Pandemic-Associated Disruption and Regional Divergence in Childhood Obesity Trajectories in Brazil, 2008-2025

Background: National averages may conceal whether Brazilian regions experienced different departures from, and returns to, their projected pre-pandemic childhood obesity trajectories.We evaluated long-term trends, pandemic-associated deviations, and changes in regional inequality among children aged 5-10 years monitored by the Brazilian Food and Nutrition Surveillance System (SISVAN) from 2008 through 2025.Methods: We conducted a nationwide ecological interrupted time-series study using publicly available surveillance data.Joinpoint regression estimated annual percent changes.Segmented regression estimated the 2020 level change and subsequent slope change nationally and in each region.From these coefficients, we calculated annual and cumulative fitted deviations from continuation of the pre-pandemic trajectory.Regional absolute and relative gaps were assessed at prespecified years.Results: Prevalence increased from 10.45% in 2008 to 15.41% in 2025 and peaked at 17.84% in 2021.The national model showed a 2.54-percentage-point level increase in 2020 (P=0.001) and a -0.646-percentagepoint annual slope change thereafter (P=0.002).The fitted deviation from the projected pre-pandemic trajectory decreased from +2.54 points in 2020 to approximately zero in 2024 and -0.69 in 2025.Regional recovery diverged: the South and Southeast crossed their projected trajectories in 2023, the Northeast and Center-West in 2025, whereas the North remained 1.28 points above projection in 2025.The maximum regional prevalence gap widened from 3.18 points in 2008 to 10.78 in 2021, then narrowed to 6.66 in 2025.Conclusions: The pandemic period was associated with a transient national departure from the prepandemic trajectory, but recovery was regionally asynchronous.Persistent excess relative to projection in the North contrasted with faster reversal in the South and Southeast, while regional inequality remained wider in 2025 than in 2008.Interpretation must account for changes in surveillance coverage.