Abstract
Background
Emerging research suggests morning cannabis use may be associated with using more cannabis and experiencing more cannabis-related problems. This paper examined whether months when young adults reported morning cannabis use (use between 6:00AM and 12:00PM) were associated with cannabis use frequency, negative cannabis-related consequences, and changes in cannabis use disorder (CUD) symptoms.
Methods
Participants were 778 young adults (Mage=21.11 years, 58.5% female) enrolled in a longitudinal study on substance use and social role transitions. Eligible participants were 18–23 years old at screening and reported past-year alcohol use. Participants completed a baseline survey, 24 consecutive monthly surveys, and a follow-up survey 30 months after baseline. Aims were tested using multilevel models and multiple regression.
Results
Analyses were limited to cannabis use months (N=4,719; 28.9% of sampled months) and participants who reported cannabis use at least once (N=542; 69.7% of all participants). Morning use was reported in 12.3% of cannabis use months and at least once by 23.6% of participants who reported using cannabis. Relative to non-morning use months, morning use months were associated with greater cannabis use frequency and more negative consequences. However, the association between morning use and negative consequences was not statistically significant after controlling for cannabis use frequency. The percentage of cannabis use months with morning use was positively associated with increased CUD symptoms at the 30-month follow-up, relative to baseline.
Conclusions
Morning cannabis use may be a useful marker of high-risk cannabis use and may contribute to the maintenance and worsening of CUD over time.
Keywords: Cannabis, Marijuana, Morning cannabis use, Cannabis use disorder, Young adults
1. Introduction
Understanding patterns of cannabis use among young adults in the United States (US) is important given ongoing increases in historical patterns of use (Patrick et al., 2023), cannabis potency (Chandra et al., 2019; Freeman et al., 2021), and the legality of recreational cannabis. At the same time, perceptions of harm related to cannabis use are decreasing, particularly for young adults (Waddell, 2022). In 2022, cannabis use among US young adults reached historic highs, with 42.6% reporting past-year use, 28.5% reporting past-month use, and 10.8% reporting daily or near daily use (Patrick et al., 2023). Although research on young adult cannabis use is quickly growing, there is much less research on cannabis use than on other commonly used substances, like alcohol and tobacco. Therefore, a deeper examination of cannabis use patterns is warranted, particularly those related to higher levels of use and use-related negative consequences.
Research on alcohol and tobacco indicates that using these substances in the morning or shortly after waking may signify a problematic pattern of use. For example, time from waking in the morning until one’s first cigarette of the day is an indicator of nicotine dependence, and time of first use is strongly associated with high-risk nicotine use behaviors (Branstetter et al., 2020; Guertin et al., 2015; Selya et al., 2016). Likewise, morning alcohol use items have been used in diagnostic and screening tests (Babor et al., 2001; Cherpitel, 1999), because morning alcohol use is a primary indicator of alcohol dependence (York, 1995). Given the scarce literature on correlates of morning cannabis use, it is unknown whether morning use may have utility in identifying problematic patterns cannabis use and/or serve as a risk factor for the development of cannabis use disorder (CUD).
To better understand potential risks of morning cannabis use, it is important to determine the degree to which this behavior is associated with levels of overall cannabis use, negative cannabis-related consequences, and symptoms of CUD over time. In a cross-sectional study of adults who used cannabis daily, Earleywine and colleagues (2016) found that adults who used cannabis in the morning (between 6:00AM and 12:00PM) reported experiencing greater cannabis-related problems than those who did not use cannabis in the morning when adjusting for cannabis use quantity on an average occasion. In a cross-sectional study of college students who reported past-month cannabis use, Hetelekides and colleagues (2023) found that students in latent classes characterized by morning cannabis use tended to report using cannabis more frequently and in greater quantities and experiencing more negative cannabis-related consequences than students in classes not characterized by morning use. In an ecological momentary assessment (EMA) study of adolescents and young adults who used cannabis frequently, Shrier and colleagues (2013) found that participants were more likely to take 6+ hits of cannabis when use occurred during the morning (between 6:00AM and 12:00PM) or evening (between 6:00PM and 12:00AM) than when use occurred in the afternoon (between 12:00PM and 6:00PM) or after midnight (between 12:00AM and 6:00AM). Together, these cross-sectional and daily findings suggest that morning cannabis use may be associated with using more cannabis and experiencing more cannabis-related problems.
However, these studies only examined cannabis use behaviors over short periods of time and did not examine within-person associations between morning use and cannabis-related outcomes. Examining both within-person and between-person associations longitudinally may provide additional insight into the amount of risk associated with morning cannabis use, and whether morning use contributes to the etiology and maintenance of CUD. Further, these studies did not examine within-person associations between morning use and cannabis-related outcomes. Examining both within- and between-person associations may help inform whether morning use events (in this case, months with morning use events) confer greater risk of harm relative to when an individual uses at other times of day, independent of typical patterns of cannabis use (i.e., within-person associations), or whether morning use tends to be a characteristic of young adults who engage in higher-risk cannabis use on average (i.e., between-person associations). Further, previous studies on morning use examined use behaviors and related outcomes over short periods of time. Examining longitudinal associations between patterns of morning use across 24 months and CUD at a later timepoint may help inform prevention/intervention efforts by demonstrating whether morning use is an indicator of and/or potentially contributes to the development of CUD.
