Relative age effect in the probability of ADHD diagnosis in Finland: emergence at school entry

Abstract Background Children born later in the year have elevated probability to be diagnosed with attention-deficit/hyperactivity disorder (ADHD). We examine when this relative age effect (RAE) - suggested to indicate overdiagnosis - emerges, and has it changed across cohorts. Methods We used register-based data on 355,421 children born in Finland in 2005-2010. Information on ADHD medication purchases and healthcare records from specialized and primary care (ICD10-code F90) were used to identify age- and gender-specific probability of diagnosis. RAE at ages 5 to 10 was estimated by regressing the probability of diagnosis on relative age, measured as the day of birth within a year. RAE is presented as percentage point (pp) change in diagnosis probability by one month increase in relative age. Results The probability of ADHD diagnosis at ages 5 to 6 remained below 0.5% for boys in all cohorts and was even lower for girls. RAE was either non-existent or slightly negative, meaning that children born towards the end of the year were less likely than others to receive diagnosis before school age. ADHD diagnoses started to increase after school entry, and RAE emerged at age 7 among boys and by age 8 among girls in all cohorts. Among boys, the overall probability to receive ADHD diagnosis was highest at age 8, and increased from 0.7% to 1.4% between 2005 and 2010 cohorts, whereas RAE increased from 0.07 pp to 0.10 pp. Among girls, the probability of diagnosis at age 8 increased from 0.2% to 0.4%, whereas RAE remained around 0.02 pp. Conclusions The lack of relative age effect in the probability of ADHD diagnosis before school age and its emergence immediately after school entry aligns with the idea that ADHD diagnoses are influenced by grade-specific assessments, with those relatively young subject to overdiagnosis. Relative age effect persists across cohorts, and even increases among boys in younger cohorts. Key messages • Relative age does not predict the probability of ADHD diagnosis before school age, but immediately after school entry the probability of diagnosis increases for those youngest in class. • For the most recent cohorts, relative age effect in ADHD diagnoses following school entry increased among boys and remained persistent among girls.