Abstract
Introduction.
This study examined US young adults’ cannabis-tobacco use classes.
Methods.
Latent class analysis used 2023 data among young adults (ages 18-34, purposively recruited via Facebook to represent ~50% past-month cannabis use), specifically the 2,267 reporting past-month cannabis and/or tobacco use. Indicators included: cannabis, cigarette, and e-cigarette use (0 days, infrequent [1-10], frequent [11-30]) and any cigar, hookah, smokeless tobacco, and nicotine pouch use. Multivariable regressions examined sociodemographics, adverse childhood events (ACEs), mental health, and personality characteristics in relation to class.
Results.
Five classes were identified: (#1) ‘primarily cannabis’ (36.6%): all used cannabis (74.0% infrequent), <14% tobacco products; (#2) ‘frequent cannabis-cigarette’ (34.2%): 86.7% cannabis (82.9% frequent), 65.9% cigarettes (51.6% frequent), 59.4% e-cigarettes (33.6% frequent), <37% other tobacco; (#3) ‘product-dabbling’ (16.0%): 79.8% cannabis, 71.8% cigarettes, 66.7% e-cigarettes (largely infrequently used each), <40% other tobacco; (#4) ‘frequent poly-product’ (7.7%): 93.9% cannabis, 90.9% cigarettes, 98.2% e-cigarettes (~half frequently used each), >84% other tobacco; and (#5) ‘primarily e-cigarette’ (5.5%): all used e-cigarettes (51.0% frequent), <9% other tobacco. Correlates of class #4 membership were: being Black (vs. White) and more mental health symptoms vs. other classes; being Hispanic vs. #2 and #5; being heterosexual (vs. other) vs. classes #1-#3; being older and male and higher extraversion vs. #1 and #5; in non-legalized states vs. #1; more ACEs vs. #1, #3, and #5; and higher neuroticism and less openness vs. #1-#2.
Conclusions.
The frequent poly-product class represented characteristics (e.g., race/ethnicity, sexual minority, mental health) implicated in substance use related disparities, underscoring the need for targeted intervention.
Keywords: Cannabis, Marijuana, Tobacco, E-cigarettes, Risk factors, Mental health
INTRODUCTION
Between 2012 and 2024, 24 states and Washington, DC legalized nonmedical (‘recreational’) cannabis use, with most jurisdictions establishing nonmedical retail markets that have sold increasingly diverse cannabis products.1 Additionally, the past 15 years have marked a rapidly evolving tobacco retail market in the US, including various products such as e-cigarettes, cigars, hookah, smokeless tobacco (SLT), and nicotine pouches.2 These shifts have coincided with increased cannabis and tobacco use among adults ages 18-34; in those ages 18-25, 26-29, and 30-34, past-month cannabis use prevalence in 2023 was 25.2%, 27.2%, and 23.1%, respectively, and past-month tobacco use prevalence was 15.7%, 20.9%, and 22.6%.3 Cannabis is traditionally smoked in the form of dried herb and can also be infused in food (e.g., edibles), tinctures, lotions and balms, or vaporized as oils and concentrates (i.e., wax).3 Similarly, combustible tobacco use (e.g., cigarettes, cigars, pipes) is common but has decreased among young adults in recent years, while alternative tobacco product use, such as e-cigarettes/vapes, has surged.2 The increasing diversification of cannabis and tobacco products raises concerns around young adult co-use or poly-product use,4 which is associated with higher frequency and intensity of use.5,6
Psychosocial factors (e.g., mental health conditions, adverse childhood experiences [ACEs], personality traits) have been associated with cannabis and tobacco use patterns, frequency, and trajectories.7-12 These relationships may be bidirectional, as psychosocial factors can increase vulnerability to substance use and multiple product experimentation, and greater cannabis and tobacco product use are linked to adverse mental and physical health outcomes (i.e., substance use disorder).7-12 Mental health symptoms (i.e., anxiety, depressive) are widely documented to have reciprocal associations with tobacco and cannabis use.7-12 ACEs (i.e., traumatic experiences during childhood [≤18 years old] involving violence, abuse, neglect, or household dysfunction13) are linked to substance use, including tobacco and cannabis use, use characteristics (e.g., early use, chronic use), and potential determinants of substance use (e.g., impulsivity, dysregulated stress).13-15 Further, ACEs are associated with diagnoses of anxiety and depressive disorders, further compounding risks of substance use and poly-product use.13,14 Certain personality traits have also been linked to both cannabis and tobacco use. For example, higher neuroticism (e.g., emotional instability, stress sensitivity) is associated with increased likelihood of both cannabis16 and tobacco use,17 lower agreeableness has been associated with tobacco use, and lower conscientiousness (e.g., self-discipline) has shown associations with both cannabis and tobacco use.18 While higher extraversion and openness to experiences (e.g., adventurousness, creativity) have shown relationships with tobacco use,18 findings are mixed with regard to cannabis use.15,19,20
Given the diversity of cannabis and tobacco products, person-centered approaches, like latent class analysis (LCA), are strategically suited to assess these use behavior profiles.5,21 A 2023 study of cannabis and tobacco use among youth (ages 15-17) and young adults (ages 18-24) across states in the US identified 6 distinct classes among young adults (i.e., cigarettes, cigarettes/cigars, cannabis, blunts, cigarettes/e-cigarettes/cannabis-vaping), as well as sociodemographic differences in use patterns (e.g., cigarettes/cigars more prevalent in Black young adults, dual/poly-use more common in Whites).22 Two 2016 LCA studies of college students ages 18-25 – one in Georgia23 and one in Pennsylvania24 (both states where nonmedical cannabis was/is not legalized) found 3 cannabis-tobacco use classes (light poly-tobacco, small-cigar/hookah/cannabis, heavy poly-tobacco;23 non-hookah tobacco, hookah/cannabis, poly-use24).
These and other prior studies using LCA to understand cannabis and tobacco use profiles inform the current study. However, these studies included restricted young adult age ranges, and they largely have failed to include (or account for) variability in legal nonmedical cannabis status.22-24 Moreover, relatively few studies have accounted for cannabis use, the vast array of tobacco products, and certain use characteristics, specifically use frequency.22,25-27 Furthermore, prior research has largely focused on sociodemographic characteristics of use, with few assessing key psychosocial factors, like ACEs, mental health symptoms, or personality characteristics.22,25-27 This study aimed to advance our understanding of cannabis-tobacco use behaviors, including use frequency, among young adults of a broader age range (i.e., 18-34) across cannabis regulator contexts, and assess how key psychosocial factors are associated with different cannabis-tobacco use profiles. Specifically, this study identified cannabis and tobacco use classes among US young adults who reported any past-month cannabis and/or tobacco use (accounting for use frequency of those most prevalent), as well as their psychosocial and sociodemographic correlates, accounting for nonmedical cannabis legalization. We hypothesized that classes representing high levels of use of various products would report the most mental health symptoms and ACEs, as well as certain personality characteristics (e.g., lower conscientiousness, agreeableness, and emotional stability; higher extraversion and openness), while classes representing lower level single product use would report the least mental health symptoms and ACEs, as well as distinct personality characteristics.
