rethinkTHC Search
Menu
Study breakdown

Cannabis exposure tracked with all four major mental health problems across US states and time periods

Cross SectionalModerate evidence
The takeaway

An analysis of 410,138 NSDUH respondents found cannabis exposure was significantly associated with all four measures of mental illness across US regions and time, with serious mental illness doubling as cannabis use increased.

Mental health epidemiologists, cannabis policy makers, and public health officials in legalized states.

SMI doubled (3.62% to 7.06%) as cannabis use increased; legalization AFE 12.91%

What the researchers found

Across 410,138 NSDUH respondents (76.7% response rate), cannabis exposure was significantly associated with any mental illness, major depression, serious mental illness (SMI), and suicidal ideation in geospatial models adjusted for demographics and other substance use. SMI rates doubled (3.62% to 7.06%) as cannabis use increased across regions. Cannabis decriminalization was associated with a 3.28% attributable fraction of SMI, and legalization with 12.91% attributable fraction.

Why it matters

This is one of the first studies to apply formal causal inference methods to the cannabis-mental health relationship at a population level, finding dose-response and temporal-sequential relationships consistent with causation.

The numbers in context

410,138 respondents; 76.7% response rate; SMI doubled from 3.62% to 7.06% with cannabis increase; decriminalization AFE 3.28%; legalization AFE 12.91%; significant for all 4 MH outcomes.

How the study worked

Ecological cohort study using NSDUH geographically-linked substate data (2010-2012 and 2014-2016) with two-stage geotemporospatial robust generalized linear regression and formal causal inference analysis.

What this study cannot tell us

Ecological design (associations at area level may not hold for individuals); cross-sectional components within the cohort; NSDUH is self-report; cannot control for all confounders at the individual level; causal inference methods applied to observational data have inherent limitations.

How to read the evidence

Moderate: very large sample with formal causal inference methods, but ecological design limits individual-level conclusions.

When this study was published

Published 2020.

The bigger picture

If the causal inference is correct, cannabis legalization may be contributing measurably to population-level mental health burden. The 12.91% attributable fraction for serious mental illness under legalization would have substantial public health implications.

Questions still open

  • Do individual-level longitudinal studies confirm these population-level associations? Would cannabis policies that limit THC potency reduce mental health impacts?

Common questions

Does cannabis legalization increase mental illness?
This study found cannabis legalization was associated with a 12.91% attributable fraction of serious mental illness at the population level. Serious mental illness rates doubled as cannabis use increased across US regions. However, ecological data cannot definitively prove individual-level causation.
Which mental health conditions are linked to cannabis?
All four measured: any mental illness, major depression, serious mental illness, and suicidal ideation were significantly associated with cannabis exposure after adjusting for demographics and other substance use.

Read the original research

Co-occurrence across time and space of drug- and cannabinoid- exposure and adverse mental health outcomes in the National Survey of Drug Use and Health: combined geotemporospatial and causal inference analysis.

BMC public health, 20(1), 1655

Citation

Reece, Albert Stuart; Hulse, Gary Kenneth. (2020). Co-occurrence across time and space of drug- and cannabinoid- exposure and adverse mental health outcomes in the National Survey of Drug Use and Health: combined geotemporospatial and causal inference analysis.. BMC public health, 20(1), 1655. https://doi.org/10.1186/s12889-020-09748-5

Explore the wider topic