Content characteristics, information quality, and user engagement of anxiety disorder–related videos on TikTok and Bilibili: a cross-sectional comparative study

Social media platforms have become important sources of mental health information, but the quality and reliability of anxiety disorder–related short videos remain unclear. This cross-sectional study evaluated 238 Chinese-language anxiety disorder–related videos from TikTok and Bilibili. Video characteristics, content coverage, uploader type, user engagement, and information quality were assessed using the Global Quality Score (GQS), modified DISCERN (mDISCERN), and JAMA benchmark criteria. Platform and uploader differences were analyzed with appropriate nonparametric tests and Holm adjustment. An adjusted ordinal logistic regression model with uploader-level cluster-robust standard errors was used to evaluate factors associated with mDISCERN scores. Bilibili videos were significantly longer than TikTok videos (rank-biserial r = 0.811, 95% CI: 0.728–0.882), while TikTok videos showed higher user engagement. Prevention-related content was more frequently presented on Bilibili (OR = 2.999, 95% CI: 1.636–5.497; adjusted P = 0.002). Although TikTok had higher mDISCERN scores in unadjusted comparisons, this association disappeared after adjustment for uploader type (aOR = 1.19, 95% CI: 0.53–2.68; P = 0.672). Non-psychiatric physicians showed higher adjusted mDISCERN scores than individual users (aOR = 2.40, 95% CI: 1.02–5.65; P = 0.046). User engagement metrics were not consistently associated with information quality. Differences in anxiety disorder–related video reliability between platforms may partly reflect differences in uploader composition rather than platform effects alone. Uploader type was associated with information reliability, but these observational findings do not establish causality. Not applicable.