Does Marijuana Affect Diabetes and Blood Sugar?

Marijuana does affect blood sugar and diabetes risk, but not in the straightforward way most people expect. Large surveys consistently find that cannabis users have lower rates of type 2 diabetes and lower fasting insulin levels than nonusers, yet people with type 1 diabetes who use cannabis tend to have worse blood sugar control and face a higher risk of dangerous complications. The relationship between cannabis and metabolism turns out to depend heavily on the type of diabetes, how often someone uses, and which compounds in the plant are doing the work.

What the Surveys Actually Show

If you look only at population-level data, cannabis seems almost protective against type 2 diabetes. A 2023 meta-analysis pooling seven studies found that people exposed to cannabis had roughly half the odds of developing type 2 diabetes compared to those who had never used it.¹ An earlier cross-sectional study using U.S. national health data found even steeper numbers, with marijuana users showing substantially lower age-adjusted odds of diabetes after controlling for other health and lifestyle factors.²

These findings are striking, but they come with a major caveat. A Swedish study following 18,000 people over time initially found a similar inverse link between cannabis use and type 2 diabetes. However, once the researchers adjusted for age, smoking status, body mass index, and occupation, the apparent protection vanished almost entirely.³ The takeaway is that cannabis users in these studies tend to be younger and leaner, and those demographic differences alone may explain most of the statistical gap. None of the existing large studies can prove cannabis itself prevents diabetes.

Fasting Insulin and Insulin Resistance

Even if the diabetes-prevention story is overblown, something interesting is happening with insulin. A well-known analysis of U.S. adults found that current marijuana users had about 16% lower fasting insulin levels and 17% lower insulin resistance scores compared to people who had never used.⁴ That effect was specific to current users and faded in people who had quit, suggesting it is tied to active cannabis exposure rather than some lasting change.

The picture gets more nuanced when you look at body weight. In obese adults with moderate insulin resistance, heavy lifetime marijuana use was linked to roughly a 12% drop in fasting insulin and a 10% drop in insulin resistance. But in lean adults, no such association appeared.⁵ A separate study confirmed this pattern, finding that the link between marijuana use and lower insulin resistance held in obese adults but not in people who were already at a healthy weight.⁶ If cannabis does affect insulin sensitivity directly, the benefit seems to show up mainly in people whose metabolism is already under stress.

The Metabolic Paradox

One of the stranger findings in cannabis research is that regular users tend to have lower body mass indexes than nonusers, despite eating more. Across multiple studies, cannabis users consumed an average of several hundred extra calories per day compared to nonusers, yet their BMI was the same or lower.⁷ This implies either an increase in metabolic rate, changes in how the body partitions nutrients, or some combination of both. Since obesity is the single biggest modifiable risk factor for type 2 diabetes, the fact that regular cannabis users tend to stay leaner could explain part of the lower diabetes rates seen in surveys, without requiring any direct effect on blood sugar.

Researchers have proposed that repeated cannabis exposure may “downregulate” the body’s cannabinoid receptors over time, effectively dampening a system that normally promotes fat storage and appetite. That hypothesis fits the data, but it remains a hypothesis. The paradox is real and reproducible; the explanation for it is still being worked out.

How Cannabis Interacts With Insulin at the Cellular Level

The biological plausibility for a cannabis-metabolism link is solid. Pancreatic beta cells, the ones that make insulin, carry a full set of cannabinoid receptors and produce their own endocannabinoids to regulate insulin release in an autocrine fashion.⁸ When cannabinoid receptor 1 (CB1) on these cells is activated, it triggers a chain of signals that reorganizes the cell’s internal scaffolding and promotes the release of insulin-containing vesicles.⁹ The response is also glucose-dependent: endocannabinoid receptor activation boosts insulin secretion at low glucose levels but does not push it further when glucose is already high.¹⁰

That sounds protective, but it is only half the story. An overactive endocannabinoid system also promotes fat storage, disrupts lipid metabolism, and can trigger programmed cell death in the very beta cells that produce insulin. Chronic overactivation of the system may even promote inflammation in pancreatic islets, setting the stage for diabetes rather than preventing it.¹¹ The dose, duration, and specific cannabinoids involved likely determine which direction things tip.

