This cross-sectional study used latent class analysis (LCA) on 17,732 US adults reporting past 30-day cannabis use from the 2022–2023 National Survey on Drug Use and Health (NSDUH) to identify patterns of cannabis consumption modality (smoking, vaping, edibles, polymodal) and their associations with demographics, use frequency, and cannabis use disorder (CUD).
Four modality classes emerged: primary smoking (48.2%), smoking and vaping (20.5%), mainly edibles (16.9%), and polymodal use (14.4%). Adults aged 18–25 were more likely to be polymodal users (aRRR = 3.58, 95% CI: 2.48–5.18) vs. those ≥50; females favored edibles (aRRR = 1.78, 95% CI: 1.48–2.13). Polymodal users had the highest daily use risk (aRRR = 2.67) and elevated CUD risk; edible-only users had lower daily use (aRRR = 0.49) and lower CUD risk compared to primary smokers.
Cross-sectional design prevents causal inference. Self-reported modality data may undercount stigmatized routes. The survey may not fully capture emerging modalities (e.g., dabbing, tinctures).
Clinicians should ask about *how* patients use cannabis, not just whether they use it — polymodal and combined smoking/vaping patterns are linked to higher CUD risk, especially in younger adults. Edible-predominant users may need different harm-reduction messaging than smokers or vapers.
