Medical cannabis patients with severe anxiety showed no meaningful improvement over 12 months | Research

The takeaway

Among 526 medical cannabis patients followed quarterly for a year, those with minimal or moderate anxiety showed slight decreases, but the severe anxiety group remained stable, suggesting medical cannabis may not adequately treat severe anxiety.

Medical cannabis prescribers, anxiety disorder clinicians, medical cannabis program administrators

21% of patients had severe anxiety that did not improve

What the researchers found

Latent class growth analysis identified three anxiety profiles: Minimal (43%), Moderate (36%), and Severe (21%). The Minimal and Moderate groups showed anxiety decreases, but reductions were not clinically meaningful. The Severe group remained stable without improvement. Younger age and female sex predicted Moderate and Severe profiles. Lifetime PTSD and anxiety as primary qualifying condition predicted the Severe group only.

Why it matters

Anxiety is one of the most common reasons people seek medical cannabis, yet this longitudinal data shows that patients with severe anxiety do not improve over 12 months of use. This finding challenges the assumption that medical cannabis effectively treats anxiety and highlights the need for additional support for severely anxious patients.

The numbers in context

n=526; 3 profiles: Minimal (43%), Moderate (36%), Severe (21%); Minimal and Moderate showed slight decreases; Severe remained stable; younger age and female sex predicted worse profiles; PTSD predicted Severe profile

How the study worked

Longitudinal study of 526 Pennsylvania medical cannabis patients (59% female) who completed at least two quarterly GAD-7 anxiety assessments over 12 months (2021-2023). Latent class growth analysis identified anxiety trajectories and their predictors.

What this study cannot tell us

No control group or randomization; patients self-selected into medical cannabis treatment. Cannot determine whether anxiety would have been worse without cannabis. Cannabis products, doses, and formulations varied across patients. Co-use of anxiety medications was noted but not fully characterized.

How to read the evidence

Longitudinal design with validated anxiety measure and latent class analysis provides moderate evidence, limited by absence of a control group and variable treatment regimens.

When this study was published

2025 publication; data collected 2021-2023

The bigger picture

The rapidly expanding medical cannabis market relies heavily on anxiety as a qualifying condition. This study suggests that while less anxious patients may see modest benefits, the patients who arguably need help the most -- those with severe anxiety -- are not getting meaningful relief from medical cannabis alone.

Questions still open

Should medical cannabis programs provide mandatory mental health follow-up for patients with severe anxiety? Are specific cannabis formulations (high CBD, low THC) more effective for severe anxiety than others?

Common questions

Does medical cannabis help with anxiety?

This study found that patients with minimal or moderate anxiety showed slight improvements, but these were not clinically meaningful. Patients with severe anxiety showed no improvement over 12 months of medical cannabis use.

Who was most likely to have severe anxiety?

Younger patients, women, those with lifetime PTSD, and those who listed anxiety as their primary qualifying condition were most likely to fall in the Severe anxiety group that did not improve with medical cannabis use.

Read the original research

Assessing anxiety longitudinally among medical cannabis patients in Pennsylvania.

The American journal of drug and alcohol abuse, 51(6), 750-760

View on PubMed ↗Publisher’s paper ↗

Citation

Ataiants, Janna; Fedorova, Ekaterina V; Cocchiaro, Benjamin; Kleber, Lyric; Sayyid, Abdallah; Ardeleanu, Katherine; Lankenau, Stephen E. (2025). Assessing anxiety longitudinally among medical cannabis patients in Pennsylvania.. The American journal of drug and alcohol abuse, 51(6), 750-760. https://doi.org/10.1080/00952990.2025.2588273