Cannabis Addiction: The Numbers Behind Cannabis Use Disorder

The public debate offers two answers to whether cannabis is addictive, and neither is right. It is not harmless and it is not comparable to opioids. The evidence supports a specific middle position.

Cannabis Use Disorder Is a Real Diagnosis

Cannabis use disorder appears in DSM-5, and cannabis withdrawal was added as a distinct diagnosis in the same edition, reflecting evidence accumulated over the preceding decades.

Diagnosis rests on criteria familiar from other substance use disorders: using more or longer than intended, unsuccessful attempts to cut down, craving, use interfering with work or relationships, continued use despite problems, tolerance and withdrawal. Severity is graded by how many criteria are met.

The important point is that this is a clinical category with defined criteria, not a rhetorical device.

The Numbers

Estimates vary by population and method, and the broad picture is consistent.

Roughly 9 percent of people who ever use cannabis develop dependence, based on epidemiological work.

Among daily users the figure is far higher, with estimates commonly ranging from about 25 percent to as high as 50 percent depending on the study and criteria.

Risk is substantially higher for people who start in adolescence, with some estimates around one in six for those beginning as teenagers.

For comparison, dependence rates are higher for nicotine, heroin and cocaine. That comparison is often used to dismiss the issue, and it does not follow. A 9 percent overall rate applied to a very large number of users produces a large number of affected people, and cannabis use disorder is among the most common substance use disorders in many countries by sheer volume.

Withdrawal Is the Part People Miss

Withdrawal is the strongest evidence for physical dependence and the most practically important thing for anyone trying to stop.

Symptoms include irritability, anxiety, sleep disturbance including vivid dreams, decreased appetite, restlessness, low mood, and physical symptoms such as headache, sweating and abdominal discomfort.

Timeline. Onset within one to two days, peaking around days two to six, mostly resolving over two to four weeks, with sleep disturbance often lasting longest.

Withdrawal is not medically dangerous the way alcohol or benzodiazepine withdrawal can be. It is uncomfortable enough to drive relapse, and sleep disruption is the single most common reason people resume.

The trap is that these symptoms closely resemble the problems people use cannabis to manage. Anxiety and insomnia during withdrawal feel like proof cannabis was treating an underlying condition, when they are the dependence unwinding. The only way to distinguish them is to get past the two to four week window.

What Raises Risk

Daily or near-daily use, the strongest behavioural predictor.

Early first use, particularly in adolescence.

Higher potency products. Risk associations scale with THC concentration, and modern products are substantially stronger than those studied in older research.

Using to cope with anxiety, low mood, trauma or insomnia, rather than socially.

Co-occurring mental health conditions.

Family history of substance use disorder.

The Gateway Question

Cannabis use does statistically precede use of other drugs, and that observation supports several explanations. Candidate accounts include a shared underlying propensity toward substance use, social access through illicit markets bringing contact with other drugs, and a genuine pharmacological sensitisation effect.

Evidence for the strong pharmacological version is weak. Most people who use cannabis do not progress to other drugs. Treating sequence as proof of causation is the error, and it is worth noting that prohibition itself creates the market conditions in the access explanation.

What Treatment Works

Effective approaches exist and are underused.

Cognitive behavioural therapy, addressing use patterns and triggers.

Motivational enhancement therapy, working on ambivalence about change.

Contingency management, providing structured incentives for abstinence, which has strong evidence across substance use disorders.

Combined approaches generally outperform single ones.

There is no approved pharmacotherapy for cannabis use disorder. Various agents have been trialled, including targeting withdrawal symptoms, without an established treatment emerging. Managing sleep during withdrawal is where the most practical help currently sits.

Signs Worth Taking Seriously

Using more than intended, failed attempts to cut back, needing more for the same effect, using on waking, using alone rather than socially, spending more than you can afford, avoiding situations where you cannot use, and continuing despite problems it is causing.

If several of those apply, it is worth a conversation with a clinician rather than a debate about whether cannabis is addictive.

Bottom Line

Cannabis use disorder is a recognised diagnosis affecting roughly 9 percent of people who ever use and a much larger share of daily users, with adolescent onset and high potency raising risk. Withdrawal is real, peaks in the first week, resolves over two to four weeks, and mimics the symptoms people use cannabis to treat. Behavioural treatments work and no medication is approved.

Disclaimer: This article is for educational purposes only and does not constitute medical advice. Always consult with a qualified healthcare professional before making any health-related decisions.