Background and Objectives Social media, particularly TikTok, plays a pivotal role in shaping perceptions of feeding and eating disorders (FEDs). While it offers recovery-oriented content, it also exposes users to harmful narratives that may normalize disordered eating patterns. Understanding how FED-related content engages users is crucial for informing digital health strategies and content moderation policies. This exploratory study aimed to explore the impact of TikTok short videos on FED-related discourse, analysing how different content types of influence user engagement and whether recovery-focused messages generate higher interaction than weight-loss or pro-anorexia (pro-ana) content. Rather than providing a platform-level estimate, the dataset represents an exploratory snapshot of the Italian TikTok landscape during the collection period. Methods An observational cross-sectional study was conducted, involving three independent researchers who analysed 202 TikTok videos collected via 13 predefined FED-related hashtags. Videos were classified based on thematic content (pro-recovery, pro-ana, anti-pro-ana, or neutral) and emotional appeal, while engagement metrics (views, likes, comments, shares, saves) were statistically analysed. To account for the large variance in video age (mean ≈ 530 days, SD ≈ 343), all engagement metrics were normalized by the number of days each video had been online, resulting in per-day averages (e.g., views/day, likes/day). Undownloadable or muted videos (n=58) were excluded, which may have biased the dataset toward content that passed TikTok’s moderation filters. Results Recovery-focused videos were the most prevalent and received significantly higher engagement than pro-ana videos specifically promoting weight-loss behaviours (p=.003). Emotional expression was positively correlated with increased user interaction (p=.007). Additionally, creators with larger follower bases achieved greater engagement (p < .001). No significant differences were observed in broader comparisons between pro-ana and pro-recovery content, although these analyses were likely underpowered. Conclusion Our results underscore TikTok’s ambivalent role in FED discourse: while it amplifies recovery-oriented narratives, harmful content persists. Strengthening content moderation and digital health strategies is crucial to foster safer online spaces for vulnerable users.
Scrolling to heal or hurt? TikTok’s impact on eating disorders / De Marco, E., Polizzotto, G., Catellani, S., Bassi, L., Traina, S., Marchi, M., Galeazzi, G.M., Pingani, L.. - In: JOURNAL OF PSYCHOPATHOLOGY. - ISSN 2499-6904. - 32:2(2026), pp. 121-134. [10.36148/2284-0249-1582]
Scrolling to heal or hurt? TikTok’s impact on eating disorders
de Marco E.;Polizzotto G.;Catellani S.;Bassi L.;Marchi M.;Galeazzi G. M.;Pingani L.
2026
Abstract
Background and Objectives Social media, particularly TikTok, plays a pivotal role in shaping perceptions of feeding and eating disorders (FEDs). While it offers recovery-oriented content, it also exposes users to harmful narratives that may normalize disordered eating patterns. Understanding how FED-related content engages users is crucial for informing digital health strategies and content moderation policies. This exploratory study aimed to explore the impact of TikTok short videos on FED-related discourse, analysing how different content types of influence user engagement and whether recovery-focused messages generate higher interaction than weight-loss or pro-anorexia (pro-ana) content. Rather than providing a platform-level estimate, the dataset represents an exploratory snapshot of the Italian TikTok landscape during the collection period. Methods An observational cross-sectional study was conducted, involving three independent researchers who analysed 202 TikTok videos collected via 13 predefined FED-related hashtags. Videos were classified based on thematic content (pro-recovery, pro-ana, anti-pro-ana, or neutral) and emotional appeal, while engagement metrics (views, likes, comments, shares, saves) were statistically analysed. To account for the large variance in video age (mean ≈ 530 days, SD ≈ 343), all engagement metrics were normalized by the number of days each video had been online, resulting in per-day averages (e.g., views/day, likes/day). Undownloadable or muted videos (n=58) were excluded, which may have biased the dataset toward content that passed TikTok’s moderation filters. Results Recovery-focused videos were the most prevalent and received significantly higher engagement than pro-ana videos specifically promoting weight-loss behaviours (p=.003). Emotional expression was positively correlated with increased user interaction (p=.007). Additionally, creators with larger follower bases achieved greater engagement (p < .001). No significant differences were observed in broader comparisons between pro-ana and pro-recovery content, although these analyses were likely underpowered. Conclusion Our results underscore TikTok’s ambivalent role in FED discourse: while it amplifies recovery-oriented narratives, harmful content persists. Strengthening content moderation and digital health strategies is crucial to foster safer online spaces for vulnerable users.| File | Dimensione | Formato | |
|---|---|---|---|
|
document (1).pdf
Open access
Tipologia:
VOR - Versione pubblicata dall'editore
Dimensione
870.35 kB
Formato
Adobe PDF
|
870.35 kB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate

I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris





