Using Cannabis in Multiple Forms May Signal Higher Addiction Risk

Nearly half of U.S. adults who use cannabis mainly smoke it, but the roughly 1 in 7 who use it in nearly every available form carry the heaviest use patterns and the highest rates of cannabis use disorder, researchers at NYU Grossman School of Medicine reported in a national study published online August 8, 2026 in the journal Addictive Behaviors.

The team analyzed responses from 17,732 adults who reported past-month cannabis use on the 2022–2023 National Survey on Drug Use and Health, the annual federal survey of Americans aged 12 and older. Applying a statistical grouping method called latent class analysis, the researchers sorted participants by how they actually consumed cannabis rather than by pre-set categories, and four groups emerged: 48.2 percent mainly smoked, 20.5 percent both smoked and vaped, 16.9 percent mainly used edibles, and 14.4 percent fell into what the authors call the polymodal group, using cannabis in nearly every form, including dabbing heated concentrates.

The outcomes tracked steeply across those groups. About half of the polymodal group used cannabis every day, according to the NYU Langone Health announcement, compared with 28 percent of primary smokers, 31 percent of those who smoked and vaped, and about 13 percent of the edibles group. Cannabis use disorder, the clinical diagnosis for cannabis addiction, followed the same gradient, affecting about 67 percent of polymodal users, 47 percent of the smoking-and-vaping group, 36 percent of primary smokers, and 15 percent of those who mainly ate edibles.

“Our findings suggest that using cannabis in multiple forms may be a red flag for heavier use and symptoms of addiction,” said lead author Omar El-Shahawy, MD, PhD, MPH, an assistant professor in the Department of Population Health at NYU Grossman School of Medicine. “Asking patients not only whether but how they use cannabis could help clinicians identify when a fuller assessment may be warranted.”

What the Data Show by the Numbers

• 17,732 adults reporting past-30-day cannabis use formed the analysis sample, drawn from the 2022–2023 National Survey on Drug Use and Health.
• 48.2% mainly smoked; 20.5% smoked and vaped; 16.9% mainly used edibles; 14.4% used nearly every modality.
• 67% of the polymodal group met criteria for cannabis use disorder, versus 15% of the mainly-edibles group.
• Daily use was 2.67 times as likely in the polymodal group as among primary smokers, and 0.49 times as likely among mainly-edibles users, in adjusted analysis.
• Young adults aged 18 to 25 were 3.58 times as likely as adults 50 and older to land in the polymodal group, and 2.52 times as likely to be in the smoking-and-vaping group.

How the Groups Differ Beyond Frequency

The demographic splits were as sharp as the use patterns. Nearly 1 in 3 members of the smoking-and-vaping and polymodal groups were young adults aged 18 to 25, compared with about 1 in 6 primary smokers, the researchers report. More than half of the edibles group were women, and edibles were especially prevalent among college graduates; in the journal’s adjusted analysis, women were 1.78 times as likely as men to fall in the mainly-edibles class.

The survey data sit against a large national backdrop: the study’s introduction cites the most recent federal survey estimate that 15.1 percent of Americans aged 12 and older, roughly 43.8 million people, used cannabis in the past month. Daily cannabis use in the U.S. has now surpassed daily alcohol and cigarette use, earlier federal survey analysis showed.

One finding extends beyond cannabis itself. Primary smokers were more likely to also smoke cigarettes than those who mainly vaped or used edibles, compounding exposure to combustion byproducts. “The strong tie between smoking cannabis and smoking cigarettes is of concern because this subset of people are breathing in a wider range of chemicals via combustion, and their lungs are at risk,” said senior author Joseph J. Palamar, PhD, MPH, a professor in the Department of Population Health at NYU Grossman School of Medicine. The study’s introduction notes that smoking involves combustion exposing users to tar and carbon monoxide, that edibles carry delayed onset that raises the risk of accidental overconsumption, and that dabbing delivers high levels of THC rapidly.

What the Study Cannot Establish

The authors state their own constraints plainly. The analysis is cross-sectional, a snapshot of self-reported behavior, so it cannot establish that using multiple forms causes heavier use or addiction; the same data are consistent with heavier users simply reaching for more formats. The findings rely on participants’ self-reports, and some use may have been underreported. The researchers also did not examine why participants chose particular methods, leaving the motivation behind the patterns unresolved.

“Public health messages about cannabis still focus mostly on smoking, but people now use the drug in many different ways,” Dr. El-Shahawy said. “For younger adults, that means talking about the risks of vaping and combining multiple forms, while for people who prefer edibles, the message is to ‘start low and go slow’ because the effects take time to kick in.”

The team next plans to examine why people choose one method over another, including whether those who favor edibles are using cannabis for pain or sleep. The work was funded by the National Institute on Drug Abuse under grant R01DA055675, and the authors note the funder had no role in the study’s design, analysis, or publication decisions. Alongside Drs. El-Shahawy and Palamar, the research team included Mohsen Abbasi-Kangevari, MD, MPH, Michelle He, and Xiaoying Zheng at NYU Langone, and Neal Doran, PhD, at the University of California, San Diego.

References:

^{1. El-Shahawy, O., Abbasi-Kangevari, M., He, M., Zheng, X., Doran, N., & Palamar, J. J. “Cannabis use modalities are associated with sociodemographic characteristics and patterns of cannabis use among US adults: A latent class analysis” (2026), Addictive Behaviors, Volume 184, 108829. https://doi.org/10.1016/j.addbeh.2026.108829}