Table 1_Accuracy, potential harm, weight stigma, and commercial influence in short-form social media content on incretin-based therapies for obesity.docx

Normalizing to views changed the picture. Likes per 1000 views did not differ (24.7 vs 20.4, p = 0.069), and comments per 1000 views were higher on YouTube Shorts (1.2 vs 0.8, p < 0.001), where videos were also older (720 vs 221 days, p < 0.001). Clinical accuracy did not differ significantly (23.4% [95% CI 16.0–32.9%] vs 39.0% [95% CI 30.0–48.8%], p = 0.062).

Hazard potential was more common on TikTok (31.9% vs 17.0%, p = 0.015), but this came from promotion of compounded products, four times more frequent on TikTok (14.9% vs 4.0%, p = 0.009). Restricted to directly unsafe clinical content, the platforms did not differ (20.2% vs 13.0%, p = 0.176). Obesity stigma (p = 0.118) and commercial influence (p = 0.438) were comparable.</p>Conclusion<p>Most short-form videos about incretin-based obesity therapies carried no accurate clinical information, and accurate videos drew only 24% of cumulative views.

Promotion of compounded products clustered on TikTok and accounted for the platform difference in hazard potential. Evidence-based, bias-free communication is needed if these platforms are to educate the public about obesity pharmacotherapy.</p>