METHODS
Study Design
This study analyzed survey data among 4,031 young adults across all 50 states and DC who participated in the Cannabis Regulation, Marketing & Appeal (CARMA) study, a longitudinal study examining nonmedical cannabis laws, marketing, and use (approved by the George Washington University Institutional Review Board).28
Participants and Recruitment
Ads targeting eligible young adults (English-speaking US residents ages 18-34, i.e., the age group with the highest cannabis use prevalence)3 were posted on Facebook in June-November 2023. After clicking on ads, individuals were messaged via chatbot on Facebook Messenger (which verified Facebook accounts and precluded duplicates). The chatbot provided an abbreviated study overview, assessed key factors (age, state of residence, race, ethnicity, sex, past-month cannabis use), and provided individuals deemed preliminarily eligible a unique link to the study webpage. There, formal consent was obtained, eligibility was confirmed, and the baseline survey was administered. Participants were told that the study required providing valid email addresses and phone numbers and confirming their participation by clicking a link in an email sent 7 days post-baseline survey (after which they received their incentive, a $10 Amazon e-gift card). Purposive, quota-based recruitment was used to ensure representation of key subgroups (i.e., ~50% past-month cannabis use, ~50% males and females, ~40% racial/ethnic minorities). Overall, 6,908 individuals completed chatbot pre-screening, 6,128 (88.7%) were preliminarily eligible, 5,827 (95.6%) visited the study webpage, 5,672 (97.3%) were consented and eligible, 4,385 (77.3%) completed the survey and were sent confirmation emails, and 4,031 (91.9%) confirmed participation. Current analyses focused on the 2,267 (56.2%) participants who reported past-month use of cannabis, cigarettes, e-cigarettes, cigars, hookah, SLT, or nicotine pouches.
Measures
Cannabis and tobacco use.
LCA was based on variables indicating cannabis, cigarette, e-cigarette, cigar, hookah, SLT, and nicotine pouch use. Participants were provided product descriptions and photos, then asked, “In the past 30 days, how many days did you use: marijuana? traditional cigarettes? e-cigarettes or other electronic nicotine delivery devices that vape nicotine? large or little cigars? hookah, waterpipe, or nargila? SLT (e.g., chew, snus)? tobacco-free nicotine pouches?”3 Variable categorizations were based on distributions; in the overall sample (n=4,031), any past-month use was reported by 48.8% for cannabis, 22.7% cigarettes, 26.6% e-cigarettes, 13.4% cigars, 11.8% hookah, 4.7% SLT, and 5.3% nicotine pouches. Thus, cigars, hookah, SLT, and nicotine pouches were categorized as 0 vs. >0 days, and cannabis, cigarette, and e-cigarette use were categorized as 3-category ordinal variables (0 days, 1-10 days, 11-30 days), based on distributions: roughly half using in the past-month used 1-10 days (50.9% cannabis, 40.6% cigarettes, 55.6% e-cigarettes) vs. 11-30 days (49.1% cannabis, 59.4% cigarettes, 44.4% e-cigarettes).
For descriptive purposes, participants reporting past-month cannabis use were asked: 1) “How do you use marijuana most of the time?” a) dried herb; b) edibles; c) oils (i.e., cannabis oils or liquids for vaping, cannabis oils or liquids taken orally, tinctures); or d) concentrates/other (i.e., concentrates, hash or kief, topical, other); 2) “Do you currently have a medical marijuana card?” (yes/no); and 3) “Currently, do you use marijuana for medical or recreational purposes – or both?” (response options: only medical; primarily medical but occasionally recreational; equally for medical and recreational; primarily recreational but occasionally medical; only recreational; unsure).
Psychosocial factors.
The ACEs-10 item scale assessed maltreatment (psychological, physical sexual) by an adult household member), neglect (of emotional or physical needs), parental separation/divorce, domestic violence, and household member mental health problems, substance use problems, or imprisonment before age 18 (0=no, 1=yes; sum score range 0-10; Cronbach’s alpha=.81).29-31 Mental health symptoms were assessed using the Patient Health Questionnaire – 4 item (PHQ-4), which includes 2 items assessing depressive symptoms and 2 assessing anxiety symptoms in the past 2 weeks (0=not at all to 3=nearly every day), operationalized as a total sum score (range 0-12; Cronbach’s alpha=.89).32 Personality characteristics were assessed using the 10-Item Personality Inventory, which has 2 items each measuring extraversion, openness, neuroticism, conscientiousness, and agreeableness, using the stem, “I see myself as …” (1=strongly disagree to 7=strongly agree; mean subscale scores range 1-7; Cronbach’s alpha range=.35-.72, similar to prior reports).33
State nonmedical cannabis law and sociodemographics.
Participants reported state of residence, age, birth sex, sexual orientation, ethnicity, race, education level, relationship status, and whether they had children. State of residence was coded as legalized vs. not legalized nonmedical cannabis.
Data Analysis
First, descriptive analyses were conducted to characterize participants and examine response distributions. Second, LCA (using product use frequency categories for cannabis, cigarettes, and e-cigarettes and any past-month use status for cigars, hookah, SLT, and nicotine pouches) identified cannabis use classes among those reporting past-month use.34 We examined latent class solutions for models with 1–6 classes, determining the best-fitting model based on: Akaike information criterion (AIC), Bayesian information criterion (BIC), and entropy values. Lower values of AIC and BIC, and larger values of entropy indicated better model fit.35 We also used the Lo-Mendell-Rubin Adjusted Likelihood Ratio Test (LRT) to compare models with K classes to models with K-1 classes; significant p-values indicated better fit for the model with K classes.35 Other considerations included smallest class (>5%) and class interpretability. Robust Maximum Likelihood was used. Participants were categorized based on their most probable class. Participants were assigned to classes based on most likely class membership.
Third, bivariate analyses (Chi-square tests for categorical variables, ANOVAs or t-tests for continuous) characterized participants in relation to cannabis-tobacco use class. Fourth, state nonmedical cannabis legalization, sociodemographics, and psychosocial factors were examined in relation to cannabis-tobacco use class (multinomial logistic regression with pairwise comparison). Analyses were conducted in SPSS.v27 and R.v4.5.
RESULTS
Participant Characteristics
Shown in Table 1, the sample was 26.87 (SD=4.60) years old on average, 40.9% male, 31.8% sexual minority, 20.4% Hispanic, 64.2% White, 17.2% Black, 9.0% Asian, and 7.9% other race(s). Overall, 34.9% had ≥bachelor’s degree, 19.0% were married, 22.4% were cohabitating, and 35.9% had children.
Table 1.