Type 1 Diabetes and the DKA Risk

For people living with type 1 diabetes, the evidence runs in a different and more concerning direction. A rapid review of the available studies found that five separate investigations reported higher HbA1c, the standard measure of long-term blood sugar control, in type 1 patients who used cannabis. In one study, cannabis users with type 1 diabetes had an HbA1c that was about 0.4 percentage points higher than nonusers after adjusting for age, income, and how they delivered their insulin.¹² That gap may sound small, but in diabetes management a sustained 0.4-point rise in HbA1c is clinically meaningful.

Worse, cannabis use in adults with type 1 diabetes was linked to roughly a two-and-a-half-fold increase in the risk of diabetic ketoacidosis (DKA), a life-threatening emergency in which blood acids build up when the body runs out of insulin.¹³ Part of the explanation may be that cannabinoids alter gut motility and can cause severe nausea and vomiting, which in turn leads to missed meals, skipped insulin doses, and dehydration, all of which accelerate the slide into DKA.¹⁴ It is also possible that the altered state of mind from being high simply makes some people less attentive to their blood sugar monitoring and insulin routines.

Diabetic Complications

Beyond blood sugar numbers, cannabis use may worsen some of the downstream complications that make diabetes so damaging over time. A Mendelian randomization study, which uses genetic variants to approximate a randomized trial, found that a genetic predisposition toward cannabis use disorder was associated with higher risks of both diabetic retinopathy (the eye damage caused by diabetes) and diabetic hypoglycemia. There was also suggestive evidence linking it to diabetic kidney disease.¹⁵ A separate study looking at patients with type 2 diabetes and diagnosed cannabis use disorder also found elevated retinopathy risk, alongside opioid and cocaine use disorders.¹⁶

These studies are talking about disordered use, not occasional consumption, which is an important distinction. Someone who uses cannabis once a month is in a very different category from someone meeting the diagnostic criteria for cannabis use disorder. But the pattern is consistent enough that anyone with diabetes who uses cannabis regularly should be having honest conversations with their eye doctor and endocrinologist.

There is at least one bright spot in the complications picture. A small randomized, placebo-controlled trial tested inhaled cannabis in people with painful diabetic neuropathy, the burning, tingling nerve pain that often develops in the feet and hands. The study found a clear dose-dependent reduction in pain scores, with the highest dose producing significant relief for both spontaneous pain and pain triggered by touch.¹⁷ The trial was small, just 16 patients, and short-term, so it is far from the final word. But for people who have exhausted standard neuropathy treatments, it is a finding worth knowing about.

CBD, THCV, and Other Cannabinoids

Most of the research discussed so far involves whole-plant cannabis or THC specifically. But the plant contains over a hundred different cannabinoids, and some of the non-psychoactive ones are generating their own lines of research in diabetes.

CBD, the most commercially popular cannabinoid after THC, has shown anti-inflammatory effects in animal models of type 1 diabetes. In mice genetically prone to developing type 1 diabetes, CBD treatment delayed the onset of disease and reduced markers of inflammation in the pancreatic microcirculation.¹⁸ Whether this translates to humans is unknown. Type 1 diabetes in mice and type 1 diabetes in people share some features, but mouse models have a long history of producing exciting results that fail to replicate in human trials.

THCV (tetrahydrocannabivarin), a less well-known cannabinoid found in small amounts in certain cannabis strains, has attracted attention for its effects on metabolism. A trial involving 62 people with type 2 diabetes found that THCV lowered fasting plasma glucose and improved pancreatic beta-cell function.¹⁹ THCV is unusual because it may act as a partial blocker of CB1 receptors at low doses rather than activating them the way THC does, which could explain why its metabolic effects differ from those of THC.

Even beyond classic cannabinoids, beta-caryophyllene, a terpene found in cannabis as well as in black pepper and cloves, selectively activates the CB2 receptor and has shown effects on glucose and lipid metabolism in preclinical studies. It also activates certain receptors that are already targets of existing diabetes and cholesterol drugs.²⁰ The research is still early, but the diversity of biologically active compounds in cannabis means that asking “does marijuana affect diabetes” is a bit like asking “does food affect weight.” Which marijuana, which compounds, and in what proportions all matter enormously.

