When Visual Consistency Persuades: Multimodal Congruence, Health Risk, and Health Information Engagement in Digital Media

Visual content is central to online health communication, yet little is known about how multimodal relationships are associated with audiences’ click-through choices. Drawing on processing fluency and risk communication perspectives, this study distinguishes image–text congruence from inter-image coherence and examines their linear, nonlinear, and risk-contingent associations with click-through rate (CTR). We analyze 34,341 health-related articles published by 2184 authors. Clicks are modeled relative to impressions using beta-binomial regression with author and publication-date fixed effects. Image–text congruence is positively associated with the probability that an impression converts into a click, whereas inter-image coherence has an inverted U-shaped association with this probability. Moderate coherence is associated with the highest CTR, while excessive similarity corresponds to declining CTR. Higher topic health risk corresponds to a stronger association between image–text congruence and CTR and a weaker high-coherence decline. The results remain consistent across alternative model specifications, visual measures, outlier exclusion, and different observation windows. By distinguishing cross-modal congruence from within-modal coherence, this study clarifies their different persuasive functions and demonstrates that their associations with click-through choice depend on topic health risk. These findings inform research on digital persuasion and the design of online health content.