US young adult participants’ sociodemographic and use characteristics, overall and by cannabis-tobacco use class (N=2,267)
| Variables | Participants reporting any past- month cannabis or tobacco use |
Cannabis-tobacco use class** | |||||
|---|---|---|---|---|---|---|---|
| Primarily cannabis |
Frequent cannabis cigarette |
Product dabbling |
Frequent poly- product |
Primarily e-cigarette |
|||
| N=2,267 (100%) |
N=899 (39.7%) |
N=768 (33.9%) |
N=312 (13.8%) |
N=164 (7.2%) |
N=124 (5.5%) |
||
| N (%) or M (SD) |
N (%) or M (SD) |
N (%) or M (SD) |
N (%) or M (SD) |
N (%) or M (SD) |
N (%) or M (SD) |
p | |
| State legal nonmedical cannabis, n (%)* | 1142 (50.4) | 528 (58.7) | 341 (44.4) | 145 (46.5) | 75 (45.7) | 53 (42.7) | <.001 |
| Sociodemographics | |||||||
| Age, M (SD) | 26.87 (4.60) | 26.10 (4.65) | 27.74 (4.46) | 26.80 (4.62) | 28.01 (3.98) | 25.75 (4.63) | <.001 |
| Male, n (%) | 927 (40.9) | 325 (36.2) | 306 (39.8) | 170 (54.5) | 88 (53.7) | 38 (30.6) | <.001 |
| Sexual minority, n (%) | 720 (31.8) | 325 (36.2) | 249 (32.4) | 85 (27.2) | 26 (15.9) | 35 (28.2) | <.001 |
| Hispanic, n (%) | 463 (20.4) | 203 (22.6) | 134 (17.4) | 66 (21.2) | 45 (27.4) | 15 (12.1) | <.001 |
| Race, n (%) | <.001 | ||||||
| White | 1415 (62.4) | 545 (60.6) | 510 (66.4) | 186 (59.6) | 91 (55.5) | 83 (66.9) | |
| Black | 390 (17.2) | 128 (14.2) | 150 (19.5) | 55 (17.6) | 43 (26.2) | 14 (11.3) | |
| Asian | 204 (9.0) | 111 (12.3) | 24 (3.1) | 38 (12.2) | 14 (8.5) | 17 (13.7) | |
| Other | 178 (7.9) | 74 (8.2) | 60 (7.8) | 25 (8.0) | 12 (7.3) | 7 (5.6) | |
| Education ≥bachelor’s degree, n (%) | 792 (34.9) | 391 (43.5) | 150 (19.5) | 129 (41.3) | 79 (48.2) | 43 (34.7) | <.001 |
| Relationship status, n (%) | <.001 | ||||||
| Single/never married/other | 1330 (58.7) | 541 (60.2) | 435 (56.6) | 201 (64.4) | 85 (51.8) | 68 (54.8) | |
| Married | 430 (19.0) | 171 (19.0) | 116 (15.1) | 58 (18.6) | 55 (33.5) | 30 (24.2) | |
| Cohabitating | 507 (22.4) | 187 (20.8) | 217 (28.3) | 53 (17.0) | 24 (14.6) | 26 (21.0) | |
| Parent/has child(ren), n (%) | 813 (35.9) | 230 (25.6) | 356 (46.4) | 102 (32.7) | 80 (48.8) | 45 (36.3) | <.001 |
| Past-month substance use, n (%) | |||||||
| Cannabis | <.001 | ||||||
| 0 days | 299 (13.2) | 0 (0.0) | 102 (13.3) | 63 (20.2) | 10 (6.1) | 124 (100) | |
| 1-10 days | 1002 (44.2) | 665 (74.0) | 29 (3.8) | 249 (79.8) | 59 (36.0) | 0 (0.0) | |
| 11-30 days | 966 (42.6) | 234 (26.0) | 637 (82.9) | 0 (0.0) | 95 (57.9) | 0 (0.0) | |
| Cigarettes | <.001 | ||||||
| 0 days | 1351 (59.6) | 872 (97.0) | 262 (34.1) | 88 (28.2) | 15 (9.1) | 114 (91.9) | |
| 1-10 days | 372 (16.4) | 0 (0.0) | 110 (14.3) | 178 (57.1) | 74 (45.1) | 10 (8.1) | |
| 11-30 days | 544 (24.0) | 27 (3.0) | 396 (51.6) | 46 (14.7) | 75 (45.7) | 0 (0.0) | |
| E-cigarettes | <.001 | ||||||
| 0 days | 1196 (52.8) | 777 (86.4) | 312 (40.6) | 104 (33.3) | 3 (1.8) | 0 (0.0) | |
| 1-10 days | 595 (26.2) | 80 (8.9) | 198 (25.8) | 162 (51.9) | 96 (58.5) | 59 (47.6) | |
| 11-30 days | 476 (21.0) | 42 (4.7) | 258 (33.6) | 46 (14.7) | 65 (39.6) | 65 (52.4) | |
| Cigars (any) | 541 (23.9) | 0 (0.0) | 280 (36.5) | 122 (39.1) | 139 (84.8) | 0 (0.0) | <.001 |
| Hookah (any) | 474 (20.9) | 24 (2.7) | 214 (27.9) | 87 (27.9) | 149 (90.9) | 0 (0.0) | <.001 |
| Smokeless tobacco (any) | 191 (8.4) | 9 (1.0) | 11 (1.4) | 29 (9.3) | 142 (86.6) | 0 (0.0) | <.001 |
| Nicotine pouches (any) | 214 (9.4) | 0 (0.0) | 24 (3.1) | 44 (14.1) | 141 (86.0) | 5 (4.0) | <.001 |
| Psychosocial factors, M (SD) | |||||||
| ACEs | 3.41 (2.86) | 2.99 (2.67) | 3.99 (2.92) | 2.85 (2.85) | 4.18 (3.19) | 3.20 (2.62) | <.001 |
| Mental health (PHQ-4) | 4.24 (3.46) | 3.87 (3.30) | 4.44 (3.59) | 4.21 (3.42) | 5.43 (3.57) | 4.15 (3.38) | <.001 |
| Personality characteristics | |||||||
| Extraversion | 3.84 (1.55) | 3.72 (1.59) | 3.88 (1.57) | 4.06 (1.41) | 4.14 (1.28) | 3.50 (1.66) | <.001 |
| Agreeableness | 4.88 (1.17) | 4.90 (1.19) | 4.98 (1.14) | 4.80 (1.14) | 4.53 (1.15) | 4.88 (1.22) | <.001 |
| Conscientiousness | 4.99 (1.37) | 4.97 (1.40) | 5.03 (1.39) | 5.07 (1.26) | 4.73 (1.33) | 5.02 (1.37) | .087 |
| Neuroticism | 4.05 (1.45) | 3.98 (1.40) | 3.99 (1.54) | 4.29 (1.32) | 4.31 (1.41) | 4.03 (1.52) | .002 |
| Openness to new experiences | 5.24 (1.23) | 5.27 (1.19) | 5.40 (1.21) | 5.08 (1.26) | 4.83 (1.31) | 5.10 (1.27) | <.001 |
Notes: M: mean. SD: standard deviation. P-values based on Chi-square tests for categorical variables; ANOVAs or t-tests for continuous. Bold indicates significance (p<.05). *Legalized nonmedical adult use as of March 2023: Alaska, Arizona, California, Colorado, Connecticut, District of Columbia, Illinois, Maine, Maryland, Massachusetts, Michigan, Missouri, Montana, Nevada, New Jersey, New Mexico, New York, Oregon, Rhode Island, Vermont, Virginia, Washington. Responses categorized separately for multivariable regressions. Sex: other (n=1), refuse (n=24). Sexual identity: refuse (n=82). Ethnicity: DK (n=9), refuse (n=52). Race: DK (n=48), refuse (n=95). Community: other (n=4). **Cannabis form used – Primarily cannabis class (n=899): bud/flower (n=405, 45.1%), edibles (n=283, 31.5%), vaped (n=120, 13.3%), and high-potency (concentrates, tinctures, hash/keif; n=37, 5.3%). Frequent cannabis-cigarette class reporting cannabis use (n=666): bud/flower (n=493, 74.0%), edibles (n=58, 8.7%), vaped (n=57, 8.6%), and high-potency (n=44, 6.7%). Product dabbling class reporting cannabis use (n=249): bud/flower (n=139, 55.8%), edibles (n=44, 17.7%), vaped (n=30, 12.0%), and high-potency (n=31, 8.4%). Frequent poly-product reporting cannabis use (n=154): bud/flower (n=65, 42.2%), high-potency (n=26, 16.8%), vaped (n=14, 9.1%), and edibles (n=14, 9.1%).
