Universal mental health awareness programmes, mindfulness lessons and structured emotional literacy curricula have been introduced into schools at scale over the past decade on the assumption that talking about anxiety, teaching children to label and monitor their internal states, and building general mental health awareness will reduce the prevalence of anxiety disorders in the school population. This paper reviews the evidence and finds the opposite has occurred. Reported rates of anxiety and other internalising conditions among children and adolescents have risen sharply over the same period these programmes expanded. The largest and most rigorous trial of school-based mindfulness training ever conducted found no benefit on any primary outcome. Multiple controlled studies have found that some school-based mental health interventions actively increase internalising symptoms in a meaningful subgroup of participants, an effect researchers now describe openly as iatrogenic harm. This paper sets out the evidence for what Oxford researcher Lucy Foulkes has termed the prevalence inflation hypothesis, reviews the mechanism by which sustained attentional focus on internal threat cues can activate rather than deactivate a biologically predisposed anxiety response, and argues that the well-intentioned expansion of mental health talk in schools has, for a meaningful subset of children, functioned as a trigger rather than a shield.