Chronic Marijuana Use Damages Blood Vessel Function Like Tobacco Smoking

Cannabis legalization has dramatically increased use in the US, yet its cardiovascular consequences remain poorly understood. To address this gap, researchers at UCSF launched the CANDIDE (CANnabis: Does It Damage Endothelium) study, a carefully designed cross-sectional investigation published in JAMA Cardiology.

The study enrolled 55 age-matched, otherwise healthy adults (mean age 31.3 years; 37% female) living in the San Francisco Bay Area between October 2021 and August 2024. Participants were divided into three groups: chronic marijuana smokers, THC-edible users, and nonusers. All participants were non-tobacco smokers, non-vapers, and not frequently exposed to secondhand smoke—key controls that isolate cannabis-specific effects.

The primary outcome was arterial flow-mediated dilation (FMD) of the brachial artery, measured via ultrasound—a validated, noninvasive measure of endothelial function. Secondary outcomes included carotid-femoral pulse wave velocity (PWV) as a measure of arterial stiffness, and a cellular assay in which human umbilical vein endothelial cells (HUVECs) were incubated with participant serum and stimulated with vascular endothelial growth factor (VEGF) to assess nitric oxide (NO) production capacity.

Both cannabis-using groups demonstrated significantly impaired FMD compared to nonusers (nonusers: mean 10.4%; marijuana smokers: 6.0%, P=.004; THC-edible users: 4.6%, P=.003). FMD was inversely correlated with smoking frequency (r=−0.7, P<.001) and amount of THC ingested (r=−0.7, P=.03), indicating a dose-dependent relationship. However, the mechanisms appear to diverge by route of administration: serum from marijuana smokers significantly blunted VEGF-stimulated NO production in HUVECs (1.1 vs. 1.5 nmol/L in nonusers; P=.004), whereas serum from THC-edible users had no such effect (1.5 nmol/L; P=.81). Pulse wave velocity and other vascular properties showed no significant differences between groups.

These findings carry important implications. Endothelial dysfunction is a well-established early marker of atherosclerosis and precedes overt cardiovascular disease by years or decades. The magnitude of FMD impairment seen in cannabis users was comparable to that previously reported in tobacco smokers. The divergence in cellular NO data suggests that combustion byproducts in marijuana smoke may contribute a specific, circulating toxic mechanism that is absent in edible use—yet edible THC still impairs vascular function through some other pathway, possibly direct cannabinoid receptor-mediated effects on vascular tone or systemic inflammation.

Caveats include the modest sample size (n=55), cross-sectional design precluding causal inference, and the fact that participants were drawn from one geographic region. Self-reported cannabis use data introduce potential measurement error, and long-term vascular outcomes were not tracked. Nevertheless, this is among the most rigorous human studies to date directly comparing cannabis delivery methods and their vascular effects.