By Dr. Benjamin Caplan, MD | Board-Certified Family Physician, CMO at CED Clinic | Evidence Watch
Want to apply this research to your care?
CED Clinic translates emerging research into individualized clinical care. Dr. Caplan has treated 30,000+ patients.
Book a consultation →A large Canadian study of over 54,000 medical cannabis users found a modest 15% higher rate of heart failure-related emergency visits and hospitalizations compared to the general population. However, the observational design and the fact that patients sought cannabis for conditions that themselves raise cardiovascular risk mean this finding cannot establish that cannabis caused the increased risk.
Medical Cannabis Authorization Linked to Modestly Higher Heart Failure Risk in Large Canadian Cohort
A retrospective cohort study of more than 215,000 Ontario residents finds that adults authorized to use medical cannabis had a statistically significant but clinically modest association with heart failure-related emergency department visits and hospitalizations, though the observational design, residual confounding, and the health profiles of cannabis-seeking patients all prevent any causal interpretation of these results.
#72
Strong Clinical Relevance
One of the largest studies to examine heart failure risk specifically among medically authorized cannabis users, directly informing clinical counseling even though causal conclusions remain out of reach.
Heart Failure
Cardiovascular Safety
Pharmacoepidemiology
Observational Evidence
Heart failure remains one of the leading causes of hospitalization in North America, and cannabis is now one of the most commonly authorized adjunctive therapies in countries with legal medical programs. The intersection of these two realities creates urgent questions for clinicians counseling patients who already carry cardiovascular risk. Existing evidence on cannabis and heart failure has drawn primarily from populations with cannabis use disorder or recreational users, leaving a critical gap regarding medically supervised patients. Understanding whether routine medical cannabis authorization itself alters heart failure trajectory is essential for informed prescribing and patient safety conversations.
This retrospective cohort study linked electronic medical records from Ontario cannabis clinics with provincial health administrative databases covering roughly 14 million residents. Investigators identified 54,006 adults who received medical cannabis authorizations between 2014 and 2019 and matched them 3:1 with 161,265 general population controls who had no cannabis authorization. The primary clinical question was whether authorized medical cannabis use was associated with incident heart failure events, defined as emergency department visits or hospitalizations with heart failure as the primary diagnosis. Propensity-score inverse probability of treatment weighting was used to balance dozens of measured baseline covariates between the two groups. The study cohort was notably younger and more female than typical heart failure populations, with 39% of subjects aged 50 or younger and 55% female, reflecting the demographics of the medical cannabis patient population in Ontario.
The primary outcome showed a hazard ratio of 1.15 (95% CI 1.06 to 1.25), corresponding to an incidence rate of 5.87 per 1,000 person-years among cannabis-authorized patients versus 5.14 per 1,000 person-years in controls. A secondary outcome that also included outpatient heart failure physician consultations yielded a similar hazard ratio of 1.13 (95% CI 1.08 to 1.19). While consistent across both definitions, the absolute risk difference was modest: approximately 0.73 extra events per 1,000 person-years. Critically, patients sought cannabis authorizations for chronic pain, sleep disorders, and mental health conditions, all of which independently elevate cardiovascular risk and represent a significant source of confounding by indication. The authors appropriately characterize these results as hypothesis-generating and emphasize that unmeasured confounders, including actual cannabis consumption patterns, dose, product type, THC-to-CBD ratio, and lifestyle factors, could not be addressed within this design.
This study does something genuinely valuable: it looks at real patients in a real medical cannabis program rather than relying on self-reported recreational use or diagnostic codes for cannabis use disorder. That matters, because the population it captures looks a lot more like the patients I actually see. The problem is that the very conditions driving these patients to seek cannabis, including chronic pain, insomnia, anxiety, and depression, are themselves strongly associated with cardiovascular disease. A hazard ratio of 1.15 is exactly the kind of signal that residual confounding can explain in its entirety, and we should be honest about that rather than treating this as proof that cannabis harms the heart.
In my own practice, I already discuss cardiovascular considerations with patients who have existing risk factors before initiating cannabis therapy, and this study reinforces that conversation without fundamentally changing it. I pay close attention to product selection, favoring lower-THC formulations and monitored dosing in patients with known cardiac histories. I would not withdraw cannabis from a patient who is benefiting from it on the basis of this study alone, but I would use it as one more reason to ensure we are tracking blood pressure, heart rate, and symptoms over time.
This study sits at an important but early position in the research arc connecting medical cannabis to cardiovascular outcomes. Most prior data have come from studies of cannabis use disorder or from cohorts where cannabis exposure was defined by self-report, both of which introduce different forms of measurement error and selection bias. By anchoring exposure to a formal clinical authorization, the investigators create a cleaner exposure definition, though the gap between authorization and actual use remains a significant limitation. Clinicians should understand that this work is hypothesis-generating: it identifies a signal worth investigating further but does not provide the kind of evidence that should prompt changes to prescribing guidelines or withdrawal of therapy from patients already benefiting.
From a pharmacological perspective, THC is known to activate sympathetic tone, increase resting heart rate, and modulate blood pressure in ways that could theoretically stress a vulnerable myocardium. However, CBD has distinct and sometimes opposing hemodynamic properties, and the study was unable to capture which cannabinoids patients actually consumed. This makes it impossible to attribute the observed signal to any particular molecular mechanism. For practitioners managing patients with cardiovascular comorbidities, the most actionable step is to ensure routine monitoring of blood pressure, heart rate, and functional capacity in cannabis-authorized patients, and to document the indication, product type
