Abstract
Health messaging often employs gain-framing (highlighting behaviour benefits) or loss-framing (emphasizing nonengagement risks) to promote behaviour change. This study examined how neural responses to gain- and loss-framed messages predict changes in physical activity. We conducted a mega-analysis of raw fMRI and pedometer/accelerometer data from four studies (N = 240) that tracked brain activity during message exposure and real-world physical activity longitudinally. Focusing on brain regions theorized by the Affect-Integration-Motivation framework-the anterior insula, ventral striatum, ventromedial prefrontal cortex, dorsal striatum, and presupplementary motor area-we found that baseline physical activity levels moderated brain-behaviour relationships in response to message framing. More active individuals increased physical activity post-intervention when these brain regions responded more strongly to loss-framed messages, suggesting that neural sensitivity to inactivity risks may reinforce behaviour maintenance in this group. Conversely, less active individuals increased physical activity when brain responses were stronger to gain-framed messages, indicating that sensitivity to activity benefits may facilitate action initiation in this group. These findings suggest that message effectiveness depends on the interaction between framing, neural processing, and pre-existing behavioural patterns. By linking neurocognitive mechanisms with real-world outcomes, we highlight the importance of personalized, neuroscience-informed health interventions tailored to individual neural and behavioural characteristics to optimize behaviour change strategies.