A small retrospective study of 67 adults with ADHD found that those who also had alcohol or cannabis use disorders were far less likely to respond to standard pharmacological treatment. While the associations were statistically significant, the study’s limited size and observational design mean these findings are exploratory and require confirmation in larger, prospective trials before informing clinical decision-making.

Adult ADHD frequently co-occurs with substance use disorders, yet most pivotal pharmacotherapy trials systematically exclude patients with active addiction. This creates a clinical blind spot: clinicians routinely prescribe stimulants and non-stimulants to patients whose comorbidity profiles have never been adequately studied. Understanding which comorbid conditions may blunt treatment response is essential for setting realistic expectations, tailoring monitoring intensity, and designing the next generation of clinical trials that reflect actual patient populations.

Adult ADHD is increasingly recognized as a lifelong neurodevelopmental condition rather than a childhood disorder that resolves. In real-world clinical settings, many adults present with complex comorbidity profiles including substance use disorders, autism spectrum disorder, and mood conditions, yet treatment guidelines are largely derived from trials that exclude these very patients. This retrospective cohort study, drawn from a specialized Italian outpatient clinic, sought to identify whether specific comorbidities predicted poorer response to routine pharmacological ADHD treatment. Using DSM-5-TR criteria for diagnostic classification and the validated DIVA-5 structured interview for ADHD confirmation, the investigators assessed 67 consecutively evaluated adults.

The overall clinical response rate, defined by a clinician-rated Clinical Global Impression-Improvement (CGI-I) score of 1 to 3, was 71.6%. In parsimonious multivariable logistic regression models, alcohol use disorder was associated with approximately 90 to 92 percent lower odds of response (OR approximately 0.08 to 0.10, p = 0.010 to 0.026), and cannabis use disorder with approximately 76 to 80 percent lower odds (OR approximately 0.20 to 0.24, p = 0.014 to 0.028). Autism spectrum disorder showed a descriptive trend toward lower response but did not retain statistical significance after adjustment (p approximately 0.11 to 0.15). The authors appropriately characterize these findings as exploratory and hypothesis-generating, acknowledging that the small sample, single-center design, absence of standardized treatment exposure documentation, and lack of a control group preclude causal inference or clinical guidance. Larger, prospective, multi-site studies with systematic medication tracking are needed.

This study asks exactly the right question. Most of my adult ADHD patients do not look like the clean, uncomplicated participants in landmark stimulant trials. They come with layered histories of self-medication, undiagnosed substance use, and sometimes autism spectrum traits that were missed for decades. The finding that alcohol and cannabis use disorders are associated with dramatically reduced treatment response feels clinically intuitive, but I want to be careful not to confuse face validity with established fact. With only 67 patients and no controlled treatment protocol, these odds ratios could shift substantially in a larger, more diverse sample.

In practice, when I see an adult ADHD patient with active or recent substance use, I already adjust expectations and monitoring frequency. I tend to prioritize stabilizing the substance use component alongside ADHD treatment rather than treating them sequentially, and I have more frequent check-ins during the first months to detect non-response early. This study does not change that approach, but it reinforces the need to design our clinical workflows around the patients we actually see, not the idealized populations in textbooks.

This study sits very early in the research arc for understanding treatment response modifiers in comorbid adult ADHD. While it provides a statistical signal, it is best understood as a structured clinical observation that identifies candidate predictors for future investigation. The absence of treatment exposure standardization means we cannot distinguish between pharmacological non-response and issues of adherence, dosing adequacy, or medication selection. Clinicians should note the direction of these associations without incorporating the specific odds ratios into treatment planning.

From a pharmacological standpoint, the co-occurrence of active substance use disorders introduces important safety considerations. Stimulant medications carry misuse liability that may be amplified in patients with concurrent alcohol or cannabis use disorder, and non-stimulant alternatives such as atomoxetine or guanfacine may warrant earlier consideration in this population. Alcohol use can alter hepatic metabolism of multiple ADHD medications, and chronic cannabis use may independently impair executive function in ways that overlap with ADHD symptoms. The single most actionable recommendation from this study is that clinicians treating adult ADHD should systematically screen for and document substance use disorders at baseline and use that information to inform monitoring intensity and follow-up scheduling.