Cannabis Use in Australia Held Steady Over 15 Years, But Young People Show Greater Vulnerability to Disorder

A repeated cross-sectional analysis of two Australian national mental health surveys spanning 2007 to 2020–22 finds that population-level cannabis use and cannabis use disorder prevalence remained statistically stable, while the association between younger age and disorder grew significantly stronger, raising targeted public health concerns about youth risk in a shifting policy landscape.

Why This Matters

Cannabis policy is evolving rapidly in many jurisdictions, and Australia legalized medical cannabis in 2016 with growing recreational access since. Clinicians and policymakers urgently need population-level data to understand whether these shifts are translating into measurable changes in harm. This study provides the most comprehensive national comparison available for Australia, and its finding that overall use rates are stable while risk appears to concentrate among young people challenges simplistic narratives about liberalization inevitably driving population-wide surges in disorder. The timing matters because evidence-based youth-targeted interventions depend on knowing where risk is actually growing.

Clinical Summary

Cannabis use disorder remains a significant clinical concern worldwide, with rising product potency and changing regulatory environments intensifying the need for robust epidemiological surveillance. This repeated cross-sectional study, published by Mewton and colleagues, drew on two waves of Australia’s National Survey of Mental Health and Wellbeing, conducted in 2007 (n = 8,841) and 2020–22 (n = 15,893). Both waves used face-to-face interviews with the World Mental Health Composite International Diagnostic Interview (WMH-CIDI 3.0) to generate DSM-IV diagnoses. The study employed weighted logistic regression with interaction terms to test whether associations between demographic and clinical characteristics and cannabis use disorder changed across the 15-year interval.

Past 12-month cannabis use was 6.7% in both survey waves, and cannabis use disorder prevalence was 1.0% in 2007 and 0.6% in 2020–22, with overlapping confidence intervals indicating no statistically significant change. Across both waves, younger age, male sex, early initiation before age 18, polysubstance use, and comorbid mental or substance disorders remained strongly associated with CUD, with odds ratios ranging from approximately 3 to 78. The most notable shift was that the association between being aged 16 to 25 and having CUD grew significantly stronger by 2020–22 (interaction OR 2.39, 95% CI 1.20 to 4.77), while the association between polysubstance use and CUD weakened (interaction OR 0.42, 95% CI 0.20 to 0.91). The authors caution that the cross-sectional design cannot establish causation, the DSM-IV criteria used differ from current DSM-5 standards, and the study cannot distinguish medical from non-medical cannabis use. They call for continued longitudinal surveillance and targeted research into the mechanisms driving the youth-specific risk shift.

Dr. Caplan’s Take

What I find most valuable about this study is that it resists the temptation to tell a simple story. The headline finding, that overall cannabis use and disorder rates did not surge over 15 years of evolving policy, is genuinely informative. But the more clinically important finding is the one buried in the interaction analysis: young people aged 16 to 25 appear to carry a meaningfully larger share of CUD risk in 2020–22 than they did in 2007. When patients or parents ask me whether cannabis has become “safer” because it is more accepted, this is the kind of evidence I need to frame an honest answer. Stability at the population level can mask real concentration of harm in vulnerable groups.

In practice, I screen all patients under 25 more carefully for cannabis-related problems than I did a decade ago, not because of any single study but because evidence like this consistently points toward a shifting risk profile in young people. I discuss dose, potency, and frequency openly, and I remain cautious about medical cannabis recommendations for younger adults without a compelling indication and a clear monitoring plan. The data do not support alarm, but they absolutely support vigilance.

Clinical Perspective

This study sits at a descriptive stage in the research arc. It confirms that Australia has not experienced the kind of population-level prevalence increase observed in some North American jurisdictions following liberalization, which is a useful benchmark. However, it does not explain why the youth-CUD association strengthened. Possible explanations include increased access to higher-potency products among young users, shifting social contexts of use, or changes in help-seeking and detection patterns. The cross-sectional design and use of DSM-IV rather than DSM-5 criteria limit the clinical specificity of the CUD diagnosis. The evidence does not support changing screening or diagnostic practices based on this study alone, but it does support the argument that age-stratified monitoring should be standard in cannabis epidemiology.

From a pharmacological standpoint, clinicians should be aware that higher-potency THC products, which have become more available in Australia’s unregulated market, carry greater risk of dependence and psychotic symptoms, particularly in adolescents and young adults whose endocannabinoid systems are still developing. No specific drug interaction findings emerge from this survey, but the weakening polysubstance-CUD association is worth tracking: it may suggest that cannabis use is becoming more isolated rather than embedded in polydrug patterns, which has implications for how we assess risk profiles. The most actionable recommendation for clinicians now is to incorporate age-specific cannabis screening, particularly for patients aged 16 to 25, into routine mental health and substance use assessments, regardless of the patient’s stated reason for the visit.

Study at a Glance

What Kind of Evidence Is This

This is a repeated cross-sectional epidemiological study, which occupies a descriptive and associational tier in the evidence hierarchy. It uses two independent, nationally representative samples to compare prevalence and risk factor associations across time. The most important inference constraint is that independent cross-sectional samples cannot track individuals, meaning that apparent changes in associations over time could reflect shifts in the composition of the population surveyed rather than true changes in vulnerability or behavior within any particular group.

How This Fits With the Broader Literature

The stable prevalence finding stands in contrast to data from the United States, where the National Survey on Drug Use and Health has documented increases in cannabis use and disorder following state-level legalization, and from Canada, where post-legalization surveys have shown modest upticks in use among certain age groups. The Australian context is distinct because recreational cannabis remains largely illegal, with only medical access formally legalized in 2016. The youth-specific risk finding aligns with a growing international literature, including work by Hasin and colleagues in the US, suggesting that younger populations may be disproportionately affected by changes in the cannabis landscape even when population-level metrics appear reassuring.

The weakening polysubstance-CUD association is a novel contribution that has not been widely reported elsewhere and deserves replication. It may reflect a broader cultural shift toward cannabis-specific rather than polydrug use patterns, a phenomenon that has been noted anecdotally in clinical settings but has lacked epidemiological documentation at this scale.

Common Misreadings

The most likely overinterpretation is to conclude that cannabis policy liberalization in Australia has been shown to have no effect on harm. This study does not test that causal claim. Stable population-level prevalence is compatible with multiple scenarios, including rising harm in subpopulations offset by declining harm in others, or effects that have not yet manifested at the population level. Similarly, the strengthened youth-CUD association should not be read as proof that young Australians are biologically more vulnerable now than in 2007. The cross-sectional design cannot distinguish genuine changes in vulnerability from changes in product potency, use patterns, detection, or the demographic composition of the youth population itself.

Bottom Line

This study provides reassurance that Australia has not experienced a measurable population-level increase in cannabis use or cannabis use disorder over 15 years, but it simultaneously flags a meaningful and statistically significant concentration of disorder risk among young people aged 16 to 25. It does not establish why this shift has occurred. For clinicians, the practical implication is clear: age-stratified screening for cannabis-related problems, particularly in younger patients, deserves greater clinical attention now than it did a decade ago.

References

Physician-Led, Whole-Person Care

A doctor who takes the time to truly understand you.

Personal care that starts with listening and is guided by experience and ingenuity.

Health, Longevity, Wellness

One-on-One Cannabis Guidance

Metabolic Balance

Leave a Message
Metabolic Care
Medical Consulting
Cannabis Care

• Newsletter Signup Form

  Enter your email address to subscribe to our newsletter.
• Email Address*
  example@example.com
• Subscribe
• Should be Empty: