70% of people believe at least one divisive health claim. Science needs a new playbook

The assumptions health communicators have relied on for decades are spectacularly wrong. Doubts about nutrition, vaccination, and public safety recommendations are no longer a fringe view. They stem neither from a single ideology nor lack of education , and do not result from distrust of doctors or experts. In fact, our latest 2026 Edelman Trust Barometer Special Report: Trust and Health, based on a survey of 16,000 respondents across 16 countries, finds that a staggering 70% believe at least one of six divisive health claims about foods, vaccines , and medicines to be true. Understanding the worldview and concerns of the public is everything. CEOs and communicators in health must wake up to this radical new reality. Divisive health beliefs span the globe, highest in the developing nations of India (89%) and South Africa (88%), lowest in Japan, Canada, and the U.S. (50–61%). The divisive health beliefs apply equally across educational levels (university – degreed versus non-university – degreed) and are more acute among young people (79% for ages 18–34) and right-leaning voters (78%), though majorities hold for ages 55 and older (60%) and the left-leaning (64%). It is striking that only a slight majority (52%) say it is false that the risks of childhood vaccination outweigh the benefits, while slightly over one third (36%) of respondents say it is false that adding fluoride to drinking water is harmful — a public health staple in many Anglophone countries since the 1960s. The one-year drop in confidence (−10 points to 51%) in people’s ability to make informed health decisions for themselves and their families is staggering. From China to the UAE to Mexico, the majority of people feel their country is divided on key health issues, potentially leading to a profound loss of trust in the healthcare system. Artificial intelligence has already displaced medical expertise in the eyes of many; doctors are competing to influence health decisions with AI, peers, friends and …