Cannabis use disorder was associated with higher all-cause mortality

Design

Matched retrospective cohort study, Colorado health system, 2006 to 2024

Population

100,140 people aged 12 or over: 16,693 with cannabis use disorder, 83,447 without

Primary outcome

All-cause, injury-related and illness-related mortality

Effect

All-cause mortality HR 1.52 (95% CI 1.30 to 1.77); overdose HR 2.59 (1.30 to 5.19)

This matched cohort study from a Colorado health system followed 100,140 people aged 12 or over, 16,693 of them with a diagnosis of cannabis use disorder (CUD) and 83,447 without, for a median of 7.7 years. Deaths came from state vital records.

Crude all-cause mortality was 899.5 per 100,000 person-years with CUD against 374.6 without. After adjustment, CUD was associated with higher all-cause mortality (HR 1.52, 95% CI 1.30 to 1.77), injury-related mortality (1.91, 1.37 to 2.67) and illness-related mortality (1.40, 1.17 to 1.67). By cause, the hazard ratios were raised for overdose (2.59, 1.30 to 5.19), accidents (1.96, 1.01 to 3.80) and cancer (1.71, 1.20 to 2.46), but not for self-harm, cardiovascular or respiratory deaths.

A diagnosis of CUD is a marker, and people who carry it differ in many ways that a matched analysis cannot fully capture, including other substance use. The authors describe CUD as a marker of risk. The study cannot show that cannabis itself caused these deaths.

• Ask about cannabis use, and about other substances, in patients with injuries or unexplained decline.
• Treat a CUD diagnosis as a prompt to review overdose, injury and cancer risk factors.
• Overdose and accidents drove much of the excess; check for co-use of opioids and alcohol.
• Do not tell patients the study proves cannabis shortens life; say it marks higher risk.

Why it matters

It reframes CUD from a minor diagnosis to a marker that sits alongside raised overdose and injury deaths.

Don't overread it

This is an observational cohort from one US health system; residual confounding is likely and causation is not shown.

The statistics, in plain English

A hazard ratio of 1.52 means that at any time point, people with CUD died at about one and a half times the rate of matched people without it. The wide intervals on cause-specific results reflect small numbers of deaths.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.