for qualitative studies; MMAT for mixed-methods; Cochrane RoB 2 for experimental RCTs; NOS for longitudinal observational studies. Assessment undertaken independently by two reviewers; consensus via discussion or third reviewer. Synthesis and analysis: Narrative synthesis organized thematically (prevalence, demographics, usage patterns, mental health contributors). Quantitative data summarized with descriptive statistics; studies compared to draw broader conclusions. Analyses conducted using IBM SPSS Statistics v28. Ethics and funding: Secondary synthesis of published data; ethical approval not required; no external funding; authors reported no conflicts of interest.

Scope and samples: 26 empirical studies; total participants = 11,462 across diverse countries (largest share from China, 45.8%; Asia overall 66.7%). Study designs: 19 cross-sectional (73.1%), 2 longitudinal (7.7%), 2 mixed methods (7.7%), 1 experimental (3.8%), 2 qualitative (7.7%). Prevalence and usage: Pooled prevalence of TikTok use across included samples was 80.19% (8,141/10,154). Highest usage among ages 18–29; females and lower socioeconomic status groups showed higher problematic use indicators. In Trinidad and Tobago, problematic TikTok use prevalence was 8.7% among university students; Indonesian eighth graders showed higher daily usage among females (approx. 3 hours/day vs. 1.5 for males) and higher addiction categories among females. Psychological and personality correlates: Problematic use associated with loneliness, boredom, low self-esteem, procrastination, neuroticism, and depressive tendencies. A cross-sectional study reported short-term exposure (20 minutes) increased depression scores by 12% and anxiety by 15% while decreasing perceived connectedness; frequent use linked to higher FOMO and loneliness. TTUD correlated positively with neuroticism (r≈.20) and depressive symptoms (r≈.31), negatively with age (r≈−.45), and was higher among females; lower conscientiousness related to TTUD via depressive inclinations. Mechanisms and platform features: Algorithm awareness, parasocial relationships, and FOMO predicted compulsive use and fatigue; flow components (enjoyment→concentration→time distortion) predicted problematic use, with active parental mediation moderating these effects. Scale development validated a 3-factor problematic TikTok use construct (obsession, escapism, lack of control) with good reliability. Body image and social comparison: Teen users exhibited higher social comparison and poorer body image relative to non-users; negative correlations observed between body image and comparison of abilities. Academic and functional impacts: Excessive use associated with poorer academic performance; academic performance mediated the relationship between TikTok addiction and depression/anxiety. COVID-19 context: Usage increased during lockdowns; lower SES and active TikTok use predicted Social Media Disorder among rural college students. Risk behaviors: Weekly TikTok use associated with hazardous alcohol consumption and problem gambling in adults; engagement with substance-related content often portrayed positively. Neurobiological evidence: fMRI showed personalized TikTok videos activated DMN and VTA more than non-personalized content, indicating higher reward valuation, attention capture, and reduced self-control. Demographic patterns: Female users more often showed problematic use (e.g., 67.3% of problematic cases among female university students), with heightened risk among younger users (<24) and lower SES groups.

Findings across diverse methodologies converge on a consistent pattern: TikTok’s design and algorithmic personalization drive high engagement that, for vulnerable subgroups (younger users, females, lower SES, individuals with higher neuroticism or depressive symptoms), associates with increased anxiety, depression, stress, and academic disruption. Flow states (enjoyment, concentration, time distortion) and social mechanisms (emotional investment in likes/followers, parasocial ties, FOMO) appear central to compulsive use and problematic patterns. Short-term exposure can elicit measurable shifts in mood and anxiety, and longitudinal evidence indicates distress intolerance mediates links from depression/social anxiety to problematic use. The platform’s unique engagement profile distinguishes it from other social media, with higher PSMU scores and complex self-esteem dynamics. TikTok functions as both a source of mental health information and potential misinformation, with notable risks of glamorizing harmful behaviors and exacerbating body image concerns, yet also offers recovery-supportive content. The evidence supports integrating digital/media literacy, caregiver-guided mediation, and structured use parameters to mitigate risks, while acknowledging heterogeneity in effects and the potential for positive uses. Owing to the predominance of cross-sectional designs and cultural concentration in Asia, causal inferences and generalizability remain limited, underscoring a need for longitudinal and interventional studies to clarify pathways and test solutions.

Across 11,462 participants, this systematic review identifies consistent associations between TikTok use—especially intensive and compulsive patterns—and adverse mental and behavioral outcomes, including anxiety, depression, and procrastination. Individual differences (age, neuroticism, conscientiousness, self-control), usage patterns, and platform design features contribute to problematic use. Given reliance on self-report and predominantly cross-sectional designs, prospective longitudinal and interventional research is needed to delineate causality and temporal dynamics. Clinically and educationally, targeted interventions for at-risk groups, promotion of digital literacy, and support for responsible engagement with online platforms are indicated, particularly for adolescents and young adults.

Most included studies use self-reported measures, introducing potential social desirability and recall biases; objective usage tracking is limited. The heavy reliance on cross-sectional designs precludes causal inference and temporal sequencing of effects. Generalizability is constrained by cultural concentration (e.g., large proportion of studies from China/Asia) and specific demographic samples (adolescents, students). Longitudinal studies face attrition concerns; follow-up completeness varies. Mixed methods studies often do not reconcile divergences between quantitative and qualitative findings. Potential moral panic and pre-existing parental/societal anxieties around technology may bias research framing. Observed associations may reflect selection effects (individuals with existing mental health issues or socioeconomic stressors gravitating to TikTok) rather than platform-induced pathology; dose–response relationships are understudied. Content-related risks (misinformation, glamorization of substances, competitive appearance comparisons) complicate interpretation of mental health impacts. Overall, more culturally diverse, longitudinal, and experimentally controlled studies with objective metrics are needed.