Daily web survey data collection of time-varying cannabis use motives and contexts, with implications for adaptive interventions: A pilot study
Survey Research Center, Institute for Social Research, University of Michigan, Ann Arbor, United States
Center for the Study of Drugs, Alcohol, Smoking and Health, School of Nursing, University of Michigan, Ann Arbor, United States
Abstract
Background:
With the rising prevalence of daily cannabis use in the U.S., more individuals may seek treatment for adverse outcomes (e.g., cannabis use disorder) arising from frequent cannabis use. Adaptive interventions leverage self-reported motives for cannabis use to develop personalized support. Research has demonstrated that these motives can be collected on a yearly, monthly, or daily basis, and strongly predict both the frequency of cannabis use and associated adverse outcomes.
Methods:
We employed a daily web survey over a 28-day period to assess daily motives and open-ended self-reported contexts of nonmedical cannabis use. Participants, aged 22–76 (n = 48, Mean = 48.8, SD = 17.1; 64 % female; 22 % African American), completed a baseline web survey on demographics and mental health and a follow-up web survey on mental health at present and cannabis use behaviors during the study. We applied latent transition analysis with random intercepts (RI-LTA) to analyze the data.
Results:
In our descriptive analyses, we identified four types of transitions in weekly latent motive classes, particularly transitions toward motives dominated by sleep aid and cannabis availability. Participants experiencing these transitions reported higher subsequent cannabis use frequency. Open-ended contextual information revealed behaviors such as cannabis use related to managing sleep disturbances, anxiety triggered by daily stressors, and recovery from medical treatments (e.g., chemotherapy).
Conclusions:
Our findings underscore the value of monitoring short-term transitions in cannabis use motives and contexts. This approach can inform the timing and content of adaptive interventions, proactively addressing the contexts and motives to prevent adverse cannabis use outcomes. Additional replications of this approach using larger samples are needed.
Untitled section
Keywords: Motives for cannabis use, Contexts for cannabis use, Adaptive interventions, Latent transition analysis, Daily web survey
Article notes
Untitled section
Issue date 2026 Jan 1.
1.Introduction
1.2.Adaptive interventions as treatment strategies
Adaptive interventions (ADIs) are flexible treatment strategies that evolve over time to address the changing needs and behaviors of an individual and can lead to effective prevention and treatment, reducing cannabis use and associated consequences over time (Nahum-Shani et al., 2022; Nahum-Shani et al., 2012; Nahum-Shani et al., 2018; Perski et al., 2022). Recent studies tailored real-time messages supporting self-efficacy and prompting consideration of coping strategies for adolescents and young adult patients to reduce cannabis use according to self-reported momentary triggers for use, desire to use, or recent use; results showed that momentary cannabis use decreased over the course of the study, and that the odds of use were significantly lower in the intervention and follow-up phases than in the baseline phase (Shrier et al., 2014, 2018).
Lee et al. (2010) evaluated a brief web-based personalized feedback intervention for incoming college students who had recent cannabis use and suggested that the context of cannabis use may be influential during the first few months of college, and interventions prior to matriculation may have limited effect. In addition to prevention studies, Mason et al. (2025) tested a tailored text message-delivered CUD treatment with young adults and found indirect effects on reduction in cannabis use through readiness to change and protective behavioral strategies. Researchers compared several ADIs for the treatment of youth CUD and emphasized the need to explore effective tailoring strategies (Stanger et al., 2020). Developing effective ADIs requires collecting and analyzing relevant, time-varying auxiliary information, such as motives and contextual factors for cannabis use (Bonar et al., 2021, 2022, 2023).
