What diabetes and metabolic syndrome are
Diabetes mellitus is a group of metabolic disorders defined by chronically elevated blood glucose (hyperglycaemia). Type 2 diabetes accounts for around 90% of UK cases and involves insulin resistance — cells become less responsive to insulin — combined with a progressive decline in insulin production by pancreatic beta cells. Type 1 diabetes is an autoimmune condition in which the immune system destroys beta cells, causing absolute insulin deficiency.
In the UK, approximately 4.4 million people live with a diabetes diagnosis, and an estimated 1.2 million have undiagnosed type 2 diabetes. Standard management includes lifestyle changes (diet, exercise, weight loss), oral glucose-lowering medications such as metformin, and, when needed, injectable therapies including GLP-1 receptor agonists and insulin.
Metabolic syndrome is a cluster of conditions that commonly occur together: central obesity, raised blood pressure, raised fasting glucose, raised triglycerides, and low HDL cholesterol. It increases the risk of type 2 diabetes, cardiovascular disease, and stroke.
Research on cannabis and metabolic health covers three distinct areas: population studies of diabetes risk, studies of glucose metabolism and insulin sensitivity, and early work on individual cannabinoids as possible metabolic treatments. These questions rely on different study designs and should not be treated as one body of evidence.
Cannabis research on Diabetes and metabolic health
Several papers address GPR119 — a receptor targeted for type 2 diabetes and obesity drug development — and PPAR receptors, which are the target of existing diabetes medications such as pioglitazone.
Epidemiological evidence: cannabis use and diabetes risk
A 2015 meta-analysis by Alshaarawy and Anthony pooled data from eight independent epidemiological samples (meta-analysis of epidemiological studies). It reported an inverse association between cannabis use and diabetes mellitus — meaning cannabis users had lower odds of diabetes in these datasets. This is an association, not a causal finding, and the studies could not rule out confounding: cannabis users may differ from non-users in ways that independently affect diabetes risk, including body weight, diet, physical activity, and healthcare access.
A 2021 two-sample Mendelian randomisation study by Baumeister and colleagues used genetic variants associated with cannabis use to test whether cannabis use causally affects type 2 diabetes risk (Mendelian randomisation study). The study found no evidence that cannabis use causally affects type 2 diabetes. Mendelian randomisation is less susceptible to confounding than observational epidemiology, but it depends on assumptions about the genetic instruments that may not fully hold.
A 2023 systematic review and meta-analysis by Mousavi and colleagues examined the epidemiological association between cannabis use and type 2 diabetes and concluded that the studies are discrepant — meaning they point in different directions (systematic review and meta-analysis). Some show an inverse association, others show no association. The review could not reach a firm conclusion.
A 2025 observational study by Bryan and colleagues in healthy adults found that cannabis use was associated with lower levels of peripheral inflammatory markers but similar insulin sensitivity compared with non-use (cross-sectional study). Lower inflammation is of interest because chronic low-grade inflammation contributes to insulin resistance, but the study did not find that cannabis use was associated with better insulin sensitivity.
A 2020 cohort study by Barre and colleagues in patients with chronic hepatitis C infection found that cannabis use was associated with a lower risk of diabetes in this specific population (prospective cohort study). Cohort-specific findings in one disease population do not generalise to the broader population.
Clinical trial evidence: cannabinoids and metabolic outcomes
A 2016 randomised, double-blind, placebo-controlled pilot trial by Jadoon and colleagues tested CBD (100 mg twice daily) and THCV (5 mg twice daily) in 62 patients with type 2 diabetes over 13 weeks (human RCT). Neither CBD nor THCV significantly changed the primary endpoint of HDL-cholesterol compared with placebo. THCV produced a small but statistically significant reduction in fasting plasma glucose (estimated treatment difference -1.1 mmol/L, P=0.03) and improved pancreatic beta-cell function as measured by HOMA2-%B. There were no significant differences in body weight, liver triglyceride content, or insulin sensitivity between THCV and placebo. This was a single small pilot trial. It does not establish that THCV or CBD is an effective diabetes treatment. The trial is also described on the MCPH THCV entity page.
No additional human RCT of a cannabinoid for diabetes or metabolic outcomes was identified in the published literature.
Preclinical receptor and animal studies
- Two papers describe GPR119, a G protein-coupled receptor expressed in pancreatic beta cells and the gut, as a target for type 2 diabetes and obesity drug development (review articles). Endogenous ligands for GPR119 include oleoylethanolamide (OEA), a fatty acid amide related to endocannabinoids. Pharmaceutical GPR119 agonists have been studied as potential diabetes drugs. These are pharmacology reviews, not cannabis studies.
- One animal study reported that PPAR-alpha activation improves insulin sensitivity and lowers muscle lipids in high-fat-fed rats (animal study, 2001). PPAR-alpha is the target of fibrate drugs used to lower triglycerides, not a cannabinoid receptor, though some cannabinoids and endocannabinoid-related compounds interact with PPAR receptors.
- One study reports downregulation of CB1 and CB2 receptors in the bladder of insulin-resistant obese Zucker rats (animal study). This is a receptor-expression finding in a specific tissue; it does not establish a metabolic effect of cannabinoids.
- One small study examined cerebral glucose metabolism and dopamine D2/D3 receptor availability in six young adults with cannabis dependence, reporting changes in regional brain glucose metabolism after abstinence (neuroimaging study). This is a brain-function study in a small sample of dependent users, not a study of diabetes or systemic glucose metabolism.