In a community sample of young adults who reported past-year alcohol use and were followed for 24 consecutive months, the current study examined within-person and between-person associations between morning cannabis use (between 6:00AM and 12:00PM) and cannabis use frequency, negative cannabis-related consequences, and changes in CUD symptoms. The study had two aims. Aim 1 was to test whether participants reported greater cannabis use frequency and experienced more negative cannabis-related consequences in months they reported morning cannabis use compared to months they used cannabis but did not report morning use. Aim 2 was to test whether the percentage of cannabis use months that participants reported morning use was associated with increases in CUD symptoms from baseline to a follow-up survey 2.5 years later (six months after the end of the repeated monthly surveys).
2. Material and Methods
2.1. Participants and Procedure
Participants were 778 young adults enrolled in a longitudinal study on substance use and social role transitions in the greater Seattle, WA metropolitan area (Lee et al., 2018; Patrick et al., 2020). Participants were recruited between January 2015 and January 2016 via social media, newspaper, and community advertisements. Eligibility criteria included being 18–23 years old at screening, reporting past-year alcohol use, and living within 60 miles of the study office in Seattle. After screening in and enrolling in the study, participants completed a baseline survey. Starting the following month, participants completed 24 consecutive months of online surveys. A follow-up survey was completed six months after the end of the repeated monthly surveys (i.e., 30 months after baseline). More detailed information about study procedures is available elsewhere (Lee et al., 2018; Patrick et al., 2020).
The analytic sample was limited to months in which participants reported cannabis use (NMonths=4,719; 28.9% of all sampled months) and thereby to participants who reported cannabis use in at least one month (NPersons=542; 69.7% of all participants). The mean age of the analytic sample at baseline was 21.11 years (SD=1.71), and 41.5% reported their biological sex as male. Regarding race/ethnicity, the composition of the analytic sample was 58.5% White Non-Hispanic (NH), 18.8% Other NH (e.g., Black/African American, multi-racial), 13.8% Asian NH, and 8.9% Hispanic. During the repeated monthly surveys, 52.4% reported being a four-year college student in at least one month, and 58.9% reported working full-time in at least one month. Participants in the analytic sample completed an average of 21.22 (SD=4.91) out of 24 monthly surveys, with 83.6% completing at least 18 monthly surveys and 57.2% completing all 24 monthly surveys. Retention at the follow-up survey was 91.5%.
2.2. Repeated Monthly Measures
Cannabis use frequency.
Each month, participants were asked, “In the past 30 days, how many days did you use marijuana?” Participants responded by typing integers into a text box.
Morning cannabis use.
In months participants reported cannabis use, they were asked, “When did you typically use marijuana in the past 30 days? Check all that apply.” Response options were “Morning (6am–noon),” “Afternoon (noon–6pm),” “Evening (6pm–midnight),” and “Night (midnight–6am).” Months in which participants endorsed morning (between 6:00AM and 12:00PM) as a time they typically used cannabis were coded as 1, regardless of whether other times of day were also endorsed. Months in which participants did not endorse morning as a time they typically used were coded as 0. For use in descriptive statistics, four other variables were created from these responses. Months with typical afternoon cannabis use were operationalized as months when participants endorsed typically using cannabis between 12:00PM and 6:00PM, regardless of whether other times of day were also endorsed. Months with typical evening cannabis use were operationalized as months when participants endorsed typically using cannabis between 6:00PM and 12:00AM. Months with typical cannabis use after midnight were operationalized as months when participants endorsed typically using cannabis between 12:00AM and 6:00AM. Months with typical cannabis use outside of morning hours were operationalized as months when participants endorsed using cannabis between 12:00PM and 6:00PM, 6:00PM and 12:00AM, or 12:00AM and 6:00AM, regardless of whether morning use was also endorsed.
Negative cannabis-related consequences.
In months participants reported cannabis use, they completed the Marijuana Consequences Checklist (Lee et al., 2021), a 26-item measure of past-month negative cannabis-related consequences. Items included “had a panic or anxiety attack,” “had your driving affected after using marijuana,” “had problems following through on things,” and “felt paranoid.” Response options were “0 times,” “1–2 times,” “3–5 times,” “6–10 times,” and “More than 10 times.” Responses were dichotomized to indicate whether each consequence had (1) or had not (0) been experienced in the past month. The dichotomized responses were then summed to create a variable representing the total number of negative cannabis-related consequences participants experienced in the past month.
Four-year college enrollment.
Each month, participants were asked, “Tell us about your educational situation in the past month. Check all that apply.” Response options were “High school student,” “General Education Development (GED) student,” “Trade or vocational school student,” “2-year or community college student,” “4-year college or university student,” “Graduate or professional school student,” “Other certifications or coursework,” and “Not a student”. Months participants reported being a 4-year student were coded as 1, and months participants did not report being a 4-year student were coded as 0.
Full-time employment.
Each month, participants were asked, “Tell us about your work situation (paid and unpaid) in the past month. Check all that apply.” There were 14 response options including “working part-time (paid),” “working full-time (paid),” and “internship/apprenticeship (paid or unpaid).” Months participants reported “working full-time (paid)” were coded as 1, and months participants reported not working full-time were coded as 0.
