The ADHD Paradox: Why Are Recorded Diagnoses Rising? A Structured Narrative Review of Prevalence, Diagnostic Coverage and the Limits of the Current Evidence

Recorded diagnoses of attention-deficit/hyperactivity disorder (ADHD) have risen steeply in several healthcare systems, and the rise is read either as evidence that ADHD is becoming more common or that it is overdiagnosed. This structured narrative review examines how far the increase reflects underlying population prevalence and how far it reflects diagnostic coverage: recognition, referral, assessment capacity, thresholds and the incentives attached to a formal diagnosis. Evidence from prevalence meta-analyses, routinely collected healthcare data, national surveys, quasi-experimental studies and service reports was kept separate by the quantity each source estimates. The three quantities have diverged. Meta-analytic estimates of underlying prevalence cluster between 5.29% and 7.2% in children, with 5.9% in young people and 2.5% in adults; those that tested year of study directly report no temporal trend once diagnostic criteria, impairment requirements and information source are accounted for, while the others were not designed to detect one. Recorded diagnosis among UK boys aged 10–16 years rose from 1.4% in 2000 to 3.5% in 2018, and reached 1.19% across all ages in England in 2025, with incidence accelerating from 2020. Screen-positive reporting among English adults rose from 8.2% in 2007 to 13.9% in 2023/24. Recorded prevalence spans a 72-fold range, from 3.60% in males aged 18–24 years to 0.05% at 65 years and over, where the estimated coverage gap approaches 93.5%. The same divergence appears in Danish, Swedish, Norwegian and United States data. Definitional changes alone raise estimated prevalence by 27% to 47%. The increase is better read as a change in diagnostic coverage than in disease frequency, and overdiagnosis near the diagnostic threshold and under-recognition in the tail occur simultaneously, in different population strata.