Differential diagnosis and comorbidity of ADHD and anxiety in adults

The aim of this study was to examine symptom profiles of people diagnosed with attention-deficit/hyperactivity disorder (ADHD) and/or anxiety (ANX) in order to determine the validity of widely used ADHD and ANX rating scales for differential diagnostic use and to develop modified measures that take symptom overlap into account.

A cross-sectional design was used to assess differences in rating scale scores between clinical (n = 52) and control (n = 74) samples as well as differences among subgroups of the clinical sample (22 ADHD; 16 ADHD + ANX; 14 ANX).

Participants completed an online questionnaire where they responded to the Conners Adult ADHD Rating Scale (CAARS; Conners, Erhardt, & Sparrow, ) and State Trait Anxiety Inventory scales (STAI; Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, ).

Results showed that the CAARS and STAI had limited sensitivity and specificity and may lack in ability to differentially diagnose ADHD and/or ANX. Cluster analysis was used to guide the proposal of modifications for the two scales, which were to use inattentive items only for the CAARS and to exclude state ANX-present items on the STAI for use in differential diagnosis. Further parametric analysis supported these proposed modifications.

Clinicians should be made aware of the limitations of the CAARS and STAI scales in terms of specificity, when used to inform differential diagnosis of ADHD and ANX. Further analysis on the psychometric properties of these modified scales is needed in order to confirm that they are valid and reliable scales.

Clinical implications It is possible that widely used self-report rating scales are not valid for use in the context of assessing adult ADHD when ANX is present. Clinicians should take alternative approaches to measuring ADHD symptoms in the context of ANX. Findings of the present study suggest the use of inattentive items only for the CAARS and to exclude state ANX-present items on the STAI for differential diagnostic use. Limitations of the study The sample sizes of the clinical subgroups were relatively small. Diagnoses were not confirmed using a semi-structured clinical interview. Alternative cluster approaches (e.g., two-step clustering using larger samples) would provide further insight.