A prospective study associated several baseline clinical ratings with later stimulant response in adults with ADHD. Its small sample and naturalistic treatment design do not establish a validated rule for selecting amphetamine for a new patient.
DiSalvo and colleagues followed 36 medication-naïve adults with ADHD for an average of 116 days after stimulant prescribing. Treatment involved amphetamine or methylphenidate formulations selected in clinical practice, rather than randomized assignment between drug classes.
The researchers examined demographics and several clinical scales using regression and classification analyses. The final sample contained 18 responders and 18 nonresponders, leaving relatively few observations for evaluating multiple candidate predictors.
Executive-function impairment and better quality of life emerged as indicators of response. Other reported problems, including internalizing and thought-related difficulties, were associated with lower efficacy. Individual scale components did not all imply the same direction of association.
These findings concern relationships among baseline ratings and an observed response definition. They do not show that changing a baseline score would cause a better response, or that one medication class would outperform another for a person with that score.
Prospective follow-up improves the ordering of measurement, but external validation remains necessary. A model developed and assessed in a small clinical sample can perform less well in a new population. Medication doses were not analyzed, and the clinic’s limited demographic diversity restricts generalization.
Clinical depression diagnoses were not included as a confounder in the models. The study supports further testing of accessible clinical predictors; it does not replace clinical evaluation, establish an amphetamine-specific response mechanism or provide an individual impairment test.
See a controlled comparison that answers a different treatment question.
Clinical predictors of stimulant efficacy in adults with ADHD. doi:10.1177/20451253261428807.