Mapping e-cigarette communication and dissemination on social media: an integrated infodemiology framework

Drawing upon the infodemiology and issue-attention cycle frameworks, we examined communication and information diffusion of electronic cigarettes (e-cigarettes) on social media. We analyzed 83,527 Facebook posts and 14,985 Instagram posts using machine learning algorithms and manual content analysis to identify how topics, sentiments, and Key Opinion Leaders (KOLs) evolve throughout various stages of the issue-attention. Regression models were employed to investigate the influence of topics and emotions on information dissemination across platforms. We identified six key topics: (a) Advertisement, promotion and marketing; (b) Policy and regulation; (c) Community building and interactions; (d) Types of flavors; (e) Popular science discourse; and (f) Health concerns. Positive sentiments toward e-cigarettes were prevalent on both platforms, with trust emerging as the foremost emotion. However, sentiment patterns diverged over time, with Facebook showing increasing positivity while Instagram experienced a decline. Significant disparities were observed between Facebook and Instagram in topic distribution, diffusion patterns, and influential figures. Policy and regulation-related content showed the highest diffusion potential on both platforms. Media entities dominated as KOLs on Facebook, while e-cigarette brands and community accounts had a stronger presence on Instagram. The diversity of KOLs increased during peak public interest stages, with media gaining more control over key topics. Our findings reveal the dynamic nature of e-cigarette discourse, highlighting how narratives evolve in response to platform characteristics and public attention shifts. This study contributes to understanding the complex interplay between content, platforms, and KOLs, offering valuable insights for public health officials and tobacco regulatory bodies.