LCA Among Participants Reporting Past-Month Cannabis or Tobacco Use
The 5-class solution was chosen based on model fit indices and conceptual/theoretical interpretability (Table 2; Figure 1). The AIC and BIC were lower for the 5-class solution relative to 1-4 class solutions. Although the AIC and BIC were lower for the 6- vs. 5-class solution, these differences were minimal. Additionally, class sizes became small (representing <5% of the sample) for solutions greater than 5 classes.
Table 2.
Model fit characteristics for latent class analysis (LCA) assessing cannabis-tobacco use (N=2,267)*
| Model fit | Class sizes | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| # of classes |
# of Parameters Estimated |
Residual DF |
Maximum Log- Likelihood |
AIC | BIC | LR/ Deviance |
Chi- squared |
Entropy | Adjusted LRT^ |
Class 1 | Class 2 | Class 3 | Class 4 | Class 5 | Class 6 |
| 1 | 10.00 | 421.00 | −10471.37 | 20962.74 | 21020.00 | 3060.17 | 75262.27 | − | - | 2267 (1.0) | |||||
| 2 | 21.00 | 410.00 | −9639.03 | 19320.06 | 19440.31 | 1393.02 | 1679.37 | 0.74 | <.001 | 1799 (.79) | 468 (.21) | ||||
| 3 | 32.00 | 399.00 | −9415.82 | 18895.64 | 19078.88 | 947.55 | 1259.25 | 0.65 | <.001 | 1237 (.55) | 830 (.37) | 200 (.09) | |||
| 4 | 43.00 | 388.00 | −9341.11 | 18768.22 | 19014.45 | 798.51 | 998.38 | 0.71 | <.001 | 750 (.33) | 687 (.30) | 634 (.28) | 196 (.09) | ||
| 5 | 54.00 | 377.00 | −9286.50 | 18681.00 | 18990.22 | 691.17 | 879.96 | 0.56 | <.001 | 899 (.40) | 768 (.34) | 312 (.14) | 164 (.07) | 124 (.06) | |
| 6 | 65.00 | 366.00 | −9243.82 | 18617.63 | 18989.84 | 605.00 | 801.44 | 0.63 | <.001 | 872 (.39) | 641 (.28) | 470 (.21) | 111 (.05) | 97 (.04) | 76 (.03) |
Notes: *Indicator variables in LCA: number of days of use of cannabis, cigarettes, and e-cigarettes (categorized as 0, 1-10, 11-30 days), and any vs. no use of cigars, hookah, smokeless tobacco, and nicotine pouches in the past 30 days. ^Vuong–Lo-Mendell-Rubin likelihood ratio test (VLMR-LRT).
Figure 1. Latent classes of cannabis and tobacco use among US young adults reporting any past 30-day use of cannabis or tobacco (N=2,267).
The 5-class solution consisted of classes representing: (1) ‘primarily cannabis’ (36.6%): all used cannabis but most used infrequently (74.0% 1-10 days) and <14% used tobacco products (3.0% cigarettes [all 11-30 days], 13.6% e-cigarettes [8.9% 1-10 days], 2.7% hookah, 1.0% SLT, 0% cigars or nicotine pouches); (2) ‘frequent cannabis-cigarette’ (34.2%): 86.7% used cannabis (82.9% using 11-30 days), 65.9% cigarettes (51.6% using 11-30 days), 59.4% e-cigarettes (33.6% 11-30 days), and <40% other tobacco products (36.5% cigars, 27.9% hookah, 1.4% SLT, 3.1% nicotine pouches); (3) ‘product-dabbling’ (16.0%): 79.8% used cannabis (all using 1-10 days), 71.8% cigarettes (57.1% 1-10 days), 66.7% e-cigarettes (51.9% 1-10 days), 39.1% cigars, 27.9% hookah, 9.3% SLT, and 14.1% nicotine pouches; (4) ‘frequent poly-product’ (7.7%): 93.9% used cannabis (57.9% 11-30 days), 90.9% cigarettes (~45% 1-10 and 11-30 days), 98.2% e-cigarettes (58.5% 11-30 days), and >84% other tobacco products (84.8% cigars, 90.9% hookah, 86.6% SLT, 86.0% nicotine pouches); and (5) ‘primarily e-cigarette’ (5.5%): all used e-cigarettes (with ~50% using 1-10 and 11-30 days, respectively), 8.1% cigarettes (all 1-10 days), 4.0% nicotine pouches, and 0% cannabis, cigars, hookah, or SLT.
To further characterize cannabis use behaviors across classes, among those with past-month cannabis use, the largest proportion smoked bud/flower (range: 42.2% in frequent poly-product to 74.0% in frequent cannabis-cigarette). The primarily cannabis class had particularly prominent use of edibles (31.5% vs. 17.7% in others) and vaped oils (13.3% vs. 12.0% in others), and the frequent poly-product class reported prominent use of concentrates/tinctures/hash/kief (16.8% vs. 8.4% in others, p’s<.001; see Table 1 footnote for details).
Sociodemographic Characteristics of Latent Classes
Table 1 shows bivariate analyses characterizing the sociodemographic, substance use, and psychosocial characteristics of each class. In bivariate analyses, all associations were significant, except conscientiousness (p=.087; see Table 1 for details).
Based on the hypothesis that the classes reporting poly-product use and more frequent use would display higher psychosocial risk factors, multinomial regression first compared the frequent poly-product use class to the other 4 classes (Table 3), then other group comparisons (Supplementary Table 1), specifically the frequent cannabis-cigarette use class to the remaining 3 classes, the primarily cannabis class to the product dabbling and primarily e-cigarette classes, and the product dabbling class to the primarily e-cigarette class.
Table 3.