Compounds studied for Diabetes and metabolic health
THCV has the most direct human metabolic evidence of any cannabinoid. In the Jadoon 2016 RCT in 62 patients with type 2 diabetes, THCV (5 mg twice daily for 13 weeks) reduced fasting plasma glucose by an estimated -1.1 mmol/L relative to placebo (P=0.03) and improved beta-cell function. The primary endpoint of HDL-cholesterol was not met. No significant effect was seen on body weight, liver fat, or insulin sensitivity. This is one small pilot trial; it is not licensing-level evidence. Preclinical studies in mice report that THCV reduces body weight and food intake and improves glucose tolerance, but these are animal findings that have not been replicated in the human trial.
CBD was tested in the same Jadoon 2016 RCT and did not significantly change any metabolic endpoint compared with placebo at 100 mg twice daily. Preclinical studies report anti-inflammatory and metabolic effects of CBD in animal models of diabetes, but these have not been confirmed in human metabolic trials.
THC has not been tested in a human RCT for diabetes or metabolic outcomes. CB1 receptor activation by THC is known to stimulate appetite acutely, but the long-term metabolic consequences of THC use are not established in controlled trials.
No human or animal study of CBG for diabetes or metabolic health was identified.
Beta-caryophyllene is a CB2 receptor agonist. CB2 activation has been linked to reduced inflammation in animal models, and chronic inflammation contributes to insulin resistance. No human trial has tested beta-caryophyllene for diabetes or metabolic outcomes.
Limonene has been studied in animal models for effects on lipid metabolism. No human metabolic trial exists.
Geraniol has limited preclinical data on glucose metabolism. No human metabolic trial exists.
Limits of the evidence for Diabetes and metabolic health
What the evidence establishes
Multiple large epidemiological studies report an inverse association between cannabis use and diabetes — cannabis users in these datasets had lower odds of having diabetes. This is a consistent observational finding.
A Mendelian randomisation study did not find evidence of a causal effect of cannabis use on type 2 diabetes risk, which weakens the case that the epidemiological association reflects a causal protective effect.
One small human RCT (Jadoon 2016, n=62) tested CBD and THCV in type 2 diabetes. THCV reduced fasting glucose and improved beta-cell function; CBD showed no significant effects. The primary endpoint (HDL-cholesterol) was not met by either compound.
The GPR119 receptor system, which responds to endocannabinoid-related lipids, is a legitimate pharmacological target for type 2 diabetes drug development, though this is pharmaceutical chemistry, not cannabis research.
What the evidence does not establish
That cannabis improves diabetes outcomes. The epidemiological association is not supported by the Mendelian randomisation evidence, and the single human RCT was a small pilot that did not meet its primary endpoint.
That THCV is an effective diabetes treatment. One small trial found a reduction in fasting glucose but did not meet the primary endpoint. This is hypothesis-generating, not confirmatory.
That any cannabis product, cannabinoid ratio, or terpene profile improves metabolic health in humans. No data exist to support such a claim.
That cannabis use is safe for people with diabetes. Cannabis can affect appetite, food intake, and perception, which may have implications for glycaemic control. A 2020 cross-sectional study noted that cannabis use can alter perception of, and adherence to, treatment in insulin-dependent diabetes.
The long-term metabolic effects of regular cannabinoid use — on weight, insulin sensitivity, beta-cell function, and microvascular and macrovascular diabetic complications — are unknown.
Related conditions
- Chronic pain (NICE context) ; the evidence behind NICE guideline NG144
- Sleep disorders ; negative CBN trials and modest THC findings
- Stroke ; cannabis, cerebrovascular risk, and post-stroke recovery research
- Alzheimer’s and dementia ; preclinical neuroprotection research
Read next
- THCV ; studied in one human trial for type 2 diabetes
- CBD ; examined in preclinical metabolic models
- Browse the cannabinoid reference
- Browse the terpene reference
Sources
1. Alshaarawy O, Anthony JC. Cannabis smoking and diabetes mellitus: results from meta-analysis with eight independent replication samples. Epidemiology. 2015;26(4):586-594. doi:10.1097/EDE.0000000000000314. PMID: 25978793. Study record
2. Baumeister SE, Nolde M, Alayash Z, et al. Cannabis use does not impact on type 2 diabetes: a two-sample Mendelian randomization study. Addict Biol. 2021;26(6):e13020. doi:10.1111/adb.13020. PMID: 33432723. Study record
3. Mousavi SE, Tondro Anamag F, Sanaie S. Association between cannabis use and type 2 diabetes mellitus: a systematic review and meta-analysis. Phytother Res. 2023;37(2):517-528. doi:10.1002/ptr.7630. PMID: 36184502. Study record
4. Jadoon KA, Ratcliffe SH, Barrett DA, et al. Efficacy and safety of cannabidiol and tetrahydrocannabivarin on glycemic and lipid parameters in patients with type 2 diabetes: a randomized, double-blind, placebo-controlled, parallel group pilot study. Diabetes Care. 2016;39(10):1777-1786. doi:10.2337/dc16-0650. PMID: 27573936. Study record
5. Bryan AD, Skrzynski CJ, Giordano G, et al. Cannabis use is associated with less peripheral inflammation but similar insulin sensitivity as nonuse in healthy adults. Am J Med. 2025;138(1):78-86. doi:10.1016/j.amjmed.2024.07.037. PMID: 39181467. Study record
6. Barre T, Nishimwe ML, Protopopescu C, et al. Cannabis use is associated with a lower risk of diabetes in chronic hepatitis C-infected patients. J Viral Hepat. 2020;27(12):1280-1288. doi:10.1111/jvh.13367. PMID: 32681671. Study record
