Cannabis Use Was Linked to Suicide Risk, but the Link Disappeared After Accounting for Other Factors | Research

The takeaway

In a 33-year follow-up of 50,087 Swedish men, cannabis use was initially associated with higher suicide risk (OR 1.62), but the association was fully explained by pre-existing psychological and behavioral problems.

Read this if you have heard that cannabis increases suicide risk and want to see what a large long-term study actually found.

Association between cannabis and suicide (OR 1.62) eliminated after adjusting for confounders (OR 0.88)

What the researchers found

Researchers followed 50,087 men conscripted for Swedish military service over 33 years, during which 600 suicides or deaths from undetermined causes occurred.

Cannabis use measured at conscription was associated with increased suicide risk in the crude analysis (OR 1.62, 95% CI 1.28-2.07).

However, after adjusting for confounding factors, including markers of pre-existing psychological and behavioral problems, the association was completely eliminated (adjusted OR 0.88, 95% CI 0.65-1.20).

The authors concluded that cannabis use is unlikely to have a strong direct effect on suicide risk. The initial association was explained by the fact that people who used cannabis also tended to have pre-existing psychological vulnerabilities.

Why it matters

This study addressed a critical question with unusual rigor: a large population, 33-year follow-up, and objective outcome measurement (completed suicide, not just ideation). The finding that confounding fully explained the association challenges claims that cannabis directly causes suicide.

The numbers in context

50,087 men followed for 33 years. 600 suicides or deaths from undetermined causes (1.2% of cohort). Crude OR for ever-use: 1.62 (95% CI 1.28-2.07). Adjusted OR: 0.88 (95% CI 0.65-1.20).

How the study worked

Longitudinal cohort study of 50,087 Swedish military conscripts with cannabis use measured non-anonymously at conscription. Suicides were identified through 33 years of follow-up using the National Cause of Death Register. Logistic regression with adjustment for psychological and behavioral confounders.

What this study cannot tell us

All-male military conscript cohort may not generalize to women or non-military populations. Cannabis use was measured at a single time point (conscription) and may not reflect lifetime use patterns. Confounders may not capture all relevant factors.

How to read the evidence

Large longitudinal cohort (50,087) with 33-year follow-up and objective outcome measurement (completed suicide). Strong design, though limited to male conscripts and single baseline measurement.

When this study was published

Published in 2009. Subsequent studies have continued to examine the cannabis-suicide relationship with mixed findings, though most agree that confounding plays a major role.

The bigger picture

The debate about cannabis and suicide often conflates association with causation. This study demonstrates the importance of accounting for confounding: people who use cannabis may be at higher baseline risk for suicide due to underlying factors, not because of cannabis itself.

Questions still open

Would similar results be found in women or in populations with different cannabis use patterns? Could heavy or chronic use have effects not captured by the binary "ever use" measure? Are there subgroups for whom cannabis might independently affect suicide risk?

Common questions

Does cannabis cause suicide?

This large study found no evidence that cannabis directly increases suicide risk after accounting for pre-existing psychological problems. People who use cannabis may be at higher risk, but the risk appears to come from underlying factors rather than cannabis use itself.

Why did the association disappear after adjustment?

People who used cannabis at military conscription also tended to have more psychological and behavioral problems. These underlying problems, not the cannabis use, explained the higher suicide rates. This is why raw associations can be misleading without proper statistical adjustment.

Read the original research

Cannabis and suicide: longitudinal study.

The British journal of psychiatry : the journal of mental science, 195(6), 492-7

View on PubMed ↗Publisher’s paper ↗

Citation

Price, Ceri; Hemmingsson, Tomas; Lewis, Glyn; Zammit, Stanley; Allebeck, Peter. (2009). Cannabis and suicide: longitudinal study.. The British journal of psychiatry : the journal of mental science, 195(6), 492-7. https://doi.org/10.1192/bjp.bp.109.065227