1.4.Contextual factors for cannabis use
The measurement of motives, often referred to as reasons, for cannabis use dates back to the theoretical motivational model and its empirical support that informed the prevention and treatment of alcohol use (Cooper, 1994; Cox and Klinger, 1988, 1990, 2011). Cooper et al. (2016) reviewed the literature on motives for alcohol, cannabis, and tobacco use and substantiated the existence of similar types of motives for cannabis use, but it should be noted that researchers have also identified more detailed motive subscales for cannabis use (Blevins et al., 2016; Bohnert et al., 2018; Lee et al., 2009). While previous studies have primarily assessed motives for cannabis use as static traits, time-varying motives, as momentary processes in daily life, have stemmed from the interaction between contextual factors, such as the broader context, including historical factors (e.g., past cannabis use reinforcement) and traits (e.g., individual differences in impulsivity), as well as the immediate context, including internal states (e.g., negative affect, withdrawal, distress tolerance), situations (e.g., cannabis availability, day of week), and alternative incentives (e.g., coping skills, reinforcing activities) (Votaw and Witkiewitz, 2021). It is important to consider these contextual factors as well as the timing of interventions over time (Lee et al., 2010). However, no studies to date have assessed these contextual factors and the interactions among these factors in relation to motives and the health outcomes of cannabis use.
1.5.Research questions
Previously published secondary analyses of longitudinal data from four studies (West et al., 2024) have revealed associations between specific types of transition in motives and the likelihood of greater future cannabis use. The time-varying daily motives for cannabis use were assessed from established scales, and these data provided little information about the contexts in which transitions occurred, which would help clinicians to develop interventions. Furthermore, in three of those four studies, the samples were exclusively youth and young adults. The present study extends this previous assessment by collecting open-ended self-reported contexts of use in a small but socio-demographically diverse sample and analyzing transitions in motives and associated health outcomes. Our two primary research questions are as follows:
- Are short-term transitions in time-varying motives for cannabis use associated with distal measures of cannabis use behaviors in a sample with a wide range of socio-demographic characteristics?
- Are certain types of transitions associated with particular cannabis use contexts that could contribute to the development of ADIs?
2.Methods
2.1.Participants
The present study was approved by our University IRB (Study ID #HUM00202179). Participants were recruited from the Michigan Institute for Clinical & Health Research (MICHR) database website through an advertisement where prospective participants could either click a link, or click the “I Am Interested!” button, and we sent the link to the email address on file with MICHR. Prospective participants completed the screener survey on the web, and those who had used cannabis in the past 30 days were eligible to participate in the study. We set balanced quotas for five age groups (18–30, 31–50, 51–60, 61–70, and 70 +), sex, and ethnic minorities to ensure a diverse sample. Selected participants (n = 50) received web surveys by email and text (if a phone number was provided) and completed a baseline survey on demographics and mental health, 28 daily surveys on motives for cannabis use (if cannabis use was indicated yesterday), and a follow-up survey on mental health at present and a variety of cannabis use behaviors over the past month. Only one participant opted out after the baseline survey; for the 28-day survey, 49 participated at least 19 days, 48 at least eight days, and 47 at least one day; and all 49 participants completed the follow-up survey.
2.2.Measures
2.2.1.Baseline survey on demographics and mental health
Eligible participants completed a baseline survey that assessed demographics and mental health. We assessed positive mental health using the Flourishing Scale (Diener et al., 2010), depression using the Patient Health Questionnaire (PHQ-9; Kroenke et al., 2001), anxiety using the Generalized Anxiety Disorder scale (GAD-7; Spitzer et al., 2006), and loneliness using the Three-Item Loneliness Scale (Hughes et al., 2004).
2.2.2.Daily survey on motives for cannabis use
One week after the baseline survey invitation, participants received a survey once a day for 28 consecutive days and were asked to complete it each day. We asked participants how many times they had used cannabis on the previous day, separated by at least 1 h in between each use. If they had used at least once, they were asked the motives behind that use. We assessed ten motives that were modified from the Comprehensive Marijuana Motives Questionnaire (Lee et al., 2009) and included a pain relief motive (Brammer et al., 2022). These motives had high levels of endorsement and tended to change over time, drawing on the results of previous secondary analyses (West et al., 2024). We asked participants to indicate whether or not (yes/no) they used cannabis yesterday to enjoy the feeling or get high, to avoid feeling left out, to expand awareness or understand things differently, to relax or improve mood, to have fun or make a social gathering more fun, to feel more self-confident or more creative, to help you sleep, because it is a low-risk substance, because it was readily available, or to help relieve physical pain, and we asked them to explain other reasons, if any. Participants could endorse multiple motives within a single day. To assess the contextual factors, we also asked participants to describe what they were doing when they used cannabis in an open-ended text box. Once each day, a reminder text and e-mail were both sent to non-responsive participants.
