Background: The goal of this study is to gauge the consistency of Attention Deficit/Hyperactivity Disorder (ADHD) latent class models that are generated by different informants such as adolescents and parents. The consistency of adolescent-derived latent classes from two different samples was assessed and these results were then compared to the class structure generated by parent-report ADHD information.
Methods: Self-reported DSM-IV Criterion A ADHD symptoms of 497 adolescent males from a population-based twin study in the state of Missouri (USA) were subjected to principal components and latent class analysis, and findings were compared to previous results obtained from identical analyses using an adolescent sample from Porto Alegre, Brazil (N = 483).
Results: The bi-dimensional structure of self-reported ADHD symptoms was similar for both male adolescent groups, but explained less than 40% of the symptom variance in either sample. Two factors, one with loadings on inattention symptoms only and the other with loadings on hyperactive-impulsive symptoms only, were identified in the Missouri sample. Specific ADHD latent classes did not replicate well across the Missouri and Brazilian samples, and both groups were characterized by the presence of several combined symptom classes but few inattentive or hyperactive-impulsive classes.
Conclusions: While adolescent-report information across two different cultures can at least in part reproduce the two-factor structure of ADHD, results from latent class analysis suggest that adolescent reporting on their own symptoms is markedly different from the type of information parents provide about ADHD symptoms in their offspring. The current findings indicate that if male adolescents endorse any ADHD symptoms there is a tendency for them to report combined type problems.