Optimization of self- or parent-reported psychiatric phenotypes in longitudinal studies

Area of Science:

• Neuroscience
• Psychiatry
• Developmental Psychology

Background:

• The Adolescent Brain Cognitive Development (ABCD) study collects extensive psychiatric, neuroimaging, and genetic data from US adolescents.
• Reliability and validity of psychiatric data in the ABCD study require examination to understand psychiatric diseases.

Purpose of the Study:

• To explore and optimize the reliability and validity of psychiatric diagnostic constructs within the ABCD study.
• To assess the accuracy of parent- and child-reported psychiatric data.

Main Methods:

• Examined parent- and child-reported psychiatric data from 11,876 children over 4 years.
• Derived specific constructs for psychiatric diagnoses using longitudinal information.
• Calculated disorder rates and compared them to epidemiological literature.

Main Results:

• Broad diagnostic constructs (single time point) showed higher prevalence rates than epidemiological studies.
• Narrow diagnostic constructs (majority of assessments) better approximated literature-reported prevalences for most disorders (e.g., OCD: 2.6% vs. 2.3%).
• Narrow constructs also improved approximation of literature-reported comorbidity rates.

Conclusions:

• Self- or parent-report data in the ABCD study tend to overestimate psychiatric disorder prevalence, especially for lifetime prevalence.
• Accurate disorder classification necessitates supplementing self/parent reports with in-depth assessments or clinician-administered interviews.