Multinomial logistic regression assessing sociodemographic correlates of cannabis-tobacco use classes among US young adults reporting past 30-day use of any cannabis or tobacco product (N=2,267)
| Variables | Frequent poly-product (vs. Primarily cannabis) |
Frequent poly-product (vs. Frequent cannabis-cigarette) |
Frequent poly-product (vs. Product dabbling) |
Frequent poly-product (vs. Primarily e-cigarette) |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| aOR | 95% CI | p | aOR | 95% CI | p | aOR | 95% CI | p | aOR | 95% CI | p | |
| State nonmedical cannabis law | ||||||||||||
| Legalized (ref: not legalized) | 0.57 | 0.39, 0.85 | .005 | 0.85 | 0.57, 1.26 | .416 | 0.94 | 0.61, 1.45 | .772 | 1.06 | 0.62, 1.82 | .822 |
| Sociodemographics | ||||||||||||
| Age | 1.06 | 1.01, 1.12 | .012 | 1.01 | 0.96, 1.05 | .846 | 1.04 | 0.99, 1.10 | .131 | 1.11 | 1.04, 1.19 | .001 |
| Male sex (ref: female) | 2.76 | 1.78, 4.28 | <.001 | 1.27 | 0.82, 1.96 | .285 | 1.18 | 0.73, 1.90 | .506 | 2.78 | 1.52, 5.09 | <.001 |
| Sexual minority (ref: heterosexual) | 0.41 | 0.25, 0.69 | <.001 | 0.49 | 0.29, 0.82 | .007 | 0.49 | 0.28, 0.86 | .012 | 0.68 | 0.35, 1.34 | .264 |
| Hispanic (ref: non-Hispanic) | 1.25 | 0.80, 1.97 | .330 | 2.03 | 1.28, 3.22 | .003 | 1.38 | 0.83, 2.28 | .216 | 2.55 | 1.24, 5.26 | .011 |
| Race (ref: White) | ||||||||||||
| Black | 2.78 | 1.70, 4.52 | <.001 | 2.20 | 1.37, 3.54 | .001 | 2.00 | 1.17, 3.42 | .011 | 3.70 | 1.75, 7.84 | <.001 |
| Asian | 0.89 | 0.42, 1.90 | .761 | 2.34 | 0.97, 5.65 | .058 | 0.80 | 0.36, 1.80 | .589 | 0.55 | 0.22, 1.40 | .213 |
| Other | 0.96 | 0.42, 2.19 | .930 | 0.99 | 0.44, 2.25 | .988 | 0.73 | 0.31, 1.76 | .487 | 1.42 | 0.43, 4.68 | .567 |
| Education ≥bachelor’s degree (ref: <) | 1.17 | 0.76, 1.80 | .473 | 2.98 | 1.92, 4.61 | <.001 | 1.31 | 0.81, 2.09 | .270 | 1.52 | 0.84, 2.76 | .165 |
| Relationship (ref: single/other) | ||||||||||||
| Married | 1.32 | 0.80, 2.19 | .283 | 2.17 | 1.30, 3.63 | .003 | 1.94 | 1.09, 3.43 | .024 | 1.25 | 0.62, 2.51 | .538 |
| Cohabitating | 1.02 | 0.59, 1.76 | .942 | 0.83 | 0.48, 1.41 | .486 | 1.12 | 0.61, 2.04 | .713 | 0.90 | 0.44, 1.82 | .764 |
| Parent/has child(ren) (ref: no) | 2.37 | 1.48, 3.78 | <.001 | 1.06 | 0.67,1.67 | .817 | 1.49 | 0.89, 2.51 | .128 | 1.31 | 0.69, 2.50 | .410 |
| Psychosocial factors | ||||||||||||
| ACEs | 1.17 | 1.08, 1.25 | <.001 | 1.07 | 1.00, 1.15 | .060 | 1.17 | 1.08, 1.26 | <.001 | 1.13 | 1.02, 1.24 | .019 |
| Mental health (PHQ-4) | 1.21 | 1.13, 1.29 | <.001 | 1.15 | 1.07, 1.22 | <.001 | 1.10 | 1.02, 1.18 | .011 | 1.17 | 1.07, 1.29 | <.001 |
| Personality characteristics | ||||||||||||
| Extraversion | 1.25 | 1.09, 1.44 | .002 | 1.13 | 0.98, 1.30 | .084 | 1.06 | 0.91, 1.23 | .465 | 1.33 | 1.10, 1.60 | .003 |
| Agreeableness | 0.84 | 0.70, 1.01 | .069 | 0.78 | 0.65, 0.94 | .009 | 0.82 | 0.67, 1.01 | .056 | 0.82 | 0.63, 1.05 | .112 |
| Conscientiousness | 0.92 | 0.79, 1.08 | .327 | 0.91 | 0.78, 1.07 | .269 | 0.91 | 0.76, 1.08 | .275 | 0.88 | 0.71, 1.09 | .232 |
| Neuroticism | 1.40 | 1.18, 1.65 | <.001 | 1.41 | 1.19, 1.66 | <.001 | 1.20 | 1.00, 1.44 | .053 | 1.25 | 1.00, 1.57 | .051 |
| Openness to new experiences | 0.78 | 0.65, 0.93 | .006 | 0.74 | 0.62, 0.88 | <.001 | 0.94 | 0.78, 1.15 | .564 | 0.91 | 0.72, 1.15 | .416 |
Notes: aOR: adjusted odds ratio. CI: confidence interval. Nagelkerke R-square=.276. Bold indicates significance (p<.05).
Frequent poly-product use vs. all other classes.
In multinomial logistic regression (Table 3), the frequent poly-product class reported more mental health symptoms than all other classes, more ACEs than others except the frequent cannabis-cigarette class, higher levels of extraversion than the primarily cannabis and primarily e-cigarette classes, higher neuroticism and lower openness than the primarily cannabis and frequent cannabis-cigarette classes, and lower agreeableness than the frequent cannabis-cigarette class.
In terms of sociodemographics, the frequent poly-product class was more likely Black (vs. White) compared to other classes; more likely Hispanic than the frequent cannabis-cigarette and primarily e-cigarette classes; older and more likely male (vs. female) than the primarily cannabis and primarily e-cigarette classes; less likely sexual minority (vs. heterosexual) than the primarily cannabis, frequent cannabis-cigarette, and product dabbling classes; more likely ≥bachelor’s degree educated than frequent cannabis-cigarette; more likely married (vs. single) than the frequent cannabis-cigarette and product dabbling classes; more likely to have children than the primarily cannabis class; and less likely in legalized states than the primarily cannabis class.
Frequent cannabis-cigarette vs. remaining 3 classes.
Shown in Supplementary Table 1 (top set of results), the frequent cannabis-cigarette use class reported more ACEs than the primarily cannabis and product dabbling classes; more mental health symptoms than the primarily cannabis class; higher extraversion than the primarily cannabis and primarily e-cigarette classes; higher openness compared to the product dabbling and primarily e-cigarette classes; and lower neuroticism than the product dabbling class.
