TikTok has emerged as a major source of information and social interaction for youth, raising urgent questions about how substance use discourse manifests and circulates on the platform. This paper presents the first comprehensive analysis of publicly visible, algorithmically surfaced substance-related content on TikTok, drawing on hashtags spanning all major substance categories. Using a mixed-methods approach that combines social network analysis with qualitative content coding, we examined 2,333 substance-related hashtags, identifying 16 distinct hashtag communities and characterizing their structural and thematic relationships. Our network analysis reveals a highly interconnected small-world structure in which recovery-focused hashtags such as \textit{\#addiction}, \textit{\#recovery}, and \textit{\#sober} serve as central bridges between communities. Qualitative analysis of 351 representative videos shows that Recovery Advocacy content (33.9\%) and Satirical content (28.2\%) dominate, while direct substance depiction appears in only 26\% of videos, with active use shown in just 6.5\% of them. These findings suggest that the algorithmically surfaced layer of substance-related discourse on TikTok is predominantly oriented toward recovery, support, and coping rather than explicit promotion of substance use. We further show that hashtag communities and video content are closely aligned, indicating that substance-related discourse on TikTok is shaped through organic community formation within platform affordances rather than widespread adversarial evasion of moderation. This work contributes to social computing research by showing how algorithmic visibility on TikTok shapes the organization of substance-related discourse and the formation of recovery and support communities.

翻译：TikTok已成为青少年获取信息和进行社交互动的主要平台，这引发了关于药物使用话语如何在平台上呈现和传播的紧迫问题。本文首次对TikTok上公开可见、算法推荐的药物相关内容进行了全面分析，涵盖了所有主要药物类别的主题标签。通过结合社交网络分析和定性内容编码的混合方法，我们研究了2,333个药物相关主题标签，识别出16个不同的主题标签社群，并刻画了它们的结构性与主题性关联。我们的网络分析揭示了一个高度互联的小世界结构，其中以康复为中心的主题标签（如#addiction、#recovery和#sober）充当了社群间的核心桥梁。对351个代表性视频的定性分析表明，康复倡导内容（33.9%）和讽刺性内容（28.2%）占主导地位，而直接描绘药物使用的内容仅出现在26%的视频中，其中展示主动使用的视频仅占6.5%。这些发现表明，TikTok上算法推荐的药物相关话语层主要面向康复、支持和应对，而非明确宣扬药物使用。我们进一步证明，主题标签社群与视频内容高度一致，这表明TikTok上的药物相关话语是通过平台可供性内的有机社群形成来塑造的，而非广泛存在的对抗性规避审核行为。这项工作通过展示TikTok上的算法可见性如何塑造药物相关话语的组织以及康复与支持社群的形成，为社交计算研究做出了贡献。