How-To Guides

In this nationally representative 2022 sample, 11.6% of US veterans reported cannabis use in the past 6 months and 2.9% screened positive for probable cannabis use disorder (CUD). Of the 2,441 veterans who completed the survey, 284 reported cannabis use and 70 had probable CUD, while 85.5% reported no cannabis use.

The Pacific region had the highest prevalence of both outcomes, with 18.6% reporting cannabis use and 8.8% screening positive for probable CUD. Regional differences across the 9 US Census Bureau–defined regions were statistically significant overall (c7b2b9b6 = 73.33, P < .001).

Yes. The study found that the Pacific region's probable CUD rate of 8.8% was significantly higher than every other region (all uncorrected Ps < .05), including New England, where probable CUD was 3.5%. The authors also noted that the Pacific region's CUD rate exceeded a contemporary estimate for the general adult US population of 7%.

Yes for cannabis use, but not for probable CUD. New England had a cannabis use prevalence of 13.4%, which did not significantly differ from the Pacific region's 18.6% (P = .24), but its probable CUD prevalence was 3.5%, which was significantly lower than the Pacific region's 8.8%.

No. The authors found elevated cannabis use and probable CUD in some states that did not have legalized recreational cannabis at the time of data collection. Examples included Oklahoma, which had 20.5% cannabis use, and North Carolina and Indiana, which were among the top 5 states for probable CUD at 9.0% and 4.9%, respectively.

Among states with at least 30 respondents, the highest cannabis use prevalences were observed in:

• Oregon: 26.7%
• California: 22.4%
• Oklahoma: 20.5%
• Massachusetts: 16.7%
• New York: 15.7%
• Alabama: 15.0%
• Illinois: 14.9%
• Arizona: 14.3%
• Ohio: 14.1%
• Virginia: 13.7%

Among states with at least 30 respondents, the highest probable CUD prevalences were:

• Oregon: 13.3%
• Washington: 9.6%
• North Carolina: 9.0%
• California: 6.7%
• Indiana: 4.9%
• New York: 4.5%
• Pennsylvania: 3.7%
• Texas: 3.2%
• Wisconsin: 3.1%
• Kentucky: 3.1%

Veterans with probable CUD were younger than veterans who used cannabis without probable CUD, and both groups were younger than veterans who reported no cannabis use. Veterans with probable CUD were also more likely to be other or multiracial, less likely to be married or partnered, less likely to have completed college, less likely to have served 10 or more years in the military, and more likely to report the VA as their main source of health care.

Yes. Although cannabis use and probable CUD were more common in younger veterans, the mean ages were still 59 years for veterans who used cannabis and 53 years for veterans with probable CUD. The authors interpreted this as supporting continued clinical monitoring of cannabis-related problems in aging veteran populations.

Probable CUD was identified with the 3-item Cannabis Use Disorders Identification Test–Short Form (CUDIT-SF), a validated DSM-5–aligned screening tool. It asks about inability to stop using cannabis, spending a great deal of time getting or using cannabis or recovering from it, and memory or concentration problems after use; a score of 2 or higher indicated probable CUD.

Cannabis use was measured by asking veterans whether they had used any cannabis over the past 6 months, including marijuana, hashish, THC, pot, grass, weed, or reefer. Veterans who reported cannabis use but did not screen positive on the CUDIT-SF were classified as cannabis users without probable CUD.

The findings suggest that routine screening may be especially important in higher-prevalence regions, particularly the Pacific region, and in VA settings. The authors noted that veterans with probable CUD were more likely to use the VA as their primary source of health care and suggested that systematic CUD screening within VHA primary care and mental health services may be particularly important in high-prevalence regions.

No. The study was cross-sectional, so it cannot establish causation. The authors suggested that regional variability in medical and recreational cannabis legalization likely contributed to the observed differences, but they also noted that other factors, such as mental health conditions, substance use comorbidity, and access to behavioral health services, may also help explain the regional pattern.

The main limitations were that the study was cross-sectional, relied on self-reported cannabis use and a screening measure for probable CUD, and had smaller sample sizes in some regions and states. According to the authors, these factors mean the study cannot support causal inference, may be affected by underreporting or overreporting, and may have less precise prevalence estimates in some geographic areas.