Regarding sociodemographics, the frequent cannabis-cigarette class was older and less likely Asian and ≥bachelor’s degree educated than the 3 other classes; more likely male than the primarily cannabis and primarily e-cigarette classes; less likely Hispanic than primarily cannabis and product dabbling; less likely married than primarily cannabis; more likely to have children than primarily cannabis; and less likely to live in a state with legal nonmedical cannabis compared to primarily cannabis.
Comparisons among primarily cannabis, product dabbling, and primarily e-cigarette.
Shown in Supplementary Table 1 (bottom set of results), the primarily cannabis class reported lower mental health symptoms and extraversion and more openness to experiences than the product dabbling class. The primarily cannabis class was also less likely male than product dabbling; more likely sexual minority and Hispanic than primarily e-cigarette; and more likely to have children and to live in a state with legalized nonmedical cannabis than the product dabbling and primarily e-cigarette classes.
Compared to the primarily e-cigarette class, the product-dabbling class was older, more likely male, and more extraverted.
To further explore whether different types of ACEs were associated with use class, we conducted Chi-square analyses on the individual ACE items (Supplementary Table 2). The frequent poly-product use class had the highest representation of individuals reporting each ACE, except in 3 instances where the frequent cannabis-cigarette use class had the highest representation (for emotional abuse, parental separation/divorce, and household substance use problems).
DISCUSSION
This LCA study of US young adults identified 5 cannabis-tobacco use classes representing varied products used and use frequency, as well as unique psychosocial and sociodemographic profiles. Findings showed similarities to classes found in other studies (poly-product,22-24 primarily cannabis,22 primarily e-cigarettes22) but also distinct classes (e.g., frequent cannabis-cigarette). These distinctions may be related to including frequency of cannabis, cigarette, and e-cigarette use, as well as past-month use status various other tobacco products (unlike many previous LCA studies assessing cannabis and tobacco use),22-27 or to characteristics of the sample, which was recruited via social media in 2023 to represent ~50% reporting cannabis use and ~50% living in states with legalized nonmedical cannabis. That is, the more recent cannabis regulatory context and the proportion residing in legalized states and using cannabis may have impacted the use profiles identified.
As hypothesized, a particularly high-risk class emerged – the frequent poly-product use class. Mental health symptoms and neuroticism were highest in this class, followed by the frequent cannabis-cigarette class, primarily cannabis, and product dabbling. Similarly, the number of ACEs was highest among the frequent poly-product use and frequent cannabis-cigarette classes, followed by the primarily cannabis and product dabbling classes. Moreover, bivariate findings indicated that the frequent poly-product use class had the highest representation of 7 ACEs, while the frequent cannabis-cigarette use class had the highest representation of 3 (emotional maltreatment, parental separation/divorce, household substance use), which coincides with previous research linking distinct ACE profiles to different substance use outcomes.14 Notably, the class with the least ACEs and mental health symptoms was the e-cigarette class, about half of whom used ≤10 days in the past month and few using other products. These findings align with the hypothesis that more mental health symptoms9-12 (and a related factor – neuroticism16,17) and more ACEs are associated with using multiple substances and more frequently.13-15 The most extraverted class was frequent poly-product use, then the frequent cannabis-cigarette and product dabbling classes. The frequent cannabis-cigarette and primarily cannabis classes reported the highest openness, adding to the limited research documenting such associations (typically with single product use).18,19 Agreeableness showed one significant difference (i.e., lower among the frequent poly-product vs. frequent cannabis-cigarette class), and conscientiousness did not distinguish any use class, which differs from prior research showing associations between these factors and single product use.18 These differences may be due to most prior studies focusing on single product use (rather than classes of cannabis-tobacco use) and highlight the advantages of person-centered approaches in elucidating important nuances in use behaviors and related factors.
Notably, compared to the 4 other classes, the primarily cannabis class was more likely to live in states with legalized nonmedical cannabis, suggesting that young adults in these states may opt for cannabis versus tobacco products within such legal contexts. A 2019 review36 and more recent research using survey data37 and tax receipt data38 have not shown such trends post-legalization of nonmedical cannabis; however, these prior studies looked at national population-level trends – rather than trends among young adults, who may differ from older adults due to cohort effects (i.e., entering adolescence and young adulthood with greater likelihood of nonmedical cannabis legalization). Studies focused on young adults have failed to account for potential displacement of tobacco with cannabis, thus highlighting opportunities for future research. Perhaps relatedly, the primarily cannabis and primarily e-cigarette classes were younger and more likely female, while the other 3 classes were older and more likely male. These findings may reflect the evolving retail context, as younger populations have had greater access to cannabis and e-cigarettes earlier in their lives, and diverse product designs and flavors may appeal to younger people and females.5,6 Furthermore, as shown in the present study, existing literature has documented the association between male sex and poly-substance use.39
Certain trends related to racial/ethnic status, education, and sexual orientation resonate with known disparities in single-product and poly-substance use3 but also highlight important nuances, likely elucidated by using a person-centered approach to identify cannabis-tobacco use classes. Regarding race, the frequent poly-product class was more likely Black compared to other classes and more likely Hispanic than the primarily e-cigarette class. Underscoring the particular trend of cigarette use among non-Hispanic White young adults, relative to most classes, the frequent cannabis-cigarette class was less likely Asian or Hispanic.22 In terms of education and sexual orientation, the highest-risk class – the frequent poly-product use class – was generally the most likely college-educated and least likely sexual minority, which somewhat contradicts a previous review showing higher rates of frequent poly-product use among sexual minority groups.39 Instead, the primarily cannabis, frequent cannabis-cigarette, and product dabbling classes were more likely sexual minority than the frequent poly-product class; additionally, the primarily cannabis class was more likely sexual minority than the primarily e-cigarette class. Future research is needed to replicate these findings and assess potential mechanisms of use among certain subpopulations.
Regarding social roles, the frequent poly-product class was more likely married than the frequent cannabis-cigarette and product dabbling classes, and the primarily cannabis class was more likely to have children than the product dabbling and primarily e-cigarette classes. Interestingly, the primarily cannabis class was more likely married than the frequent cannabis-cigarette class, but frequent cannabis-cigarette was more likely to have children than primarily cannabis. In general, these findings are difficult to interpret and contradict some literature suggesting substance use generally decreases as individuals transition to such social roles (e.g., spouse, parent).40,41 However, this study did not assess partner substance use status, which is a crucial influence on substance use dynamics within marital relationships.42 Furthermore, a recent review suggested that legalization likely increases parental cannabis use, as well as maternal cannabis use during pregnancy and postpartum,43 which may contextualize the finding that the primarily cannabis class was more likely to have children. Nonetheless, these findings warrant replication and additional research.
Findings have implications for research, practice, and policy. The use classes and their distinct psychosocial and sociodemographic profiles highlight the need for person-centered research to better understand substance use behaviors. Prevention and treatment efforts should address these complex use profiles and the associated mechanisms of use. For example, classes that used more products and more frequently reported more mental health symptoms and ACEs, showed different personality characteristics, and differed by sociodemographics (e.g., sex, race/ethnicity, education), suggesting different use motives or mechanisms (e.g., to cope with minority stress or trauma vs. for social reasons or sensation-seeking) that may be targeted through comprehensive interventions. Relatedly, further research to understand why certain personality characteristics are linked to distinct use profiles. Finally, given the association between legalized nonmedical cannabis and primarily using cannabis, it is crucial to understand the impact of regulatory context.
