Trends in Autism Spectrum Disorder Cumulative Incidence and Incidence Density: A Systematic Review

Claims of an "autism epidemic" imply a true increase in ASD occurrence and changing underlying risk yet largely rely on rising prevalence estimates rather than incidence. Both measures are influenced by diagnostic practices, service access, documentation, and ascertainment; prevalence is additionally affected by service-related migration. Incidence studies follow a defined population previously not identified, to examine outcomes over a specified period and study area; incidence is more likely to capture temporal changes in underlying risk. We systematically reviewed autism incidence studies across differing methodological designs.

Following PRISMA 2020 guidelines, we identified population-based studies reporting ASD or Autistic Disorder incidence in individuals under 18 years. Studies were classified as Category A (birth-defined cohorts with definable risk periods) or Category B (dynamic population annual incidence density). Due to heterogeneity, incidence estimates were compared narratively.

Twenty-seven studies met inclusion criteria (14 Category A; 13 Category B). While ASD incidence increased, patterns varied by incidence framework, diagnostic definition, risk period, and ascertainment comprehensiveness (population coverage and detection of undiagnosed cases). High-comprehensiveness birth cohort studies tended to show inconsistent patterns and lower male-to-female ratios, whereas studies with moderate or low comprehensiveness often showed consistent increases. Increases were stronger for broader ASD category.

ASD incidence increased in dynamic, administratively defined study populations with moderate or low ascertainment comprehensiveness. In contrast, birth-defined cohorts with fixed risk periods and high ascertainment comprehensiveness showed inconsistent patterns. Overall, the available data do not provide compelling evidence of a true increase in ASD occurrence-an "autism epidemic."