Cannabis During Pregnancy and Gestational Diabetes

A question that comes up less often but matters a great deal is whether cannabis use during pregnancy affects gestational diabetes risk. A large Canadian population-based study found that prenatal cannabis use was associated with higher odds of gestational diabetes. The same study also linked cannabis use in pregnancy to substantially increased risks of preterm birth, low birth weight, and congenital anomalies.²¹ This is an observational study, so it cannot rule out that the kinds of people who use cannabis during pregnancy may also have other health or socioeconomic factors at play. But the sheer breadth of adverse outcomes associated with prenatal cannabis use in this study makes a strong case for avoidance.

Drug Interactions With Diabetes Medications

If you use cannabis and also take medications for diabetes or related conditions, the pharmacokinetic interactions deserve attention. THC and CBD are both metabolized by the same liver enzyme systems that process many common drugs, including insulin analogues, metformin, statins, beta-blockers, and blood thinners like warfarin.²² CBD in particular is a potent inhibitor of certain liver enzymes, which means it can slow down the breakdown of other medications and effectively raise their blood levels. For diabetes drugs with a narrow safety window, that could mean the difference between a therapeutic dose and a dose that causes dangerously low blood sugar.

This is not a theoretical concern. People who start a high-dose CBD oil alongside their existing diabetes regimen without telling their doctor are running an uncontrolled experiment on their own liver metabolism. If you are on insulin, a sulfonylurea, or any medication where blood sugar lows are a real danger, your provider needs to know about your cannabis use so they can monitor and adjust doses appropriately.

Why the Type 1 and Type 2 Stories Diverge

It is worth pausing on why cannabis appears to have a mostly neutral-to-favorable metabolic profile in population studies of type 2 diabetes but a mostly harmful one in type 1 diabetes. Type 2 is driven primarily by insulin resistance and excess body fat. Cannabis users tend to be leaner and show better insulin sensitivity, both of which work in favor of type 2 risk. Type 1, on the other hand, is an autoimmune disease in which the body destroys its own insulin-producing cells. You cannot solve a type 1 problem by being thinner or more insulin-sensitive. What you can do is lose control of a tightly managed insulin regimen by being intoxicated, nauseated, or inattentive, and that is exactly where cannabis appears to cause harm.

Self-management behavior probably explains more of the type 1 signal than any direct biochemical action. Cannabis can make you forget to bolus before a meal, delay treating a low, or simply not notice that your continuous glucose monitor has been alarming for the last hour. For someone whose blood sugar is managed by an external insulin supply and constant vigilance, those lapses carry real danger.

Peripheral Vascular and Kidney Risks

Beyond the eyes and nerves, the same rapid review that flagged higher HbA1c in type 1 users also found that people with type 2 diabetes who used cannabis recreationally had higher risks of peripheral arterial disease and heart attack. Worse kidney function was reported in two separate studies, one each for type 1 and type 2 diabetes.¹² Smoking anything, including cannabis, introduces carbon monoxide and particulate matter that damage blood vessel walls, which matters when diabetes has already compromised vascular integrity. Edibles and vaporizers remove some of that risk, but the cannabinoid-specific effects on blood vessels and kidneys may persist regardless of how the drug is consumed.

What “Not Enough Evidence” Really Means Here

Much of the cannabis-diabetes literature is observational, meaning researchers looked at people who already chose to use or not use cannabis and compared outcomes. That study design is notoriously bad at separating the effect of the substance itself from the dozens of lifestyle and demographic factors that differ between users and nonusers. The Swedish cohort study is a vivid example: the raw association between cannabis and lower diabetes risk almost entirely disappeared once age and BMI were accounted for.³

We have very few randomized controlled trials, the gold standard for proving cause and effect, testing cannabis compounds in people with diabetes. The neuropathy trial had 16 people. The THCV trial had 62. These are far too small to make population-wide recommendations. Until larger and longer trials are run, the honest scientific position is that cannabis plausibly affects glucose metabolism through real biological pathways, that observational data hints at lower type 2 risk and worse type 1 outcomes, and that no one should be using cannabis as a diabetes treatment based on the current evidence. The gap between “interesting signal” and “proven therapy” is wide, and we are firmly on the “interesting signal” side of it.