Limitations
Despite representation from all 50 states and DC and diversity of the sample, this study is limited in generalizability. Specifically, the purposive sampling of ~50% young adults reporting past-month cannabis use and potential selection bias resulting from recruitment via social media (specifically Facebook) may have impacted findings. Relatedly, this sample was comprised of an imbalance of the sexes (i.e., 41% were male); however, given that substance use rates are typically lower among females vs. males, the sample composition facilitated examination of substance use in females. Additionally, self-reported measures introduce potential bias, dichotomizing substance use variables (rather than using continuous measures) may have limited the nuances in substance use among the classes, and this study was not inclusive of all potential determinants of substance use. Finally, cross-sectional data preclude causal inference and examination of temporal relationships. Thus, future research using longitudinal designs and assessing a larger range of variables with representative samples is needed.
Conclusions
This study identified distinct cannabis and tobacco use classes among US young adults with implications for public health interventions. The frequent poly-product use class showed indicators of high-risk substance use behaviors and related psychosocial risk factors (e.g., mental health symptoms, ACEs), while the class that used the fewest products and the least frequently – the primarily e-cigarette class – showed the lowest levels of these psychosocial factors. Moreover, the sociodemographic profiles of these use classes reflect known disparities in substance use. These findings underscore the importance of future longitudinal research assessing potential transitions between use classes and tailored interventions that address substance use, mental health, and related disparities.
Supplementary Material
IMPLICATIONS.
Relatively few studies have assessed classes of cannabis and tobacco use among young adults, with few assessing key psychosocial determinants. This study examined cannabis and tobacco use classes among US young adults and psychosocial and sociodemographic correlates. Findings documented a frequent poly-product use class, displaying high-risk substance use behaviors and related psychosocial risk factors (e.g., mental health symptoms, ACEs), as well as classes using less often and fewer products with lower psychosocial risks. Moreover, sociodemographic profiles of these classes reflect known substance use disparities. Findings underscore the importance of tailored interventions addressing substance use, mental health, and related disparities.
Funding.
This work was supported by the National Institute on Drug Abuse (R01DA054751, MPIs: Berg, Cavazos-Rehg).
Footnotes
Declaration of Interest. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this paper.
Ethics Approval. This study was approved by the George Washington University Institutional Review Board (NCR224124). All participants provided informed consent.
Data Availability.
The datasets used and/or analyzed in the current study are available from the corresponding author on reasonable request.
REFERENCES
- 1.Chapekis A, Shah S, Most Americans now live in a legal marijuana state – and most have at least one dispensary in their county. Pew Research Center, 2024. https://www.pewresearch.org/short-reads/2024/02/29/most-americans-now-live-in-a-legal-marijuana-state-and-most-have-at-least-one-dispensary-in-their-county/ [Google Scholar]
- 2.Cornelius ME, Loretan CG, Jamal A, et al. Tobacco Product Use Among Adults - United States, 2021. MMWR Morb Mortal Wkly Rep. 2023;72(18):475–483. doi: 10.15585/mmwr.mm7218a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Substance Abuse and Mental Health Services Administration. Key substance use and mental health indicators in the United States: Results from the 2023 National Survey on Drug Use and Health. Rockville, MD: Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration. Retrieved from: https://www.samhsa.gov/data/release/2023-national-survey-drug-use-and-health-nsduh-releases#detailed-tables. [Google Scholar]
- 4.Cho J, Goldenson NI, Kirkpatrick MG, Barrington-Trimis JL, Pang RD, Leventhal AM. Developmental patterns of tobacco product and cannabis use initiation in high school. Addiction. 2021;116(2):382–393. doi: 10.1111/add.15161 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Berg CJ, LoParco CR, Romm KF, et al. Cannabis use characteristics and associations with problematic use outcomes, quitting-related factors, and mental health among US young adults. Subst Abuse Treat Prev Policy. 11 2025;20(1):1. doi: 10.1186/s13011-025-00634-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lanza HI, Bello MS, Cho J, et al. Tobacco and cannabis poly-substance and poly-product use trajectories across adolescence and young adulthood. Prev Med. 2021;148:106545. doi: 10.1016/j.ypmed.2021.106545 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Peters EN, Budney AJ, Carroll KM. Clinical correlates of co-occurring cannabis and tobacco use: a systematic review. Addiction. 2012;107(8):1404–17. doi: 10.1111/j.1360-0443.2012.03843.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Do VV, Ling PM, Chaffee BW, Nguyen N. Concurrent Use of Tobacco and Cannabis and Internalizing and Externalizing Problems in US Youths. JAMA Netw Open. 2024;7(7):e2419976. doi: 10.1001/jamanetworkopen.2024.19976 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Vaughn MG, Salas-Wright CP, Jackson DB. The complex genetic and psychosocial influences on polysubstance misuse. Curr Opin Psychol. 2019;27:62–66. doi: 10.1016/j.copsyc.2018.08.008 [DOI] [PubMed] [Google Scholar]
- 10.Belfiore CI, Galofaro V, Cotroneo D, et al. A Multi-Level Analysis of Biological, Social, and Psychological Determinants of Substance Use Disorder and Co-Occurring Mental Health Outcomes. Psychoactives. 2024;3(2):194–214. [Google Scholar]
- 11.Fluharty M, Taylor AE, Grabski M, Munafo MR. The Association of Cigarette Smoking With Depression and Anxiety: A Systematic Review. Nicotine Tob Res. 2017;19(1):3–13. doi: 10.1093/ntr/ntw140 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gobbi G, Atkin T, Zytynski T, et al. Association of cannabis use in adolescence and risk of depression, anxiety, and suicidality in young adulthood: a systematic review and meta-analysis. JAMA Psychiatry. 2019;76(4):426–434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Control CfD. About Adverse Childhood Experiences (ACEs). Centers for Disease Control. Accessed February 24, 2025, https://www.cdc.gov/aces/about/index.html#:~:text=Adverse%20childhood%20experiences%2C%20or%20ACEs,attempt%20or%20die%20by%20suicide. [Google Scholar]