2.3. Baseline and Follow-Up Measures
CUD symptoms.
At baseline and follow-up, participants completed the Cannabis Use Disorders Identification Test – Revised (CUDIT-R; Adamson et al., 2010), a 9-item measure designed to screen for CUD symptoms in the past six months and potential CUD. The nine items had varying response options. The CUDIT-R was scored by summing the last eight items. The internal consistency of the CUDIT-R was α=0.83 at both timepoints. CUDIT-R scores ≥ 8 represent hazardous levels of cannabis misuse; however, this cutoff may be lower among young adults (Schultz et al., 2019).
2.4. Analyses
Aim 1, which tested associations between morning cannabis use and cannabis use frequency and negative cannabis-related consequences in a given month, was tested using Poisson multilevel models estimated using maximum likelihood estimation based on the Laplace Approximation in the glmmTMB package (Brooks et al., 2017) of R 4.2.2 (R Core Team, 2022). A truncated Poisson model was used to test associations between morning use and cannabis use frequency, since all cannabis use months had at least one cannabis use day. Overdispersion was accounted for using a monthly-level random effect, as necessary. Random slopes were specified for monthly-level associations when doing so improved model fit based on likelihood ratio tests. Given that some aspects of cannabis use (i.e., vaping, daily use) tend to be more common among young adults not in college full-time than among those enrolled in college full-time (Patrick et al., 2023) and that full-time employment has been linked with lower cannabis use frequency (Bears Augustyn et al., 2020), these models controlled for four-year college enrollment and full-time employment. To disaggregate the within-person and between-person associations at the monthly (Level 1) and person levels (Level 2), respectively, monthly-level cannabis-related variables were person-mean centered (Hamaker & Muthén, 2020). The four-year college enrollment and full-time employment variables varied across months but did not vary enough for within-person and between-person associations to be examined separately. Rather, these variables were left uncentered, and the coefficients for these variables in model results represented a composite of within- and between-person associations (Hamaker & Muthén, 2020; Hox et al., 2017). A time in years variable (i.e., study month divided by 12) was included at the monthly level and was centered at its midpoint. All person-level variables were grand-mean centered. All models controlled for sex, age at baseline, and race/ethnicity at the person level.
Aim 2, which tested whether the percentage of cannabis use months in which morning use was reported predicted change in CUD symptoms from baseline to the follow-up 2.5 years later, was tested using a negative binomial regression in the MASS package (Venables & Ripley, 2002) in R. CUD symptoms at the follow-up were regressed on the percentage of cannabis use months in which morning use was reported (divided by 10 so that a one-unit change represented a 10% difference), number of months cannabis use was reported, baseline CUD symptoms, sex, age at baseline, and race/ethnicity. Multiple imputation was performed using the MICE package (van Buuren & Groothuis-Oudshoorn, 2011) in R, according to recommendations in van Buuren (2018), to account for missing data in CUD symptoms at the follow-up. Multiple imputation is expected to produce unbiased estimates under the assumption that data are missing at random. Five-hundred imputed datasets were created, and all variables included in the regression were entered as covariates in the imputation model. Summary parameters and standard errors were estimated according to Rubin’s Rules, which account for uncertainty within and between imputed datasets (Rubin, 2014). All variables in this model were grand-mean centered.
3. Results
3.1. Descriptive Statistics
Descriptive statistics are presented in Table 1. Cannabis use was reported in a median of 7 sampled months (M=8.71, SD=7.04, range: 1–24), which corresponded to a median of 35.00% of sampled months (M=41.24%, SD=30.39%, range: 4.17–100.00%). Morning cannabis use was reported in 12.25% of all cannabis use months and at least once by 23.62% of individuals who reported cannabis use. Among participants who reported morning cannabis use in at least one month, the median percentage of cannabis use months with morning use was 25.54% (M=39.55%, SD=33.40%, range: 4.17–100.00%). Ignoring the nesting of person-months within persons, the median number of past-month cannabis use days was 7.00 (M=12.40, SD=11.18, range: 1–30). Negative cannabis-related consequences were reported in 90.64% of cannabis use months, and the median number of past-month negative consequences was 5.00 (M=6.47, SD=5.21, range: 0–26). The median number of CUD symptoms was 5.00 at baseline (M=6.29, SD=6.24, range: 0–30) and 3.00 at the follow-up (M=5.00, SD=5.61, range: 0–32).
Table 1.
