A double-edged hashtag: Evaluation of #ADHD-related TikTok content and its associations with perceptions of ADHD

The study investigates how ADHD is portrayed and perceived on TikTok, a highly popular but understudied platform for mental health information. The research examines the psychoeducational quality of #ADHD content and its associations with young adults’ perceptions of ADHD. Motivated by the platform’s algorithmic curation, brevity-focused format, and creator incentives, the authors explore potential benefits (accessibility, destigmatization, community-building) and risks (misinformation, lack of nuance, overgeneralization). The central questions are: the popularity and reach of #ADHD content (RQ1); alignment of claims with DSM-5 diagnostic criteria and evidence-based treatments (RQ2); consumption patterns among young adults with differing ADHD statuses (RQ3); how young adults evaluate content compared to psychologists (RQ4); associations between consumption and perceptions of ADHD prevalence and symptom severity (RQ5); whether participants choose to watch a psychologist’s evaluation (RQ6); and whether viewing TikTok and the psychologist’s video relates to changes in confidence about one’s ADHD status (RQ7).

Prior work documents widespread health misinformation on social media across topics such as vaccination and noncommunicable diseases, with #autism and #ADHD among TikTok’s most-viewed health hashtags. Multiple studies of autism content on TikTok report high proportions of inaccurate or misleading videos (around 41%–32%). For ADHD, earlier work found 52% of videos misleading and only 21% useful, often centering anecdotal lived experiences without diagnostic nuance. Social media content can shape self-diagnosis behaviors, potentially leading to overgeneralization of normal experiences as pathological. Echo-chamber effects, confirmation bias, and algorithmic amplification may increase perceived credibility and sharing of repeated claims. Misinformation can undermine trust in professionals and evidence-based treatment (e.g., CBT portrayals), paralleling vaccine misinformation dynamics. Users’ reliance on social media for health information has been linked to doctor shopping and suboptimal patient-provider interactions. The present study extends this literature by systematically evaluating top #ADHD TikTok content and empirically testing its relation to young adults’ perceptions of ADHD.

Study 1: Ethical review determined no consent required for publicly available TikTok videos. A new TikTok account was used to search the hashtag #ADHD on January 10, 2023; the 100 most-viewed videos were screen-recorded (two excluded for lacking ADHD presentation/treatment content). Video metrics collected: views, comments, saves, likes, shares, length, captions/hashtags, creator-stated credentials in video and profile, ADHD/mental health posting history, and potential for financial gain (product sales, donation links). Research assistants noted distinct ADHD-related claims per video, classifying them as symptom or treatment claims; second and third authors verified classifications. Two licensed clinical psychologists (20+ years expertise) independently rated symptom claims: DSM-5 accuracy (Yes/No); symptom category (inattention, hyperactivity, impulsivity); realism of severity/impairment; if non-DSM, whether strongly linked to ADHD (e.g., working memory), better explained by another specific disorder, transdiagnostic symptom, or normal human experience; presence of nuance (acknowledging variability/applicability); and a global psychoeducation recommendation score (1–5). Inter-rater reliability: 84.8% agreement on DSM-5 accuracy (Cohen’s kappa = .686); category agreement 96.4% for impulsivity (k = .687), 94.9% hyperactivity (k = .739), 86.7% inattention (k = .672); global score ICC = .775 (95% CI .660–.851). Severity realism showed unacceptable reliability. Treatment claims were coded by an advanced PhD student and a postdoctoral fellow: treatment type, empirical support based on RCTs/meta-analyses/guidelines, and presence of nuance; perfect agreement reported. Study 2: Approved by UBC BREB; pre-registered; data collected Jan 9–Mar 25, 2024. Participants: N = 843 undergraduates (age 18–25; M = 20.23, SD = 1.62), predominantly women (79.4%), racially diverse (Asian 54.3%, White/European 35.3%). ADHD groups: formal diagnosis (n = 198), self-diagnosis (n = 421), no ADHD (n = 224), defined via pre-screening and in-study confirmations/diagnostic source descriptions; inconsistent responders removed. Procedure: demographics, ADHD diagnostic history, confidence in ADHD status (1–7), typical ADHD-TikTok consumption and perceptions, ASRS ratings for average person with ADHD and without ADHD, and evaluation of the top 5 and bottom 5 videos from Study 1 (presented in random order, blind to selection criteria) using the same global recommendation scale (1–5). Optional viewing of a psychologist’s explanatory video followed by re-rating confidence. Measures: ADHD-TikTok consumption (4 items; 1–5 Likert; α = .844); perceptions of typical TikTok info (4 items; 1–5 Likert; α = .877); recommendation scale (1–5 as in Study 1); ASRS-with (18 items; α = .892) and ASRS-without (18 items; α = .966). Analyses: hierarchical multiple regression for RQs 3–6 (demographics step 1; diagnostic status step 2; consumption step 3; VIF < 10; gender coded 0 = woman, 1 = man; forward-stepwise thresholds p < .005 enter, p ≥ .10 remove); binomial logistic regression for RQ6 (choice to watch psychologist video); repeated measures ANCOVA for RQ7 (confidence across three timepoints; Greenhouse-Geisser correction; covariate = typical consumption).

