Cannabis Cardiovascular Risk: Legalization Study and USMLE

A new analysis of more than 50 million U.S. hospital discharge records, published in BMJ Open, links the legalization of recreational cannabis to faster growth in hospital admissions for cardiovascular disease. Admissions for atrial fibrillation, heart failure and myocardial infarction rose faster in the Western U.S., where legalization came early, than in regions without it. The study is observational, so it shows an association and cannot prove that cannabis caused the increase.

What happened

The researchers used the National Inpatient Sample (NIS) for 2012–2018. NIS participating states cover about 97% of the U.S. population. In total, they analyzed 50,042,564 hospital discharge records.

They focused on the Western region: Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington and Wyoming. This region began legalizing recreational cannabis in January 2014, nearly five years before other regions. By 2018, over 98% of people in the West lived in, or bordered, a state with legal cannabis sales.

The team looked at admissions for six diagnoses:

• acute heart failure
• heart attack
• atrial fibrillation/flutter
• ischemic stroke
• venous thromboembolism (VTE)
• ventricular tachycardia

For each, they estimated baseline regional admission rates per 100,000 people from weighted NIS data for the two years before the first legal recreational sale. They then compared the trends before and after legalization in the West with trends in regions that had not enacted such laws.

Before legalization (2012–13), admission rates for these diagnoses were lower in the West than elsewhere, but they rose faster afterward. After adjustment, the estimated change in admission rates for each year of legal cannabis sales was statistically significant for three conditions:

• atrial fibrillation: just over 2% higher
• heart failure: nearly 4% higher
• heart attack: just over 3% higher

Rates for the other outcomes (ischemic stroke, VTE, ventricular tachycardia) were also higher, but those increases were not statistically significant.

Patients admitted in the West were younger on average (51 vs. 55 years), more often lived in higher-income ZIP codes (26% vs. 19%) and less often identified as white.

Why it matters

Legal access to cannabis keeps expanding. As of June 2025, medicinal use was legal in 41 U.S. states and recreational use in 24. The authors note that legalization has been associated with more widespread use, particularly among older adults, who already carry a higher baseline cardiovascular risk.

Earlier research has examined legalization in relation to psychiatric conditions, substance use, emergency visits and traffic accidents. According to the authors, this is the first study to examine its association with cardiovascular hospitalizations. They call for a cautious approach to legalization, paired with data collection on cardiovascular outcomes.

For a future physician, this is also a useful case of how to read a policy-level study. The exposure is a law, not a measured dose of cannabis in individual patients.

What to know for the USMLE — high-yield points

• Cannabinoid pharmacology: THC acts mainly on CB1 receptors (G-protein-coupled, Gi/o), which are abundant in the CNS and are also found in the cardiovascular system.
• Classic acute hemodynamic effects: Acute use is associated with tachycardia and increased sympathetic tone with reduced vagal tone. Supine blood pressure may rise, and orthostatic hypotension can occur.
• Myocardial oxygen supply–demand mismatch: A faster heart rate raises oxygen demand. If cannabis is smoked, carboxyhemoglobin can reduce oxygen delivery. Together these can provoke myocardial ischemia or MI, even in young patients without obvious risk factors.
• Arrhythmias: Sympathetic activation and autonomic shifts can trigger atrial fibrillation and, rarely, ventricular arrhythmias. This is a plausible mechanism, but this study did not test it.
• Heart failure: Tachycardia, increased afterload and ischemia can worsen decompensation in patients with existing disease. Again, this is a biological rationale, not a finding of the study.
• Study design: difference-in-differences (DiD). This method compares the change over time in an exposed group (the West) with the change in an unexposed group (other regions). It controls for stable baseline differences between groups. Its key assumption is parallel trends: without the intervention, both groups would have followed similar trajectories.
• Association vs. causation: Observational studies can show association. For causality, think of the Bradford Hill criteria: temporality, strength, consistency, dose-response, biological plausibility and others. Randomized trials remain the gold standard for causal inference.
• Ecological-level exposure: Here the exposure is a regional law, not individual cannabis use. Inferring individual-level risk from population-level data risks the ecological fallacy.
• Confounding and secondary trends: Simultaneous changes in other risk factors, coding practices, access to care or illicit use can bias results. Know how each could mimic or mask an effect.
• Statistical significance vs. effect size: Not significant does not mean no effect. Stroke, VTE and ventricular tachycardia rates were higher, but the study did not show these increases to be statistically significant.

Limitations

The authors and the design set clear boundaries on how far these findings can be pushed.

• Observational design. The study cannot establish cause and effect. The authors say so explicitly.
• No state-level resolution. The researchers could not separate individual states, so the analysis treats the West as one region.
• No data on route of use. They could not look at the method of use (smoked or not), which matters mechanistically for exposures such as carbon monoxide.
• Illegal use not captured. They could not account for trends in illegal use.
• Exposure is a policy, not individual use. The data show hospital admissions in a region, not which patients used cannabis.
• Differences between regions. Patients in the West were younger, lived in higher-income ZIP codes and less often identified as white. The analysis adjusted for potentially influential factors, but residual differences cannot be excluded.
• Time window. The data end in 2018, so the study says nothing about later years.

FAQ

Does this study prove that cannabis legalization causes heart attacks?

No. It is an observational difference-in-differences analysis, and the authors state that it cannot establish cause and effect. It shows that admissions for heart attack, heart failure and atrial fibrillation rose faster in the Western region after legalization than in comparison regions.

Which cardiovascular conditions showed a significant change?

Atrial fibrillation (just over 2% higher per year of legal sales), heart failure (nearly 4%) and heart attack (just over 3%). Ischemic stroke, VTE and ventricular tachycardia also trended higher, but those increases were not statistically significant.

Why is this relevant for USMLE Step 1 and Step 2 CK?

It touches two testable areas. One is the cardiovascular pharmacology and pathophysiology of THC, including tachycardia and ischemia. The other is biostatistics and epidemiology: reading study designs, understanding association vs. causation, and recognizing confounding and the ecological fallacy.

Key takeaways

• In an analysis of more than 50 million U.S. discharge records (2012–2018), recreational cannabis legalization in the Western region was associated with faster growth in admissions for atrial fibrillation, heart failure and heart attack.
• The finding is an association from observational, region-level data. It does not prove causation, and key details such as individual use, route of use, and state-level differences were not available.
• For the USMLE, connect the CB1-mediated sympathetic and ischemic effects of THC with the logic of difference-in-differences and the limits of causal inference.