Descriptive Statistics
| Monthly-Level Variables (N = 4,719 person-months) | |||||
|---|---|---|---|---|---|
|
| |||||
| Variable | n | M (SD) or % | Min. | Max. | ICC |
|
| |||||
| No. of cannabis use days | 4,719 | 12.40 (11.18) | 1 | 30 | 0.79 |
| Morning cannabis use | 4,663 | 12.3% | 0 | 1 | 0.91 |
| No. of negative consequences | 4,719 | 6.39 (5.23) | 0 | 26 | 0.66 |
| Four-year college enrollment | 4,719 | 36.5% | 0 | 1 | 0.89 |
| Full-time employment | 4,719 | 32.8% | 0 | 1 | 0.64 |
|
| |||||
| Person-Level Variables (N = 542 individuals) | |||||
|
| |||||
| Variable | n | M (SD) or % | Min. | Max. | |
|
| |||||
| No. of cannabis use months | 542 | 8.71 (7.04) | 1 | 24 | |
| % of months with morning use | 542 | 9.36 (23.35) | 0.00 | 100.00 | |
| CUD symptoms at baseline | 542 | 6.28 (6.23) | 0 | 30 | |
| CUD symptoms at follow-up | 542 | 5.03 (5.57) | 0 | 32 | |
| Male sex | 542 | 41.5% | 0 | 1 | |
| Age at baseline | 542 | 21.11 (1.71) | 18.05 | 24.05 | |
| Race/ethnicity | |||||
| Asian NH | 542 | 13.8% | 0 | 1 | |
| Hispanic | 542 | 8.9% | 0 | 1 | |
| Other NH | 542 | 18.8% | 0 | 1 | |
| White NH | 542 | 58.5% | 0 | 1 | |
Note. ICC = Intraclass correlation coefficient; CUD = Cannabis use disorder; NH = Non-Hispanic.
Although the purpose of this paper was to compare months with and without morning use, it is important to note that most months in which morning use was reported also included typical cannabis use during other periods of the day. In 97.72% of morning use months, cannabis was also typically used outside of the morning hours (i.e., 12:00 PM to 6:00 AM). More specifically, typical afternoon cannabis use (12:00–6:00 PM) was reported in 76.36% of morning use months, typical evening use (6:00 PM to midnight) was reported in 90.89% of morning use months, and typical use after midnight (midnight to 6:00 AM) was reported in 50.79% of morning use months.
3.2. Aim 1: Associations between Morning Cannabis Use, Cannabis Use Frequency, and Negative Cannabis-Related Consequences
Table 2 presents the results of a truncated Poisson multilevel model that tested the association between morning cannabis use and cannabis use frequency (i.e., number of past-month cannabis use days). At the within-person monthly level, morning use in a given month was positively associated with number of cannabis use days such that participants reported 21% more cannabis use days, on average, in months with morning use compared to months without morning use. The model-predicted number of cannabis use days was 3.12 (95% CI: 2.82–3.45) in months without morning use and 3.80 (95% CI: 3.38–4.26) in months with morning use. At the person level, the percentage of months with morning use was positively associated with participants’ average number of past-month cannabis use days such that each 10% increase in the percentage of months with morning use was associated with a 25% higher average number of past-month cannabis use days. Number of cannabis use months during the study period was positively associated with participants’ average number of past-month cannabis use days such that each additional month of cannabis use was associated with an 11% higher average number of past-month cannabis use days.
Table 2.
Truncated Poisson Multilevel Model Testing Associations between Morning Cannabis Use and Number of Past-Month Cannabis Use Days in Months with Cannabis Use
| Fixed Effects | Outcome: Number of Cannabis Use Days |
||
|---|---|---|---|
| Rate Ratio | 95% CI | ||
|
| |||
| Level 2: Person Level | |||
| Intercept | 3.10*** | 2.79, 3.45 | |
| % of cannabis use months with morning use (10% increments) | 1.25*** | 1.20, 1.30 | |
| Number of cannabis use months | 1.11*** | 1.09, 1.12 | |
| Male sex | 1.39*** | 1.15, 1.69 | |
| Age at baseline | 0.96 | 0.91, 1.02 | |
| Race/ethnicity | |||
| Asian NH | 0.56*** | 0.42, 0.76 | |
| Hispanic | 0.80 | 0.56, 1.14 | |
| Other NH | 0.98 | 0.76, 1.26 | |
| Level 1: Month Level | |||
| Morning cannabis use | 1.21*** | 1.13, 1.29 | |
| Four-year college student | 1.05 | 0.97, 1.13 | |
| Employed full-time | 1.01 | 0.97, 1.06 | |
| Time in years | 1.08 | 1.00, 1.18 | |
|
| |||
| Random Effects | SD | ||
|
| |||
| Intercept | 0.99 | ||
| Morning cannabis use | 0.12 | ||
| Time in years | 0.65 | ||
| Overdispersion | 0.30 | ||
Note. NMonths = 4,663, NPersons = 541. NH = Non-Hispanic.
p < .05;
p < .01;
p < .001.
Table 3 presents the results of two Poisson multilevel models testing associations between morning cannabis use and number of past-month negative cannabis-related consequences. The first model tested this association without controlling for the number of past-month cannabis use days. At the within-person monthly level, morning cannabis use in a given month was positively associated with the number of negative cannabis-related consequences such that participants reported 8% more negative consequences, on average, in months with morning use compared to months without morning use. The model-predicted number of negative consequences was 3.65 (95% CI: 3.40–3.93) in months without morning use and 3.96 (95% CI: 3.62–4.33) in months with morning use. At the person level, the percentage of months with morning use was positively associated with participants’ average number of past-month negative cannabis-related consequences such that each 10% increase in the percentage of months with morning use was associated with a 6% greater average number of past-month negative consequences. Number of cannabis use months was positively associated with participants’ average number of past-month negative cannabis-related consequences such that each additional month of cannabis use during the study period was associated with a 4% greater average number of past-month negative consequences.
Table 3.
