Genetic analysis finds no causal link between cannabis use and type 2 diabetes risk | Research

The takeaway

A Mendelian randomization study using genetic data from nearly one million people found no evidence that cannabis use causes type 2 diabetes.

People with diabetes concerned about cannabis use, metabolic health researchers, anyone interested in cannabis and metabolic effects

OR = 1.00 for lifetime cannabis use and type 2 diabetes (no effect)

What the researchers found

Using genetic variants as proxies for cannabis use, researchers found no causal effect of either lifetime cannabis use (OR 1.00, 95% CI 0.93-1.09) or cannabis use disorder (OR 1.03, 95% CI 0.99-1.08) on type 2 diabetes risk.

Why it matters

Previous observational studies have produced conflicting results about cannabis and diabetes risk. By using genetic variants as natural experiments, this study provides stronger evidence that the relationship (or lack thereof) is unlikely to be due to confounding.

The numbers in context

184,765 participants for lifetime cannabis use analysis; 2,387 cases/48,985 controls for cannabis use disorder; 74,124 cases/824,006 controls for type 2 diabetes; OR = 1.00 (95% CI 0.93-1.09) for lifetime use

How the study worked

Two-sample Mendelian randomization using 19 genetic variants for lifetime cannabis use and 14 for cannabis use disorder, linked to type 2 diabetes outcomes from genome-wide association studies totaling over 900,000 participants.

What this study cannot tell us

Mendelian randomization assumes genetic variants only affect diabetes through cannabis use, which may not hold perfectly. European-ancestry population may not generalize to other groups. Cannabis use disorder had a relatively small case sample.

How to read the evidence

Mendelian randomization with large sample sizes provides stronger causal inference than observational studies

When this study was published

Published in 2021 using genome-wide association study data from European-ancestry populations.

The bigger picture

This adds to a growing body of Mendelian randomization evidence examining cannabis and metabolic health, helping separate genuine causal effects from the confounding that plagues observational studies in this area.

Questions still open

Could specific cannabinoids (THC vs CBD) have different metabolic effects not captured by overall cannabis use? Does the route of administration or frequency of use matter for metabolic outcomes? What about type 1 diabetes?

Common questions

Does cannabis cause type 2 diabetes?

This genetic study found no evidence of a causal link. The odds ratio was exactly 1.00, meaning cannabis use showed no effect on diabetes risk in either direction.

How is this different from previous studies?

Unlike observational studies that can be confounded by lifestyle factors, this Mendelian randomization study used genetic variants as natural experiments, providing stronger evidence about causation.

Read the original research

Cannabis use does not impact on type 2 diabetes: A two-sample Mendelian randomization study.

Addiction biology, 26(6), e13020

View on PubMed ↗Publisher’s paper ↗

Citation

Baumeister, Sebastian-Edgar; Nolde, Michael; Alayash, Zoheir; Leitzmann, Michael; Baurecht, Hansjörg; Meisinger, Christa. (2021). Cannabis use does not impact on type 2 diabetes: A two-sample Mendelian randomization study.. Addiction biology, 26(6), e13020. https://doi.org/10.1111/adb.13020