Abstract
Evaluating whether research aligns with the global burden of disease is essential for equitable and effective scientific progress and improvement of human health. Without systematic evaluation of this alignment, science cannot respond to shifting health needs. Here we analyzed the distribution between research and disease, linking 8.6 million disease-specific publications to two decades of global disease burden data using a triangulated large language model approach. We find that since 1999, research and disease burden have seemingly become much more aligned; however, this is mainly because of regional declines in communicable disease burden, whereas the noncommunicable disease burden has increased and globalized. Meanwhile, research effort has not changed to match changes in disease burden. Our simulations suggest that without intentional alignment, the research-disease divergence will probably widen by a third over the next two decades, and be substantially accelerated by the reduction of US public funding for international research. Aligning research with health needs will require strategic investments, improved global coordination, open science policies and stronger, more equitable international partnerships to build resilience in a fragile research ecosystem.