Development of a human analogue ADHD diagnostic system for family dogs

Dogs exhibit natural variability in inattention, hyperactivity and impulsivity traits, sometimes with extreme manifestations resembling to Attention-Deficit/Hyperactivity Disorder (ADHD) symptoms in humans. Using human-standard diagnostic methods, we developed a comprehensive approach for supporting diagnosing ADHD in dogs (N = 1872) based on a validated questionnaire that consists of two parts: a symptom section that provides factor scores for three areas (inattention, hyperactivity and impulsivity), and a functionality section that evaluates the extent to which these symptoms cause functional impairments. Based on the functionality section, we classified dogs as functionally impaired if they showed impairment in at least 4 out of the 7 questions of at least one area (n = 116; 6.2% of the sample). Then, we projected the impaired-non-impaired categories on the ADHD total scores of the symptom section to establish potential cutoff values. The ROC curve analysis resulted in an excellent AUC value (0.861). A cutoff score of 26 was established based on well-defined sensitivity and specificity trade-offs. Finally, we applied the combined thresholds of the two parts of the questionnaire to identify dogs at-risk vs. non-at-risk for ADHD (n = 79; 4.22% of the sample). We offer the first replicable method to screen dogs with suspected ADHD. A final diagnosis could be set after dog-owner pairs participate in relevant behavioural tests and an expert consultation. Allowing for comparisons between ADHD vs. typical groups of family dogs, our method not only facilitates the improvement of the wellbeing of at-risk dogs, but also makes it more feasible to use dogs as a natural model species for studying ADHD.