Study 1 (RQ1): Of 98 included videos, the 90 with complete view counts amassed 495,729,000 views (M = 5,470,322; SD = 6,410,138; range 30,800–33,900,000). Average per video: 984,684 likes; 9,728 comments; 71,302 saves; 19,911 shares. Mean video length 38.20s (SD = 36.75; range 5–216). Mean ADHD-related claims per video 2.99 (SD = 2.43; range 1–14). 93.9% of videos cited no source; 20.4% shared credentials in-video; 36.7% listed credentials on profile. Of credential-listers: 83.6% cited lived experience; 13.1% life coaches; 1.6% therapist/counsellor (license unspecified); 1.6% licensed MA-level providers; none PhD/PsyD/MD-level. 79.2% posted >1 ADHD-related video; 44.8% posted other mental health content; 50% promoted products or solicited donations. Top hashtags included #adhdtiktok, #fyp, #neurodivergent, #adhdawareness, #adhdinwomen, #adhdtok, #adhdcheck, #adhdprobs, #relatable, #foryoupage. Study 1 (RQ2): 92/98 videos (93.9%) contained only symptom claims; 4.1% only treatment claims; 2.0% both. Symptom claims: 48.7% judged to accurately reflect DSM-5 adolescent/adult ADHD symptoms (by ≥1 rater). Among accurate claims: 71.3% inattention; 27.2% hyperactivity; 16.2% impulsivity (categories not mutually exclusive; some rater disagreement). Of non-ADHD symptom claims (51.3%): 5.6% strongly associated phenomena (e.g., executive functioning/working memory deficits); 18.2% better illustrate other disorders (e.g., depression, anxiety, eating disorders); 42.0% transdiagnostic (e.g., emotion dysregulation); 68.5% normal human experience (percentages >100% due to rater disagreements). Nuance was rare: 4.1% acknowledged claims may not apply to all with ADHD; 1.4% acknowledged symptoms may also occur in people without ADHD. Global psychoeducation recommendation score: M = 1.78 (SD = 0.82; range 1–4); no video received a 5. Treatment claims (18 across 6 videos): 44.4% environmental modifications; 44.4% behavioral techniques; 11.2% medication (all negative portrayals) and workbooks. 55.6% aligned with empirically supported treatments; 38.9% based only on personal experience; 5.6% (workbook) with little empirical support; no nuance about treatment applicability. Study 2 (RQ3–RQ5): Consumption patterns: Self-diagnosed participants viewed more ADHD-TikTok than those with no ADHD (β = -0.25, p < .001), and less than those with formal diagnosis (β = 0.11, p = .002). Perceptions: Greater typical consumption predicted more favorable perceptions of typical ADHD-TikTok content (β = 0.36, p < .001) and greater likelihood to recommend typical content (β = 0.26, p < .001). Self-diagnosed participants rated typical content more favorably than no ADHD (β = -0.07, p = .040) and trended vs formal diagnosis (β = -0.07, p = .063). Video evaluations (top/bottom 5): Participants rated top 5 videos M = 2.82 (SD = 0.76) vs psychologists M = 3.60 (SD = 0.27); bottom 5 participants M = 2.32 (SD = 0.73) vs psychologists M = 1.10 (SD = 0.32). Participants rated top 5 higher than bottom 5 (t(824) = -23.68, p < .001, d = 0.61), but less favorably than psychologists for top 5 (t(842) = 7.71, p < .001, d = 0.35) and more favorably for bottom 5 (t(842) = 70.58, p < .001, d = 3.08). Greater consumption predicted higher recommendations for both top (β = 0.06, p = .025) and bottom videos (β = 0.06, p = .027). Self-diagnosed participants recommended bottom videos more than formal diagnosis (β = 0.06, p = .007). Perceptions of ADHD (RQ5): Participants estimated ADHD prevalence in general population M = 33.82% (SD = 19.50%). Self-diagnosed estimated higher prevalence than no ADHD (β = -0.16, p < .001) and formal diagnosis (β = -0.14, p < .001). Greater consumption predicted higher prevalence estimates (β = 0.13, p < .001). ASRS expectations: For average person with ADHD, self-diagnosed estimated more symptom severity than no ADHD (β = -0.16, p < .001) but less than formal diagnosis (β = 0.10, p = .007); consumption predicted higher perceived severity (β = 0.22, p < .001). For average person without ADHD, only greater consumption predicted expecting more severe ADHD symptoms (β = 0.10, p = .009). Study 2 (RQ6): 51.4% chose to watch the psychologist’s explanatory video; odds higher among self-diagnosis (OR = 1.5, p = .014) and formal diagnosis (OR = 2.5, p < .001) vs no ADHD; consumption did not predict choice. Study 2 (RQ7): No overall main effect of time on confidence, but time × diagnostic status interaction significant (F(3.38, 605.33) = 4.53, p = .003). After viewing TikToks (time 2), self-diagnosed participants increased confidence that they have ADHD; no ADHD participants decreased confidence in not having ADHD. After the psychologist video (time 3), no ADHD participants increased confidence back; self-diagnosed confidence did not change; formal diagnosis group remained consistently high across timepoints.

