Temporal trends and spatial analysis of self-reported autism spectrum disorder in Brazil: a decade of school census...

To analyze the temporal and spatial trends of children and adolescents reporting a diagnosis of autism spectrum disorder upon school enrollment in Brazil between 2014 and 2023.

This ecological study employed time series (Mann-Kendall test with Hamed-Rao modification) and spatial analyses (global Moran's I). Enrollment data (2014-2023) were obtained from the Brazilian School Census (Censo Escolar), including self-reported autism spectrum disorder.

On a national level, reported autism spectrum disorder cases increased from 8.23 to 134.49 per 10,000 enrolled students (p < 0.001; slope ~11.18). Consistent upward trends were also found on a state level (tau = 1, p < 0.001), except in São Paulo (tau = 0.78, p = 0.002). The largest and smallest increases were in Acre and São Paulo, respectively. Moran's I increased from 0.1235 to 0.3233 (p = 0.001), indicating stronger spatial clustering, with local indicators of spatial association identifying persistent clusters in the southern/southeastern regions and emerging clusters in the northeastern region.

There was a steep nationwide increase in reported autism spectrum disorder cases between 2014 and 2023, with most states trending upward. Spatial clustering strengthened (increase in Moran's I), with local indicators of spatial association showing persistent clusters in the southern/southeastern regions and emerging clusters in the northeastern region, which indicates a need for region-specific policies and integrated care.