We compared early-diagnosed autism spectrum disorder (ASD) (defined as diagnosis by age 4 years) between the 2002 and 2006 birth cohorts, in five sites of the Autism and Developmental Disabilities Monitoring Network. In the 2002 cohort, the prevalence/1000 of early-diagnosed ASD was half the 8-year-old prevalence (7.2 vs. 14.7, prevalence ratio [PR] 0.5 [0.4-0.6]). Overall, the prevalence of early-diagnosed ASD did not differ between birth cohorts (PR 1.1 [0.9-1.3]). However, in three sites with complete case ascertainment, the prevalence of early-diagnosed ASD was higher for those born in 2006 versus 2002 (PR 1.3 [1.1-1.5]), suggesting possible improvement in early identification. The lack of change in two sites may reflect less complete case ascertainment. Studies in more recent cohorts are needed.
Keywords: Autism; Autism spectrum disorder; Birth cohort effects; Community diagnosis; Prevalence.