Poisson Multilevel Models Testing Associations between Morning Cannabis Use and Number of Past-Month Negative Cannabis-Related Consequences in Months with Cannabis Use
| Fixed Effects | Outcome: No. of Negative Cannabis-Related Consequences | |||
|---|---|---|---|---|
|
| ||||
| Not controlling for no. of cannabis use days | Controlling for no. of cannabis use days | |||
|
| ||||
| Rate Ratio | 95% CI | Rate Ratio | 95% CI | |
|
| ||||
| Level 2: Person Level | ||||
| Intercept | 3.66*** | 3.39, 3.96 | 3.70*** | 3.44, 3.99 |
| % of cannabis use months with morning use (10% increments) | 1.06*** | 1.03, 1.09 | 0.98 | 0.95, 1.01 |
| No. of cannabis use months | 1.04*** | 1.03, 1.05 | 1.02*** | 1.01, 1.03 |
| Mean no. of cannabis use days | - | - | 1.04*** | 1.03, 1.05 |
| Male sex | 1.06 | 0.92, 1.22 | 0.97 | 0.85, 1.10 |
| Age at baseline | 0.94** | 0.91, 0.98 | 0.95** | 0.91, 0.98 |
| Race/ethnicity | ||||
| Asian NH | 0.96 | 0.78, 1.18 | 1.10 | 0.90, 1.33 |
| Hispanic | 1.03 | 0.81, 1.31 | 1.10 | 0.87, 1.37 |
| Other NH | 1.00 | 0.84, 1.20 | 1.06 | 0.89, 1.25 |
| Level 1: Month Level | ||||
| Morning cannabis use | 1.08** | 1.02, 1.15 | 1.01 | 0.96, 1.07 |
| No. of cannabis use days | - | - | 1.04*** | 1.03, 1.04 |
| Four-year college student | 1.07 | 1.00, 1.14 | 1.07* | 1.01, 1.14 |
| Employed full-time | 0.95* | 0.90, 0.99 | 0.95* | 0.90, 0.99 |
| Time in years | 0.89*** | 0.84, 0.95 | 0.86*** | 0.82, 0.91 |
|
| ||||
| Random Effects | SD | SD | ||
|
| ||||
| Intercept | 0.74 | 0.70 | ||
| No. of cannabis use days | - | 0.02 | ||
| Time in years | 0.45 | 0.39 | ||
| Overdispersion | 0.22 | 0.14 | ||
Note. NMonths = 4,632, NPersons = 540. NH = Non-Hispanic.
p < .05;
p < .01;
p < .001.
The second model in Table 3 tested the association between morning cannabis use and number of past-month negative cannabis-related consequences when controlling for number of past-month cannabis use days. At the within-person monthly level, morning cannabis use in a given month was not significantly associated with the number of negative cannabis-related consequences. The number of cannabis use days was positively associated with the number of negative cannabis-related consequences such that each additional day of cannabis use was associated with experiencing 4% more negative consequences. At the person level, the percentage of months with morning use was not significantly associated with participants’ average number of past-month negative cannabis-related consequences. Number of cannabis use months during the study period was positively associated with the average number of past-month negative cannabis-related consequences such that each additional cannabis use month during the study period was associated with a 2% greater average number of past-month negative consequences. Average number of past-month cannabis use days was also positively associated with the average number of past-month negative cannabis-related consequences such that each one day increase in participants’ average number of past-month cannabis use days was associated with a 4% greater average number of past-month negative consequences.
3.3. Aim 2: Morning Cannabis Use as a Predictor of Change in CUD Symptoms
Table 4 presents the results of a negative binomial regression testing whether the percentage of cannabis use months with morning use predicted change in CUD symptoms from baseline to the follow-up 2.5 years later. The percentage of months with morning use was positively associated with change in CUD symptoms such that each 10% increase in the percentage of months with morning use was associated with a 7% increase in CUD symptoms from baseline to the follow-up 2.5 years later, independent of number of cannabis use months during the study period, baseline CUD symptoms, sex, age at baseline, and race/ethnicity. Number of cannabis use months was also positively associated with change in CUD symptoms such that each additional month of cannabis use during the study period was associated with a 6% increase in CUD symptoms from baseline to follow-up. Baseline CUD symptoms were positively associated with CUD symptoms at the follow-up.
Table 4.
Negative Binomial Regression Testing Whether Morning Cannabis Use Predicted Change in CUD Symptoms across 2.5 Years
| Predictor Variables | Outcome: CUD Symptoms at Follow-Up |
|
|---|---|---|
| Rate Ratio | 95% CI | |
|
| ||
| Intercept | 4.03*** | 3.68, 4.41 |
| % of cannabis use months with morning use (10% increments) | 1.07** | 1.02, 1.11 |
| Number of cannabis use months | 1.06*** | 1.05, 1.08 |
| CUD symptoms at baseline | 1.04*** | 1.02, 1.05 |
| Male sex | 1.06 | 0.88, 1.28 |
| Age at baseline | 0.90*** | 0.86, 0.95 |
| Race/ethnicity | ||
| Asian NH | 0.76* | 0.58, 1.00 |
| Hispanic | 1.00 | 0.72, 1.37 |
| Other NH | 1.13 | 0.89, 1.43 |
Note. N = 542 participants.
p < .05;
p < .01;
p < .001.
