Study links cannabis use to nearly fourfold higher incidence of type 2 diabetes over five years

Adults with cannabis-related medical records were nearly four times as likely to be diagnosed with type 2 diabetes over a five-year period compared with people who had no record of drug use or chronic illness, researchers reported.

The team, led by Dr. Ibrahim Kamel of Boston Medical Center, analyzed electronic health records from 54 health care organizations in the United States and Europe using the TriNetX Research Network. They compared roughly 96,800 people who had cannabis-related diagnoses — a group that ranged from casual users to those with dependency — with more than 4.16 million individuals with no documented drug use or chronic conditions. Over five years, 2.2% of the cannabis-exposed group developed type 2 diabetes versus 0.6% of the control group. The increased incidence persisted after adjusting for other cardiometabolic risks, including high blood pressure, high cholesterol, heart disease and the use of alcohol or cocaine.

The researchers said possible explanations for the association include cannabis effects on appetite, metabolism and insulin sensitivity, and the tendency for cannabis use to coincide with poor dietary habits that can elevate blood sugar over time. "As cannabis becomes more widely available and socially accepted and legalized in various jurisdictions, it is essential to understand its potential health risks," Kamel said in a statement. He and colleagues recommended that clinicians discuss cannabis use with patients so they can assess overall diabetes risk and consider metabolic monitoring where appropriate.

The study has notable limitations that the authors acknowledged. It used a retrospective design, relying on previously recorded clinical data rather than prospectively following people from a baseline without disease, which limits the ability to infer causation. Cannabis use was identified through clinical diagnoses and self-report entries in medical records, so the analysis could not measure frequency, potency, mode of consumption or precise dose, and it could not distinguish between occasional and daily users. The team also noted the potential for selection or reporting biases inherent in health-record data.

Despite those caveats, the investigators said the real-world findings warrant attention as cannabis legalization and use increase in many jurisdictions. They argued the results support integrating diabetes-risk awareness into substance-use disorder treatment and routine clinical counseling. The authors called for prospective studies that can better measure use patterns, timing and biological markers to clarify whether the observed association reflects a causal effect of cannabis on metabolic health or is driven by unmeasured confounders.

Public-health experts and clinicians not involved in the study have generally emphasized the need for more detailed, longitudinal research before altering clinical guidelines. They noted that metabolic consequences of any substance can depend on dose, duration of use and interactions with other health behaviors. Until more definitive evidence is available, the researchers advised that clinicians consider asking patients about cannabis use as part of routine risk assessment for diabetes and other metabolic conditions.

The study adds to a growing body of literature examining long-term health outcomes associated with cannabis. Researchers said that clarifying mechanisms — whether behavioral, such as diet and activity changes, or biological, such as effects on insulin signaling — will be important for informing clinical practice and public messaging as cannabis products become more accessible.

Further peer-reviewed publications and prospective cohort studies will be needed to determine causality, to quantify risk by use pattern and to assess whether changes in consumption lead to changes in diabetes risk.