Obesity projections built on decades of trend data may already be obsolete — and understanding why matters enormously for public health planning, health equity initiatives, and personal risk assessment. A critical methodological gap in widely cited national forecasting models could lead policymakers and clinicians to misallocate resources at precisely the moment when intervention is most consequential.
A correspondence published in JAMA challenges the forecasting assumptions underlying a high-profile analysis of U.S. adult obesity prevalence by race and ethnicity spanning 1990 to 2022, with projections extending to 2035. The original study employed Global Burden of Disease modeling techniques — regarded as among the most rigorous available — to estimate obesity trajectories across demographic and geographic strata. However, the critique identifies a structural flaw: these models assume historical trends will persist unless disrupted by an external shock. The rapid clinical adoption of glucagon-like peptide-1 receptor agonists (GLP-1 RAs), such as semaglutide and tirzepatide, constitutes precisely such a shock, one already embedded in current prescription data yet absent from the forecast architecture.
This is not merely a statistical footnote. GLP-1 receptor agonists have demonstrated 15–22% body weight reductions in randomized trials, a magnitude without precedent in pharmacological obesity treatment. Their uptake has been uneven across income levels, insurance coverage, and racial groups — meaning that race- and ethnicity-stratified projections are especially vulnerable to distortion if differential access patterns go unmodeled. For adults tracking personal and population-level obesity risk, the implication is that 2035 forecasts circulating in policy documents may substantially overestimate future obesity burdens in well-insured populations while underestimating persistent disparities among underserved groups. This is an incremental but genuinely important methodological correction: not paradigm-shifting science, but a necessary recalibration that forecasters should act on before these projections shape a decade of health policy.