4. Discussion
Results from this intensive longitudinal study of a community sample of young adults suggest that morning was not an uncommon time for cannabis use among young adults who use cannabis. Morning was reported as a typical time of use in at least one month by more than one-fifth of young adults who reported using cannabis during the study period and in approximately one-eighth of all cannabis use months. Months young adults reported morning cannabis use (compared to use months without morning use) were associated with more frequent overall cannabis use and experiencing more negative consequences. Moreover, participants with a greater percentage of cannabis use months with morning use reported higher average levels of cannabis use frequency and negative consequences. However, after controlling for cannabis use frequency in the consequences models, there were no significant associations between morning use and number of past-month negative consequences at the within-person monthly level or the person level. Finally, the percentage of cannabis use months with morning use was positively associated with increases in CUD symptoms from baseline to the follow-up 2.5 years later.
In contrast to rates of morning drinking, which are very low even among individuals who drink heavily (e.g., 2% among college students; Piasecki et al., 2005), rates of morning cannabis use were relatively high in this sample, with nearly a quarter of cannabis-using young adults endorsing morning as a time of typical use in at least one month. This higher prevalence may be explained by increasing permissiveness toward cannabis use and decreasing perceptions of risk (Waddell, 2022). Although no known studies have assessed perceived risk and peer disapproval of morning cannabis use specifically, both perceived risk and perceived peer disapproval of regular cannabis use more generally are at historic lows (Schulenberg et al., 2021), which may influence young adults to use cannabis earlier in the day if they believe doing so is not risky and will not be negatively judged by their peers.
Findings from this monthly-level study extend results of cross-sectional (Earleywine et al., 2016; Hetelekides et al., 2023) and EMA (Shrier et al., 2013) studies that provide evidence linking morning cannabis use with greater levels of overall cannabis use. Whereas the other studies used smaller samples of individuals followed for shorter amounts of time, the present findings do so in a larger sample followed for 2.5 years. The large ICC (0.91) for morning use in this sample, along with the positive person-level associations between morning use and overall cannabis use frequency and quantity, may indicate that morning use tends to primarily be a stable characteristic of use among young adults who use cannabis frequently and in large amounts. However, the positive within-person associations between morning use and overall cannabis use frequency suggests that young adults tend to use cannabis more frequently in months they typically use in the morning compared to months they themselves use cannabis but not typically in the morning.
Findings from the present study suggest that the positive associations between morning cannabis use and number of negative cannabis-related consequences at the monthly and person levels were fully explained by more frequent cannabis use. However, it is possible that morning cannabis use may be associated with greater odds of engaging in specific risk behaviors that could lead to certain negative use-related consequences, such as driving, working, or being at school under the influence of cannabis. Thus, although morning use was not associated with negative consequences after accounting for use frequency, it may provide greater opportunities for engaging in risky behaviors and contribute to continued cannabis use over the course of the day. Furthermore, the present study was not designed to assess negative consequences that are specific to or more likely to occur because of morning cannabis use, which may partially explain our null findings. Specific consequences of morning use that could be explored in forthcoming research include work/school engagement or productivity, physical consequences such as low-energy, and shirking one’s daily responsibilities.
Providing initial evidence of risks of morning cannabis use over somewhat longer time periods, findings showed that young adults who reported morning cannabis use in a greater percentage of use months across two years experienced greater relative increases in CUD symptoms from baseline to the follow-up survey six months after the end of the repeated monthly surveys (and 2.5 years after baseline). Although morning cannabis use was not a unique contributor to experiencing negative consequences, it was longitudinally associated with symptoms of CUD. This finding may be partially explained by the high degree of between-person consistency in morning cannabis use (ICC=0.91) and overall use patterns associated with morning use – people who engaged in morning cannabis use tended to do so consistently over time and continued to engage in heavier, more frequent use as reflected in higher CUD symptoms. Given longitudinal associations with CUD symptoms, morning cannabis use may be an important factor that impedes normative age-related reductions in use (i.e., “maturing out”) (Jochman & Fromme, 2010; Waddell, 2022) and contributes to the development and maintenance of CUD symptoms.
Findings from the present study suggest that morning use is an indicator of heavy, frequent cannabis use and is correlated with the development of CUD symptoms. Indicated intervention and CUD screening may be warranted for young adults who engage in morning use, as the individual is likely to be engaging in heavy, frequent cannabis use and progressing towards CUD. Although our findings do not indicate that morning use is a causal determinant of negative cannabis-related consequences or CUD, that does not mean morning use is harmless. For instance, when cannabis is used before going to school or work, it may interfere with academic or occupational performance. Similarly, young adults are more likely to report driving under the influence of cannabis on days they use cannabis within 30 minutes of waking compared to days they use cannabis but not within 30 minutes of waking (Calhoun et al., 2023). Although it is unclear whether this finding generalizes to morning use more broadly, it is plausible that there are unique acute risks of morning use that were not identified in this study. More research is needed to determine what such risks might be.