2.2.3.Follow-up survey on mental health at present and cannabis use behaviors during the study
The day after the last daily survey was sent, participants received the follow-up survey on mental health at present and cannabis use behaviors during the study. We used the same mental health scales as in the baseline survey. Regarding cannabis use behaviors, we asked participants how many occasions they had used cannabis in the past week.
2.3.Data analysis
All analyses were conducted using R version 4.3.3 (R Core Team, 2023) and Mplus version 8.9 (Muthen and Muthen, 2017). The code is available in the supplemental materials. We applied RI-LTA (Muthén and Asparouhov, 2022) in Mplus to the daily indicators of motives for cannabis use and cannabis use behaviors collected in the study. We found that day-to-day transitions in latent motive classes were rare, so we recoded the daily indicators of motives as weekly indicators that represented whether any particular motive was endorsed on weekdays (i.e., Monday through Thursday) and/or the weekend (i.e., Friday through Sunday) in a given week. For example, if a participant endorsed a given motive on any weekday of the week, then the corresponding weekly indicator for that motive became activated.
We estimated the probabilities of membership in latent classes characterized by specific motives in Week 1 and Week 4 in the RI-LTA, which separates between-subject variation from the within-subject latent class transitions over time by allowing random intercept variation (Muthén and Asparouhov, 2022). We also estimated the probabilities of transitioning from one motive class to another between Week 1 and Week 4 as well as the corresponding means of the past-week cannabis use occasions in the follow-up survey as the distal measure of future cannabis use behavior. We excluded indicators for Week 2 and Week 3 from the analysis due to the limited sample size relative to the number of parameters that would need to be estimated if these two intermediate weeks were included.
As demonstrated by West et al. (2024), we used commonly reported measures of model fit for RI-LTA (e.g., BIC and entropy) to ultimately select the best-fitting models in terms of the number of latent motive classes at the two time points. We then included the distal outcome to estimate the transition-specific means. It should be noted that when following this manual two-step approach, the inclusion of the distal outcome in the second step could influence the latent class formation, and the transition-specific distribution of the distal outcome cannot be interpreted as the direct effect of a transition, as the distal outcome itself was also considered an indicator of the latent motive class variable (Nylund-Gibson et al., 2019). In these cases, it is indeed recommended to use multi-step methods such as the BCH method to ensure that the latent class membership does not change, as the reviewer notes (Asparouhov and Muthén, 2021). We re-ran all of our analyses using the BCH method for RI-LTA (Muthén and Asparouhov, 2025), which had not been fully developed when we first conducted the analyses for this paper, but our small sample size in the present study unfortunately led to model non-convergence (see the supplemental material for syntax). Although it is assumed that the measurement of the latent classes is the same across time in most applications of LTA, the measurement of the latent motive classes can be qualitatively different across time, and the transition probabilities are interpreted as the probability of changing between qualitatively distinct motive classes (Nylund-Gibson et al., 2023).
3.Results
3.1.Demographics
Of the 50 participants, two did not report the past-week cannabis use occasions in the follow-up survey and were excluded from the analysis. Table 1 shows the demographics and descriptive statistics.
| Mean | SD | |
|---|---|---|
| Age | 48.79 | 17.06 |
| n | % | |
| Sex | ||
| Female | 32 | 64 |
| Male | 18 | 36 |
| Sexual Orientation | ||
| Heterosexual | 42 | 84 |
| Other | 8 | 16 |
| Race | ||
| African American | 11 | 22 |
| White | 31 | 62 |
| Other | 8 | 16 |
| Citizenship | ||
| US Citizen | 48 | 96 |
| Other | 2 | 4 |
| Education Level | ||
| 8th Grade or Lower | 1 | 2 |
| High School Degree | 11 | 22 |
| Some College (But No College Degree) | 8 | 16 |
| Associate’s Degree | 4 | 8 |
| Bachelor’s Degree | 15 | 30 |
| Graduate Degree | 11 | 22 |
3.2.Latent classes, transitions, and associated distal cannabis use behavior
We identified four types of transitions from two latent motive classes on weekdays in Week 1 to two classes on weekdays in Week 4, and estimated the associated means of the past-week cannabis use occasions in the follow-up survey. We excluded one motive assessing conformity (i.e., avoid feeling left out), for which the low probability of endorsement did not distinguish between motive classes. The patterns of transitions in the weekend latent motive classes were not evident, with most participants remaining in the same classes. We thus reported and interpreted the transitions in the weekday latent motive classes. See the supplementary materials for additional results related to the weekend latent classes and transitions.
