Addictive behaviors · 2026
Latent classes of substance use patterns and cannabis use disorder Risk in U.S. Adults during recreational Legalization expansion (2015-2023).
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Kim Solomon, Saad Beshoy, Kelada Anthony
Аннотация
BACKGROUND: Traditional surveillance may not capture heterogeneous patterns of cannabis use disorder (CUD) risk during policy transitions. We used latent class analysis (LCA) to identify substance use patterns associated with varying CUD risk during U.S. recreational cannabis legalization expansion (2015-2023).
METHODS: We analyzed 326,193 U.S. adults aged 18 + from the National Survey on Drug Use and Health (2015-2023, excluding 2021). LCA identified classes based on eight indicators: past-year alcohol and cannabis use disorders, lifetime use of alcohol, cannabis, cocaine, and heroin, and arrest history and suicidal ideation. Model selection evaluated 2-15 class solutions using Bayesian Information Criterion (BIC) and entropy.
RESULTS: Five classes emerged (BIC = 1,598,194, entropy = 0.680): Minimal Substance Use (47.3%), Low-Risk Cannabis Use (36.3%), Alcohol-Dominant Comorbidity (5.6%), Polysubstance/High Legal Involvement (7.3%), and High-Risk CUD (3.4%), with the highest cannabis use disorder (42.1%), alcohol use disorder (82.7%), arrest history (29.4%), and suicidal ideation (25.9%). This class increased from 2.6% (2015-2020) to 5.8% (2022-2023) (slope=+0.46 pp/year, p = 0.010). Daily/near-daily cannabis use showed a severity gradient: 32.2% in High-Risk CUD versus 0.0% in Minimal Substance Use (χ2 = 73,297, p < 0.001). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition period (2015-2019) sensitivity analyses confirmed stability (p = 0.614), indicating observed increases reflect population changes rather than diagnostic criteria evolution.
CONCLUSIONS: LCA identified five classes with distinct CUD risk profiles. The High-Risk CUD class increased during legalization expansion, with frequency patterns validating clinical significance. Class-based surveillance may complement prevalence monitoring during policy transitions, though causal attribution requires longitudinal designs with state-level legalization timing.
METHODS: We analyzed 326,193 U.S. adults aged 18 + from the National Survey on Drug Use and Health (2015-2023, excluding 2021). LCA identified classes based on eight indicators: past-year alcohol and cannabis use disorders, lifetime use of alcohol, cannabis, cocaine, and heroin, and arrest history and suicidal ideation. Model selection evaluated 2-15 class solutions using Bayesian Information Criterion (BIC) and entropy.
RESULTS: Five classes emerged (BIC = 1,598,194, entropy = 0.680): Minimal Substance Use (47.3%), Low-Risk Cannabis Use (36.3%), Alcohol-Dominant Comorbidity (5.6%), Polysubstance/High Legal Involvement (7.3%), and High-Risk CUD (3.4%), with the highest cannabis use disorder (42.1%), alcohol use disorder (82.7%), arrest history (29.4%), and suicidal ideation (25.9%). This class increased from 2.6% (2015-2020) to 5.8% (2022-2023) (slope=+0.46 pp/year, p = 0.010). Daily/near-daily cannabis use showed a severity gradient: 32.2% in High-Risk CUD versus 0.0% in Minimal Substance Use (χ2 = 73,297, p < 0.001). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition period (2015-2019) sensitivity analyses confirmed stability (p = 0.614), indicating observed increases reflect population changes rather than diagnostic criteria evolution.
CONCLUSIONS: LCA identified five classes with distinct CUD risk profiles. The High-Risk CUD class increased during legalization expansion, with frequency patterns validating clinical significance. Class-based surveillance may complement prevalence monitoring during policy transitions, though causal attribution requires longitudinal designs with state-level legalization timing.