Cannabis Use Disorder in American Adults

In December 2025, President Trump issued Executive Order 14370, declaring it federal policy to increase medical marijuana and cannabidiol (CBD) research and directing the Attorney General to expedite the movement of marijuana from Schedule I to Schedule III. The order stated that marijuana has credible scientific support for pain, medically related anorexia, and nausea and vomiting induced by chemotherapy.

Effective April 28, 2026, the Justice Department placed state-licensed medical marijuana and FDA-approved plant-derived marijuana products in Schedule III. It created an expedited DEA-registration pathway for state-licensed growers, distributors, and dispensaries. Qualifying medical-marijuana businesses are relieved of the punitive federal tax provision that formerly prevented Schedule I and II cannabis businesses from deducting ordinary expenses. A broader proceeding to move all marijuana essentially to Schedule III was held from June 29 through July 15; a final decision on that broader proposal is still pending. Thus, although the administration describes its objective principally as “increasing research,” its actions will almost certainly facilitate and expand medical use.

The FDA has not approved the cannabis plant itself for treating any disease. It has approved purified CBD for certain severe epilepsies and synthetic cannabinoid drugs for chemotherapy-induced nausea and vomiting and AIDS-related anorexia. Evidence for pain and MS symptoms suggests only modest benefits. The evidence does not justify treating “medical cannabis” as a broadly effective therapeutic class. Most favorable trials studied standardized oral compounds or precisely formulated sprays, not the highly variable flower, concentrates, vapes, and edibles sold by dispensaries. THC concentration, CBD-to-THC ratio, contaminants, dose and bioavailability may differ greatly.

Cannabis use disorder (CUD) is a medical diagnosis describing a problematic pattern of cannabis use that causes clinically significant impairment or distress. Under DSM-5 criteria, a person must experience at least two of these symptoms within 12 months:

Using more cannabis, or for longer, than intended
Wanting or repeatedly trying unsuccessfully to reduce or stop
Spending considerable time obtaining, using, or recovering from cannabis
Craving cannabis
Failing to meet obligations at work, school, or home
Continuing despite relationship or social problems
Giving up important activities
Using in hazardous situations, such as while driving
Continuing despite knowing it worsens a physical or psychological problem
Developing tolerance
Experiencing withdrawal

Severity is classified as:
Mild: 2–3 symptoms
Moderate: 4–5
Severe: 6 or more

Frequent, legal, or medically authorized cannabis use does not by itself constitute CUD; the essential features are loss of control, adverse consequences, impairment, or distress. Severe CUD is commonly described as cannabis addiction, though cannabis use does not meet the strict pharmacological definition of addiction, as it is not associated with withdrawal symptoms as are opioid or alcohol withdrawal. I am not using the DSM V definition of addiction, as I think it incoherent.

In this discussion, addiction denotes a chronic pattern of drug use characterized by all four of the following: tolerance, habituation, dependence, and a recognizable withdrawal syndrome upon cessation. This definition is narrower than DSM-5 cannabis use disorder; consequently, the prevalence of CUD reported in the study cannot be interpreted as the prevalence of cannabis addiction.

A recent study published in JAMA Psychiatry: Trends in the Prevalence and Severity of Alcohol and Cannabis Use Disorders Among US Adults analyzed responses from 186,823 adults participating in the nationally representative National Survey on Drug Use and Health from 2021 through 2024. The survey represents the civilian, noninstitutionalized adult population. It does not include prisoners, active-duty military personnel, other institutional residents, or people experiencing homelessness who are not staying in shelters. Cannabis use disorder was determined from respondents’ answers to questions corresponding to DSM-5 criteria. The authors combined moderate and severe CUD into a single category, defined as four or more symptoms. This was not a longitudinal study following the same people. It was a series of annual cross-sectional surveys used to measure population-level changes.

Principal national findings

These percentage changes corresponded to an estimated increase in the number of adults with CUD from 15.3 million in 2021 to 19.4 million in 2024, about 4 million additional people over three years. The increases in overall CUD and moderate-to-severe CUD were statistically significant.

These percentage changes corresponded to an estimated increase in the number of adults with CUD from 15.3 million in 2021 to 19.4 million in 2024, about 4 million additional people over three years. The increases in overall CUD and moderate-to-severe CUD were statistically significant. In 2024, 36.2% of male cannabis users met criteria for CUD, while 26.6% of female cannabis users met criteria for CUD. The absolute prevalence remained highest among younger adults, but some of the largest increases occurred in middle and later life.

The paper was deliberately constructed as a comparison between cannabis and alcohol disorders. The estimated number of adults with alcohol use disorder (AUD) declined from 28.8 million to 27.1 million. AUD prevalence in men was statistically stable. AUD in women declined from 9.7% to 8.2%. Moderate-to-severe AUD generally remained stable or declined.

The evidence convincingly shows a recent increase in self-reported CUD and its severity. It does not, however, establish that legalization or increased THC potency caused the increase. The study did not compare legalization policies among states, measure the potency of products used by individual respondents, or follow individuals over time.

The most rational approach to cannabis use is not to regard it as harmless. It should be treated as a psychoactive drug whose risks rise with THC concentration, dose, frequency, and susceptibility. The long-term use of a mind-altering drug is likely to be undesirable. Prolonged recreational use may have serious adverse effects.

Complete avoidance is generally most rational for adolescents, pregnant or breastfeeding women, people with a personal or strong family history of psychosis, and those who have had substance-use disorders. Considerable caution is also warranted for those with cardiovascular disease, cognitive impairment, poor balance, or multiple medications. Medical use might be reasonable for a specific refractory symptom, but only as a very-low-dose, closely monitored trial after reviewing drug interactions, blood-pressure effects, balance, cognition, and driving. Adults over 55 may be more sensitive to cannabis because of age-related physiological changes, medical comorbidities and polypharmacy, and they face greater risks from drug interactions, cognitive effects, orthostatic hypotension and falls.

The federal government is moving ahead of the evidence concerning both the beneficial and adverse effects of prolonged cannabis use. Expanding research is entirely reasonable; conferring Schedule III status and federal accommodation to state dispensary systems is a broad policy choice. It may eventually produce better evidence, but it should not itself be mistaken for evidence of clinical effectiveness.