Fig. 1 shows the estimated probabilities of endorsing the various motives in two classes on weekdays in Week 1 and Week 4. In Week 1, one motive class was defined by multiple motives except for feeling more self-confident or more creative; another was primarily defined by enjoyment, relaxation, sleep aid, low risk, and availability; the separation of the two classes was moderate. In Week 4, compared with the multiple motives class in Week 1, the multiple motives class had relatively higher probabilities of endorsing cannabis use to feel more self-confident or more creative; the other motive class was primarily defined by sleep aid and availability.
While the cannabis use occasions asked about in the follow-up survey covered the past week that overlapped with Week 4 weekdays, these occasions also covered the Week 4 weekend, which was considered distal relative to the weekdays. Table 2 shows the estimated probabilities of the four types of transition and the estimated means of the follow-up past-week cannabis use occasions associated with each type of transition. From Week 1 to Week 4, transitioning from the motive class defined by enjoyment, relaxation, sleep aid, low risk, and availability to the motive class defined by sleep aid and availability (Transition 4) was associated with highest estimated mean of cannabis use occasions (45.00), although this transition was unlikely to happen (estimated probability of 0.05); remaining in the multiple motives class (Transition 1) or transitioning to the motive class defined by sleep aid and availability (Transition 2) was also associated with a high estimated mean of cannabis use occasions (16.00 and 12.36, respectively). In contrast, transitioning from the motive class defined by enjoyment, relaxation, sleep aid, low risk, and availability to the multiple motives class (Transition 3) was extremely likely to happen (estimated probability of 0.95) and was associated with the lowest estimated mean of cannabis use occasions (3.86). See the supplementary materials for demographic statistics and mental health assessments at baseline and follow-up for each type of transition.
| Week 1 Weekdays Motive Class | Week 4 Weekdays Motive Class | Estimated Transition Probability | Estimated Mean of Follow-up Past-Week Cannabis Use Occasions (SE) | |
|---|---|---|---|---|
| Transition 1 | Multiple Motives | Multiple Motives | 0.31 | 16.00 (1.67) |
| Transition 2 | Multiple Motives | Sleep Aid, Availability | 0.69 | 12.36 (3.66) |
| Transition 3 | Enjoyment, Relaxation, Sleep Aid, Low Risk, Availability | Multiple Motives | 0.95 | 3.86 (0.53) |
| Transition 4 | Enjoyment, Relaxation, Sleep Aid, Low Risk, Availability | Sleep Aid, Availability | 0.05 | 45.00 (2.12) |
Of the three transitions associated with high estimated means of follow-up past-week cannabis use occasions, only three participants experienced Transition 1, seven participants experienced Transition 2, and only two participants experienced Transition 4. We focused our qualitative analysis on participants going through Transitions 1, 2, and 4, which were associated with relatively high estimated means of cannabis use occasions.
3.3.Mental health in transitions 1, 2 and 4
Table 3, Table 4, and Table 5 focus on the mental health for each participant who experienced Transition 1, Transition 2, and Transition 4.