The findings demonstrate a substantial disconnect between mental health professionals’ evaluations of #ADHD TikTok content and young adults’ perceptions. Despite high reach and engagement, fewer than half of symptom claims aligned with DSM-5 criteria, nuance was rare, and treatment claims often lacked evidentiary grounding. Yet, frequent consumers perceived TikTok content as more helpful and accurate, were more likely to recommend both high- and low-rated videos, and estimated higher ADHD prevalence and greater symptom burdens for both those with and without ADHD. This suggests that algorithmic exposure, familiarity, and relatable lived-experience narratives may shape user perceptions differently from clinical standards. Importantly, young adults did differentiate between top and bottom videos, though their evaluations diverged from psychologists. Self-diagnosed individuals tended to view content more favorably, potentially reflecting reliance on relatable narratives in the context of barriers to formal diagnosis and care. Optional engagement with a psychologist’s explanatory video was relatively high among those with ADHD, and for no-ADHD participants, such professional input appeared to recalibrate confidence following exposure to TikTok videos. Overall, the results underscore both the democratizing potential of social media and the risks of misinformation and lack of nuance, highlighting the need for accessible, research-informed content and constructive dialogue between professionals and online communities.

This work characterizes the popularity and psychoeducational quality of top #ADHD TikTok videos and examines how consumption relates to young adults’ perceptions of ADHD. The study contributes evidence that widely viewed ADHD content frequently lacks alignment with DSM-5 and exhibits limited nuance, while users—especially frequent consumers and those with self-diagnoses—perceive such content more favorably and infer higher ADHD prevalence and symptom severity. These insights can inform clinical practice: professionals should engage patients about their social media experiences, provide clear explanations of diagnostic criteria and processes, and advocate for equitable, relatable, evidence-based ADHD information. Future research should empirically assess benefits of lived-experience content (e.g., destigmatization, community support), analyze creators followed by users (beyond viral videos), evaluate comment sentiment and potential “rage-bait,” and develop best practices for effective, nuanced short-form psychoeducation.

Study 1 analyzed only the 100 most popular #ADHD videos at one timepoint, excluding ads and deleted content, which may limit generalizability to less popular or previously removed material. Engagement metrics were recorded but comment sentiment (positive/negative/neutral) was not coded, potentially obscuring the nature of engagement. Study 2 relied on a university undergraduate sample (predominantly women), which may limit generalizability; all measures were self-report and ADHD diagnoses were not independently verified. Timing between video exposure and confidence assessments may have been insufficient to detect lasting effects; participants may have previously seen some videos, and some content could reflect outdated trends. Credential reporting by creators was limited and often centered on lived experience; financial incentives and subtle advertising were not fully captured. The TikTok search results can vary across users due to algorithmic personalization, potentially affecting reproducibility. Overall, these factors may constrain interpretation and generalizability.