- 14.Romm KF, Berg CJ. Patterns of Adverse Childhood Experiences and Problematic Health Outcomes Among US Young Adults: A Latent Class Analysis. Subst Use Addctn J. 2024;45(2):191–200. doi: 10.1177/29767342231218081 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Romm KF, Cohn AM, Wang Y, Berg CJ. Psychosocial predictors of trajectories of dual cigarette and e-cigarette use among young adults in the US. Addict Behav. 2023;141:107658. doi: 10.1016/j.addbeh.2023.107658 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chowdhury N, Kevorkian S, Sheerin CM, Zvolensky MJ, Berenz EC. Examination of the Association Among Personality Traits, Anxiety Sensitivity, and Cannabis Use Motives in a Community Sample. J Psychopathol Behav Assess. 2016;38(3):373–380. doi: 10.1007/s10862-015-9526-6 [DOI] [Google Scholar]
- 17.Zvolensky MJ, Taha F, Bono A, Goodwin RD. Big five personality factors and cigarette smoking: a 10-year study among US adults. J Psychiatr Res. 2015;63:91–6. doi: 10.1016/j.jpsychires.2015.02.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Malouff JM, Thorsteinsson EB, Schutte NS. The five-factor model of personality and smoking: a meta-analysis. J Drug Educ. 2006;36(1):47–58. doi: 10.2190/9ep8-17p8-ekg7-66ad [DOI] [PubMed] [Google Scholar]
- 19.Winters AM, Malouff JM, Schutte NS. The association between the five-factor model of personality and problem cannabis use: A meta-analysis. Pers Individ Dif. 2022;193:111635. doi: 10.1016/j.paid.2022.111635 [DOI] [Google Scholar]
- 20.Wang Y, Romm KF, Edberg MC, et al. Two‐part models identifying predictors of cigarette, e-cigarette, and cannabis use and change in use over time among young adults in the US. Am J Addict. 2024;33(5):559–568. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lanza ST. Latent Class Analysis for Developmental Research. Child Dev Perspect. Mar 2016;10(1):59–64. doi: 10.1111/cdep.12163 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Mattingly DT, Elliott MR, Fleischer NL. Latent Classes of Tobacco and Cannabis Use among Youth and Young Adults in the United States. Subst Use Misuse. 2023;58(10):1235–1245. doi: 10.1080/10826084.2023.2215312 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Haardorfer R, Berg CJ, Lewis M, et al. Polytobacco, marijuana, and alcohol use patterns in college students: A latent class analysis. Addict Behav. Aug 2016;59:58–64. doi: 10.1016/j.addbeh.2016.03.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Evans-Polce R, Lanza S, Maggs J. Heterogeneity of alcohol, tobacco, and other substance use behaviors in U.S. college students: A latent class analysis. Addict Behav. 2016;53:80–85. doi: 10.1016/j.addbeh.2015.10.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Krauss MJ, Rajbhandari B, Sowles SJ, Spitznagel EL, Cavazos-Rehg P. A latent class analysis of poly-marijuana use among young adults. Addict Behav. 2017;75:159–165. doi: 10.1016/j.addbeh.2017.07.021 [DOI] [PubMed] [Google Scholar]
- 26.Patrick ME, Berglund PA, Joshi S, Bray BC. A latent class analysis of heavy substance use in Young adulthood and impacts on physical, cognitive, and mental health outcomes in middle age. Drug Alcohol Depend. 2020;212:108018. doi: 10.1016/j.drugalcdep.2020.108018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Connor JP, Leung J, Chan GCK, Stjepanović D. Seeking order in patterns of polysubstance use. Curr Opin Psychiatry. 2023;36(4):263–268. doi: 10.1097/yco.0000000000000881 [DOI] [PubMed] [Google Scholar]
- 28.Berg CJ, Romm KF, LoParco CR, et al. Young Adults' Experiences with Cannabis Retailer Marketing and Related Practices: Differences Among Sociodemographic Groups and Associations with Cannabis Use-related Outcomes. J Racial Ethn Health Disparities. 2024;doi: 10.1007/s40615-024-02092-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Center on the Developing Child - Harvard University. Take the ACE quiz. Retrieved April 1, 2025. from https://developingchild.harvard.edu/media-coverage/take-the-ace-quiz-and-learn-what-it-does-and-doesnt-mean/. 2025; [Google Scholar]
- 30.Gilbert LK, Breiding MJ, Merrick MT, et al. Childhood adversity and adult chronic disease: an update from ten states and the District of Columbia, 2010. Am J Prev Med. 2015;48(3):345–349. [DOI] [PubMed] [Google Scholar]
- 31.Felitti V, Anda R, Nordenberg D, Williamson D. Adverse childhood experiences and health outcomes in adults: The Ace study. J Fam Consum Sci. 1998;90(3):31. [Google Scholar]
- 32.Löwe B, Wahl I, Rose M, et al. A 4-item measure of depression and anxiety: validation and standardization of the Patient Health Questionnaire-4 (PHQ-4) in the general population. J Affect Disord. 2010;122(1-2):86–95. doi: 10.1016/j.jad.2009.06.019 [DOI] [PubMed] [Google Scholar]
- 33.Gosling SD, Rentfrow PJ, Swann WB. A very brief measure of the Big-Five personality domains. J Res Pers. 2003;37(6):504–528. doi: 10.1016/S0092-6566(03)00046-1 [DOI] [Google Scholar]
- 34.Collins LM, Lanza ST. Latent Class and Latent Transition Analysis: With applications in the social, behavioral, and health sciences. John Wiley & Sons, Inc.; 2010. [Google Scholar]
- 35.Nylund KL, Asparouhov T, Muthén BO. Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Struct Equ Modeling. 2007;14(4):535–569. [Google Scholar]
- 36.Smart R, and Pacula RL. Early evidence of the impact of cannabis legalization on cannabis use, cannabis use disorder, and the use of other substances: Findings from state policy evaluations. Am J Drug Alcohol Abuse. 2019;45(6):644–663. doi: 10.1080/00952990.2019.1669626 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Pravosud V, Glantz S, Keyhani S, et al. Cannabis legalization and changes in cannabis and tobacco/nicotine use and co-use in a national cohort of U.S. adults during 2017–2021. Int J Drug Policy. 2024;134:104618. doi: 10.1016/j.drugpo.2024.104618 [DOI] [PubMed] [Google Scholar]
- 38.Veligati S, Howdeshell S, Beeler-Stinn S, et al. Changes in alcohol and cigarette consumption in response to medical and recreational cannabis legalization: Evidence from U.S. state tax receipt data. In J Drug Policy. 2020;75:102585. doi: 10.1016/j.drugpo.2019.10.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Goodwin SR, Moskal D, Marks RM, Clark AE, Squeglia LM, Roche DJO. A Scoping Review of Gender, Sex and Sexuality Differences in Polysubstance Use in Adolescents and Adults. Alcohol Alcoholism. 2022;57(3):292–321. doi: 10.1093/alcalc/agac006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Roghani A, Nyarko SH, Potter L. Smoking Cigarettes, Marijuana, and the Transition to Marriage among Cohabiters in the USA. Global Social Welfare. 2021;8(3):279–286. [Google Scholar]
- 41.Merrin GJ, Bailey JA, Kelly AB, et al. Continuity and change in substance use patterns during the transition from adolescence to young adulthood: examining changes in social roles. Int J Ment Health Addict. 2024:1–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Windle M, Windle RC. A prospective study of mutual influence for substance use among young adult marital dyads. Psychol Addict Behav. 2018;32(2):237–243. doi: 10.1037/adb0000329 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Wilson S, Rhee SH. Causal effects of cannabis legalization on parents, parenting, and children: A systematic review. Prev Med. 2022;156:106956. doi: 10.1016/j.ypmed.2022.106956 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and/or analyzed in the current study are available from the corresponding author on reasonable request.