Increasing incidence of autism spectrum disorder: are we over-diagnosing?

Purpose In 2016, 1 in 54 children were estimated to have autism in the USA compared to 1 in 2,500 in 1955. This study aims to consider whether there has been a worldwide rise in incidence over time that is contributing to the rise in prevalence. Design/methodology/approach A systematic review of the literature with strict inclusion criteria was performed to identify large population-based studies that include raw incidence rate data with clearly defined diagnostic criteria. The data from the included studies were pooled and analysed descriptively to compare incidence rates by decade. Findings Seven studies were included in the final quantitative analysis including incidence rate data from 1988 to 2015 with 29,026 cases, over a total of 69,562,748 person years. Considering the most robust data, the incidence rate ratio between the decade 1990–1999 and 2000–2009 provides an estimated relative risk of 4.21 (95% CI; 4.11–4.32). If we compare the limited data available in 1988–1989 and 2010–2015, there is an estimated 75 times (95% CI 49.56–115.04) increased rate of diagnosis. Originality/value The broadening of diagnostic criteria and its increasing application in clinical practice needs further consideration to ensure individuals receive the most appropriate personalised support. A true rise in the incidence of autism will influence the level of service provision required in future with the potential for significant under resourcing. More detailed assessment of the clinical characteristics of those diagnosed will help predict risk factors for specialist service involvement in future.