Cannabis and the heart: three studies – and a finding that does not fit the pattern

Within four days, three research teams report the same thing: cannabis is linked to cardiovascular disease. The headlines agreed quickly. The most interesting number of the week appears in none of them, and it does not fit the simple story of "more is worse".

Study one: 442,906 people and a surprising kink

At the American Heart Association's Scientific Sessions, running from 6 to 9 November in Chicago, a team from Tulane University in New Orleans presents an analysis of telephone surveys from the Behavioral Risk Factor Surveillance System. The basis: 442,906 adults in the US, years 2020 to 2024, average age 49, roughly half of them women.

The results, compared with people who reported no cannabis use:

• Those who used on 1 to 5 days a month had 23 per cent higher odds of reporting cardiovascular disease.
• Those who used on 16 to 30 days were at 21 per cent.
• For stroke: 40 per cent with low-frequency use, 25 per cent with heavy use.
• For heart attack, the team found no statistically significant association with frequency of use.

The authors say themselves that they were surprised. They had expected risk to rise steadily with frequency. "Both low-use and heavy-use groups had higher odds," study author Aabha Divya said.

That kink is exactly what matters for reading the study, and it is usually left out. There are at least three explanations that have nothing to do with a dose-response curve:

• Reverse causation. People who have had a heart condition often stop or cut down. The group with no current use therefore includes former users, possibly with exactly the diagnoses at issue.
• The survey is cross-sectional. Use and diagnosis were recorded at the same moment, so no temporal order can be derived.
• Self-report. Recall error on both use and diagnosis, and no data on potency, amount or method of use.

One point often raised as reassurance does not hold here either: the association appeared regardless of form, including edibles and vapes, not only smoking. That argues against the idea that this is only about combustion products.

And there is a formal limit the American Heart Association states itself: this is an abstract, not a peer-reviewed publication. The findings count as preliminary until they appear as a manuscript in a journal.

Study two: 50 million hospital discharges

Differently built is the paper published in BMJ Open on 22 October. The team used the National Inpatient Sample, a sample of hospital discharges, and compared regions where recreational use was legalised with regions where it was not, using a difference-in-differences design. Period: 2012 to 2018.

Result: after legalisation, hospital admissions for cardiovascular diagnoses rose faster, including atrial fibrillation, heart failure and heart attack. Methodologically the work is sturdier than a survey because it carries control diagnoses: one positive control (cannabis-related diagnoses, which had to rise if the effect is real) and two negative controls (opioid-related diagnoses and kidney stones, which should not rise).

Two things remain. It is an observational study, so causation is not proven. And the period ends in 2018. For the German debate that means: the study says nothing about the consequences of the 2024 cannabis act. Anyone citing it for that overstretches it.

Study three: 100 million people and a genetic detour

The third paper is the broadest and the only one that supports a causal suspicion. A team at the University of Cambridge pooled cohort studies with more than 100 million participants in a meta-analysis, published in the International Journal of Stroke.

• Cannabis use was associated with a 37 per cent higher stroke risk.
• Cocaine with 96 per cent, amphetamines with 122 per cent.
• Among people under 55, the increase for cannabis was 14 per cent, for amphetamines 174 per cent.

The additional step is Mendelian randomisation: the team tested whether genetic variants associated with cannabis use disorders are also associated with stroke. The result supports the possibility of a causal relationship, particularly for large artery stroke. Among the mechanisms the authors name is increased clotting, said to be especially pronounced with cannabis.

This is where the coverage deserves a look. Several German-language articles named the 37 per cent and a "nearly threefold" increase among younger users in the same sentence. The threefold increase concerned amphetamines, not cannabis. That is not a detail; it is the difference between a study and a misunderstanding.

What the three add up to

Single observational studies are weak. Three papers with different designs pointing the same way are considerably stronger: a cross-sectional survey, a regional difference-in-differences analysis with control diagnoses, and a meta-analysis with a genetic add-on. That is why specialists are paying attention, even though none of the three proves anything on its own.

The comment from Robert Page of the University of Colorado, who was not involved, captures the situation: this is the second observational study to suggest that cannabis exposure without tobacco smoking may be a risk factor. Until now, cannabis was treated as an innocent bystander in people who also smoke tobacco.

What is missing in Europe

What stands out is how little of this debate reaches Europe. The European discussion is dominated by potency limits, THC thresholds, pilot projects and market rules. Cardiovascular risks barely feature, although stroke is among the most common causes of death worldwide.

That is a gap with consequences. Anyone arguing about THC ceilings or the approval of stronger products is also deciding about exposure. Anyone not discussing heart risks cannot make that decision in an informed way. In the US, the cardiovascular question is now part of the regulatory debate; in the EU it is not.

Open questions

• Why is the curve not monotonic? That question is currently open and the most interesting research task of the three papers.
• How large is the effect for the concentrates that are growing in Europe? None of the studies separates by THC content.
• What does this mean for people with known heart disease? The data suggest a risk group but say nothing about individual decisions.
• And: why does the EU discuss potency but not cardiovascular consequences?

Facts and interpretation: The figures from the Tulane analysis come from the American Heart Association release of 2 November 2026 on abstract Mo3041; it is not peer-reviewed and the findings are preliminary. The BMJ Open study appeared on 22 October 2026 (DOI 10.1136/bmjopen-2025-111050) and is observational, without proof of causation. The Cambridge meta-analysis by Megan Ritson, Hugh Markus and Eric Harshfield appeared in 2026 in the International Journal of Stroke (DOI 10.1177/17474930261418926). Explaining the non-monotonic pattern through reverse causation and the healthy-smoker effect is an interpretation the authors themselves raise as a possibility, not a result. This article is not medical advice and contains no recommendation to use. Anyone using cannabis with questions about heart health should discuss it with a doctor.

Sources: American Heart Association, release of 2 November 2026; BMJ Open, Cannabis legalisation and cardiovascular hospitalisations; University of Cambridge on the meta-analysis. As of 9 November 2026.