This study had several strengths. First, compared to other studies assessing morning cannabis use, the sample was larger, and participants were followed for a longer period of time. Second, the repeated measures design allowed for within-person monthly associations to be estimated (in addition to between-person associations) and longitudinal examination of change in CUD symptoms. This study also had several limitations. First, cannabis use and negative consequences variables were assessed at the monthly level. It is unknown how use behaviors were distributed across use months, and it is unknown whether morning cannabis use occurred on days with greater quantities of cannabis use and greater consequences. Second, morning cannabis use was operationalized based on an item asking about typical monthly use. It is unknown how frequently it occurred in months it was typical or whether it occurred in cannabis use months it was not typical; this framing may also miss people who engaged in morning use but did not consider morning use to be the typical time of day they used. Relatedly, given that there may be considerable variability in young adults’ sleep and wake schedules, there may also be benefits to assessing when use occurred relative to waking. For instance, one recent study found that young adults were high for more hours and had greater odds of driving under the influence of cannabis on days they used cannabis within 30 minutes of waking compared to days cannabis was used but not within 30 minutes of waking (Calhoun et al., 2023). Third, participants were sampled from a large metropolitan area in a state and at a time in which non-medical/recreational cannabis was legal. This may limit the generalizability of the findings, as use behaviors may not be similar in states in which recreational cannabis is not legal or in non-metropolitan areas.
5. Conclusions
Findings from the present study indicate that morning cannabis use may be an important and understudied correlate of more frequent cannabis use as well as a contributor to increases in CUD symptoms over time. Continued research on contextual predictors of cannabis use, including time of day, and predictors of morning use specifically (e.g., perceived risk, normative beliefs, and motives), will be important next steps in advancing our understanding of cannabis misuse among young adults.
Highlights.
Morning cannabis use was reported by 23.6% of young adult participants.
Morning cannabis use was reported in 12.3% of cannabis use months.
Morning use months were associated with greater cannabis use frequency.
Morning use predicted changes in cannabis use disorder symptoms.
Role of Funding Source
Funding for this study was provided by NIAAA grants R01AA022087 (PI: Lee), R01AA027496 (PI: Lee), and F32AA029589 (PI: Walukevich-Dienst). NIAAA did have any role in the study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.
A portion of these findings were presented as a poster at the annual conference of Collaborative Perspective on Addiction in March 2023. The content of this manuscript is solely the responsibility of the author(s) and does not necessarily represent the official views of the NIAAA or the National Institutes of Health.
Footnotes
Conflict of Interest
All authors declare that they have no conflicts of interest.
Declarations of interest: none
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- Adamson SJ, Kay-Lambkin FJ, Baker AL, Lewin TJ, Thornton L, Kelly BJ, Sellman JD, 2010. An improved brief measure of cannabis misuse: The Cannabis Use Disorders Identification Test-Revised (CUDIT-R). Drug Alcohol Depend. 110, 137–143. 10.1016/j.drugalcdep.2010.02.017 [DOI] [PubMed] [Google Scholar]
- Babor TF, Higgins-Biddle JC, Saunders JB, Monteiro MG, 2001. The Alcohol Use Disorders Identification Test. World Health Organization, Geneva. [Google Scholar]
- Bears Augustyn M, Loughran T, Larroulet P, Henry K L., 2020. Economic effects of adolescent to adult patterns of cannabis use: Full-time employment and employment stability. J. Drug Issues. 50, 579–603. 10.1177/0022042620943538 [DOI] [Google Scholar]
- Branstetter SA, Muscat JE, Mercincavage M, 2020. Time to first cigarette: A potential clinical screening tool for nicotine dependence. J. Addict. Med. 14, 409–414. 10.1097/FADM.0000000000000610 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brooks ME, Kristensen K, van Benthem KJ, Magnusson A, Berg CW, Nielsen A, Skaug HJ, Machler M, Bolker BM, 2017. glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R J. 9, 378–400. 10.3929/ethz-b-000240890 [DOI] [Google Scholar]
- Calhoun BH, Graupensperger S, Fairlie AM, Walukevich-Dienst K, Patrick ME, Lee CM, 2023. “Wake-and-bake” cannabis use: Predictors and cannabis-related outcomes of use shortly after waking. Drug Alcohol Depend. 248, 109937. 10.1016/j.drugalcdep.2023.109937 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chandra S, Radwan MM, Majumdar CG, Church JC, Freeman TP, ElSohly MA, 2019. New trends in cannabis potency in USA and Europe during the last decade (2008–2017). Eur Arch Psychiatry Clin Neurosci. 269, 5–15. 10.1007/s00406-019-00983-5 [DOI] [PubMed] [Google Scholar]
- Cherpitel CJ, 1999. Screening for alcohol problems in the U.S. general population: A comparison of the CAGE and TWEAK by gender, ethnicity, and services utilization. J Stud Alcohol Drugs, 60, 705–711. 10.15288/jsa.1999.60.705 [DOI] [PubMed] [Google Scholar]
