What Actually Puts People at Risk for Cannabis Use Disorder? | Research

The takeaway

An umbrella review of meta-analyses found that while many risk factors for cannabis, cocaine, and opioid use disorders were statistically significant, none had convincing evidence, with antisocial behavior showing the strongest link to cannabis use disorder.

Addiction researchers, substance use prevention specialists, clinical psychiatrists

Only antisocial behavior had convincing evidence as a risk factor for cannabis use disorder

What the researchers found

Of 19 associations between 12 risk/protective factors and substance use disorders, none reached "convincing" evidence. Sensitivity analyses found convincing evidence only for antisocial behavior predicting cannabis use disorder (OR=3.34, 95% CI: 2.53-4.41). Smoking had highly suggestive evidence for nonmedical prescription opioid use (OR=3.07).

Why it matters

Despite many claimed risk factors for substance use disorders, this rigorous evaluation shows that very few have truly strong evidence, which should guide prevention efforts toward the most well-supported targets.

The numbers in context

3,072 initial references; 5 meta-analyses included; 19 associations examined; 12 risk/protective factors; cases: 4,539; total N: over 1.1 billion; 84% of associations statistically significant but none convincing; antisocial behavior and CUD OR=3.34 in sensitivity analysis.

How the study worked

Umbrella review of meta-analyses of observational studies from PubMed-MEDLINE/PsycInfo (through September 2020), grading evidence credibility from "not significant" to "convincing" using established criteria, with quality assessment via AMSTAR-2.

What this study cannot tell us

Limited to published meta-analyses which may have their own quality issues; 40% of included meta-analyses rated critically low quality; cannot examine interactions between risk factors.

How to read the evidence

Rigorous umbrella review methodology with evidence grading, limited by the quality of underlying meta-analyses.

When this study was published

Search through September 2020.

The bigger picture

The gap between the number of claimed risk factors and those with convincing evidence highlights the need for higher-quality longitudinal research on what truly predicts substance use disorder development.

Questions still open

Why do so few risk factors for substance use disorders have convincing evidence? Would newer, better-designed longitudinal studies change the evidence grades?

Common questions

What predicts cannabis use disorder?

This umbrella review found that antisocial behavior was the only risk factor with convincing evidence for predicting cannabis use disorder (OR=3.34). While many other factors were statistically significant, their evidence quality was weak.

Why aren't more risk factors considered convincing?

The "convincing" standard requires consistent results across large studies with minimal bias. Most existing meta-analyses on substance use disorder risk factors had methodological limitations that prevented their findings from reaching this threshold.

Read the original research

Risk and protective factors for cannabis, cocaine, and opioid use disorders: An umbrella review of meta-analyses of observational studies.

Neuroscience and biobehavioral reviews, 126, 243-251

View on PubMed ↗Publisher’s paper ↗

Citation

Solmi, Marco; Dragioti, Elena; Croatto, Giovanni; Radua, Joaquim; Borgwardt, Stefan; Carvalho, Andrè F; Demurtas, Jacopo; Mosina, Anna; Kurotschka, Peter Konstantin; Shin, Jae Il; Fusar-Poli, Paolo. (2021). Risk and protective factors for cannabis, cocaine, and opioid use disorders: An umbrella review of meta-analyses of observational studies.. Neuroscience and biobehavioral reviews, 126, 243-251. https://doi.org/10.1016/j.neubiorev.2021.03.014