| Participant 1 | Participant 2 | Participant 3 | |
|---|---|---|---|
| Baseline | |||
| FS | 56 | 49 | 46 |
| PHQ-9 | Mild | Mild | Mild |
| GAD-7 | Minimal | Mild | Moderate |
| UCLA-3 | Not lonely | Lonely | Not lonely |
| Follow-up | |||
| FS | 56 | 48 | 48 |
| PHQ-9 | Minimal | Mild | Moderate |
| GAD-7 | Minimal | Mild | Mild |
| UCLA-3 | Not lonely | Not lonely | Not lonely |
| Follow-up Past-Week Cannabis Use Occasions | 14 | 14 | 20 |
| Participant 4 | Participant 5 | Participant 6 | Participant 7 | Participant 8 | Participant 9 | Participant 10 | |
|---|---|---|---|---|---|---|---|
| Baseline | |||||||
| FS | 50 | 49 | 48 | 36 | 44 | 41 | 35 |
| PHQ-9 | Minimal | Minimal | Minimal | Mild | Minimal | Minimal | Moderate |
| GAD-7 | Minimal | Minimal | Minimal | Mild | Mild | Minimal | Severe |
| UCLA-3 | Not lonely | Not lonely | Not lonely | Lonely | Not lonely | Not lonely | Lonely |
| Follow-up | |||||||
| FS | 51 | 52 | 50 | 51 | 45 | 44 | 53 |
| PHQ-9 | Mild | Minimal | Minimal | Minimal | Mild | Mild | Minimal |
| GAD-7 | Moderate | Minimal | Minimal | Mild | Moderate | Minimal | Minimal |
| UCLA-3 | Lonely | Not lonely | Not lonely | Lonely | Lonely | Not lonely | Not lonely |
| Follow-up Past-Week Cannabis Use Occasions | 18 | 15 | 7 | 10 | 13 | 10 | 12 |
| Participant 11 | Participant 12 | |
|---|---|---|
| Baseline | ||
| FS | 44 | 43 |
| PHQ-9 | Minimal | Moderate |
| GAD-7 | Minimal | Mild |
| UCLA-3 | Lonely | Not lonely |
| Follow-up | ||
| FS | 45 | 44 |
| PHQ-9 | Minimal | Mild |
| GAD-7 | Minimal | Mild |
| UCLA-3 | Not lonely | Not lonely |
| Follow-up Past-Week Cannabis Use Occasions | 48 | 42 |
Participant 1 shifted from mild to minimal depression symptoms. Participant 2 shifted from feeling lonely to not feeling lonely. Participant 3 shifted from mild to moderate depression symptoms, and from moderate to mild anxiety symptoms.
Participant 4 shifted from minimal to mild depression symptoms, from minimal to moderate anxiety symptoms, and from not feeling lonely to feeling lonely. Participant 8 shifted from minimal to mild depression symptoms, from mild to moderate anxiety symptoms, and from not feeling lonely to feeling lonely. Participant 9 shifted from minimal to mild depression symptoms. Compared to these participants, the mental health of the other four participants shifted toward positive, especially Participant 7, who shifted from mild to minimal depression symptoms, and Participant 10, who shifted from moderate to minimal depression symptoms, from severe to minimal anxiety symptoms, and from feeling lonely to not feeling lonely.
Participant 11 shifted from feeling lonely to not feeling lonely. Participant 12 shifted from moderate to mild depression symptoms.
3.4.Other contextual factors for cannabis use in transitions 1, 2 and 4
Table 6, Table 7, and Table 8 show the self-reported contexts of cannabis use for each participant who experienced Transition 1, Transition 2, and Transition 4.