- Earleywine M, Luba R, Slavin MN, Farmer S, Loflin M, 2016. Don’t wake and bake: Morning use predicts cannabis problems. Addict. Res. Theory. 24, 426–430. 10.1080/16066359.2016.1177027 [DOI] [Google Scholar]
- Freeman TP, Craft S, Wilson J, Stylianou S, El Sohly M, Di Forti M, Lynskey MT, 2021. Changes in delta-9-tetrahydrocannabinol (THC) and cannabidiol (CBD) concentrations in cannabis over time: Systematic review and meta-analysis. Addiction. 116, 1000–1010. 10.1111/add.15253 [DOI] [PubMed] [Google Scholar]
- Guertin KA, Gu F, Wacholder S, Freedman ND, Panagiotou OA, Reyes-Guzman C, Caporaso NE, 2015. Time to first morning cigarette and risk of chronic obstructive pulmonary disease: Smokers in the PLCO Cancer Screening Trial. PLoS One. 10, 1–11. 10.1371/journal.pone.0125973 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hamaker EL, Muthén B, 2020. The fixed versus random effects debate and how it relates to centering in multilevel modeling. Psychol Methods. 25, 365–379. 10.1037/met0000239 [DOI] [PubMed] [Google Scholar]
- Hetelekides E, Joseph V, Pearson M, Bravo AJ, Prince M, Conner B, 2023. Early birds and night owls: Distinguishing profiles of cannabis use habits by use times with latent class analysis. Cannabis. 6, 79–98. 10.26828/cannabis/2023.01.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hox JJ, Moerbeek M, Van de Schoot R, 2017. Multilevel analysis: Techniques and applications. Routledge. [Google Scholar]
- Jochman KA, Fromme K, 2010. Maturing out of substance use: The other side of etiology. In Scheier L (Ed.), Handbook of drug use etiology: Theory, methods, and empirical findings (pp. 565–578). American Psychological Association. [Google Scholar]
- Lee CM, Cadigan JM, Fairlie AM, Lewis MA, 2018. Transitions into young adulthood: Extent to which alcohol use, perceived drinking norms, and consequences vary by education and work statuses among 18–20 year olds. Addict Behav, 79, 107–112. 10.1016/j.addbeh.2017.12.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee CM, Kilmer JR, Neighbors C, Cadigan JM, Fairlie AM, Patrick ME, Logan DE, Walter T, White HR, 2021. A marijuana consequences checklist for young adults with implications for brief motivational intervention research. Prev Sci, 22, 758–768. 10.1007/Fs11121-020-01171-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Patrick ME, Rhew IC, Duckworth JC, Lewis MA, Abdallah DA, Lee CM, 2020. Patterns of young adult social roles transitions across 24 months and subsequent substance use and mental health. J Youth Adolesc, 49, 869–880. 10.1007/Fs10964-019-01134-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Patrick ME, Miech RA, Johnston LD, O’Malley PM, 2023. Monitoring the Future Panel Study annual report: National data on substance use among adults ages 19 to 60, 1976–2022. Monitoring the Future Monograph Series. University of Michigan Institute for Social Research: Ann Arbor, MI. 10.7826/ISR-UM.06.585140.002.07.0002.2023 [DOI] [Google Scholar]
- Piasecki TM, Sher KJ, Slutske WS, Jackson KM, 2005. Hangover frequency and risk for alcohol use disorders: Evidence from a longitudinal high-risk study. J. Abnorm. Psychol. 114, 223–234. 10.1037/0021-843X.114.2.223 [DOI] [PubMed] [Google Scholar]
- R Core Team, 2022. R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/ [Google Scholar]
- Rubin BD, 2014. Multiple imputation for nonresponse in surveys. John Wiley and Sons Ltd. [Google Scholar]
- Schulenberg JE, Patrick ME, Johnston LD, O’Malley PM, Bachman JG, Miech RA, 2021. Monitoring the Future national survey results on drug use, 1975–2020: Volume II, College students and adults ages 19–60. University of Michigan Institute for Social Research: Ann Arbor, MI. [Google Scholar]
- Schultz NR, Bassett DT, Messina BG, & Correia CJ, 2019. Evaluation of the psychometric properties of the Cannabis Use Disorders Identification Test - Revised among college students. Addict. Behav. 95, 11–15. 10.1016/j.addbeh.2019.02.016 [DOI] [PubMed] [Google Scholar]
- Selya AS, Oancea SC, Thapa S, 2016. Time to first cigarette, a proxy of nicotine dependence, increases the risk of pulmonary impairment, independently of current and lifetime smoking behavior. Nicotine Tob. Res. 18, 1431–1439. 10.1093/ntr/ntv291 [DOI] [PubMed] [Google Scholar]
- Shrier LA, Walls C, Rhoads A, Blood EA, 2013. Individual and contextual predictors of severity of marijuana use events among young frequent users. Addict Behav. 38, 1448–1456. 10.1016/j.addbeh.2012.05.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
- van Buuren S, 2018. Flexible imputation of missing data. CRC Press. [Google Scholar]
- van Buuren S, Groothuis-Oudshoorn K, 2011. mice: Multivariate imputation by chained equations in R. J Stat Softw, 45, 1–67. [Google Scholar]
- Venables WN, Ripley BD, 2002. Modern applied statistics with S (4th ed.). Springer. [Google Scholar]
- Waddell JT, 2022. Age-varying time trends in cannabis- and alcohol-related risk perceptions 2002–2019. Addict. Behav. 124, 107091. 10.1016/j.addbeh.2021.107091 [DOI] [PMC free article] [PubMed] [Google Scholar]
- York JL, 1995. Progression of alcohol consumption across the drinking career in alcoholics and social drinkers. J. Stud. Alcohol. 56, 328–336. 10.15288/jsa.1995.56.328 [DOI] [PubMed] [Google Scholar]