1) To estimate the prevalence of ADHD according to the DSM-5 (dADHD), reported medical ADHD diagnosis (mADHD), and current psychostimulants use in a representative pediatric sample in Brazil;

2) To investigate the predictive value of mADHD relative to DSM-5 and reasons for false positives and false negatives. 

Concerns exist about whether Attention Deficit Hyperactivity Disorder (ADHD) is overdiagnosed in the pediatric population, with consequent over-treatment.

Sample consists of 10,912 children (5 to 18 years) recruited at the school system in 89 cities of 18 Brazilian states. Parents and teachers were interviewed using standardized and validated questionnaires. ADHD was ascertained as per the DSM-5 criteria and the mental health status with the validated Brazilian version of the Strengths and Difficulties Questionnaire (SDQ). Relative risks were modeled using univariate and multivariate analyses. 

Of 10,912 children, 1.7% of children were in use of psychostimulants, 6.6% had a mADHD. A total of 486 (4.5%) met DSM-5 criteria for ADHD (dADHD). Of children with mADHD, only 25.3% had dADHD (74.7% of false positives); however, of children with dADHD, 62.6% had never received a medical diagnosis of ADHD; of children using psychostimulants, only 26.2% had dADHD. The risk for false-positive diagnosis was higher in children studying in private vs. public schools, from bigger cities, and from higher income classes. Other risk factors were below average school performance and having the ADHD symptoms reported by the parents but not by the teacher (p<0.001). The risk for false-negative diagnosis was higher in children from smaller cities, from lower income classes, divorced parents, and below average school performance (p<0.01).

These findings suggest that social and gender disparities, as well as heuristic diagnosis instead of well-established diagnostic criteria, are associated with misdiagnosis and potentially with mistreatment of ADHD in children and adolescents.