| Participant 1 | Participant 2 | Participant 3 | |
|---|---|---|---|
| Week 1 Monday | I was camping. spending with family & friends | Anxiety; Cleaning house | Was camping which included a number of different activities such as fishing, hiking, etc. |
| Week 1 Tuesday | Every day chores | Anxiety; watching tv | Attended a social function for my partners birthday, celebrated at a number of different locations throughout the evening |
| Week 1 Wednesday | Every day activities | Anxiety; Sitting outside relaxing | Went out with a few friends to celebrate a birthday and partook then rather than drinking, then before bed for sleeping |
| Week 1 Thursday | Completing paperwork for family, phone calling for aoots, household chores, watching television | Anxiety and Depression; Watching tv | Used after work to relax while doing house chores and then before bed for sleep |
| Week 4 Monday | Relaxing, chores, watch tv, crosswords | Grieving; Nothing | Used after a fight with partner to relax and help refocus myself, then at bedtime for relaxing and sleep |
| Week 4 Tuesday | Life, sewing, traveling, watching movies, paperwork | Used in the evening to unwind and relax after work then before bed for sleep | |
| Week 4 Wednesday | Crafting, cleaning hime, relaxing, gaming | Watching tv | Used several times after work while doing house chores then before bed for relaxing and sleep |
| Week 4 Thursday | Household chores, paperwork, reading, watching tv | Depression; Nothing | Used while cleaning car and doing other miscellaneous tasks around house, then some while watching tv before bed to sleep |
| Follow-up Past-Week Cannabis Use Occasions | 14 | 14 | 20 |
| Participant 4 | Participant 5 | Participant 6 | Participant 7 | Participant 8 | Participant 9 | Participant 10 | |
|---|---|---|---|---|---|---|---|
| Week 1 Monday | Pain relief while doing everyday activities | Walking around my yard to observe my plants and garden. | Relaxing at home before bedtime. | I used it when I woke up and then periodically throughout the day | Was home didn’t go out wanted to enjoy some lone time and I was listening to music. | Watching tv. Socializing | Relaxing in my chair, listening to music |
| Week 1 Tuesday | Relaxing | Sitting in my living room watching TV. Walking around in my yard. | Relaxing at home with the family | making breakfast, driving, working, after dinner with friends and before bed | Was relaxing after a day’s work | Running errands, cooking | Relaxing |
| Week 1 Wednesday | exercising, relaxing | Sitting in my living room | Relaxing at home. | after dinner and right before bed | I was happy | Cleaning, socializing | Watching a movie |
| Week 1 Thursday | Aid with chemo-induced neuropathy; Relaxing. Taking a break. Performing chores. | Appetite enhancing; Sitting in my living room. | Having dinner with family. | before bed | Got stuck on problem, used it to enlighten myself and eventually pulled through | Social media. Socializing with a friend | I was listening to music |
| Week 4 Monday | Anxiety; Chores, relaxing | Getting ready to go to bed. | Playing with grandkids | Relaxing | |||
| Week 4 Tuesday | Anxiety, depression; PT Exercise, mood stabilizer | While watching TV before bed | Getting ready for bed. | before bed | Watching anime | Looking for work | Relaxing |
| Week 4 Wednesday | Anxiety, depression; Relaxing, chores | Sitting in my living room watching TV. | Relaxing before bed. | Working | Relax | ||
| Week 4 Thursday | To help get through a very difficult day; Anxiety, coping | Sitting on my couch while watching TV | Relaxing before bedtime. | Watching tv | Relaxing | ||
| Follow-up Past-Week Cannabis Use Occasions | 18 | 15 | 7 | 10 | 13 | 10 | 12 |
| Participant 11 | Participant 12 | |
|---|---|---|
| Week 1 Monday | I sit in my porch or in my garage and smoke about a half joint at a time. | Cleaning house, making food, watching tv |
| Week 1 Tuesday | Sitting on the porch, i don’t smoke in the house, either getting ready for bed, waking up, or because i felt like a smoke. | Taking a couple puffs off a pipe seems to quiet the endless loop of chatter reminding me of all the things that need my attention. Often helps me focus on tasks. Can also demotivate further action if tired; Planning the day. Working outside. Talking with friends on phone or in person |
| Week 1 Wednesday | Sitting on my porch each time getting ready for bed or waking up or just relaxing and watching tv | Lots of tasks to accomplish; Establishing plan of action/activities. Stayed away from morning news. |
| Week 1 Thursday | Sitting on the porch as usual | Used prior to appointment in morning and after returning home in evening |
| Week 4 Monday | Waking up, going to bed, decorating, sitting on the porch | Hanging at home |
| Week 4 Tuesday | Waking up, going to bed, doing projects, sitting on the porch | Home, few errands |
| Week 4 Wednesday | Waking up, going to bed, playing video games, sitting on the porch | Post chemo after 5 week delay from complications healing; Laying low, jittery from steroid drugs |
| Week 4 Thursday | Waking up, going to sleep, playing video games | Laying low, post chemo, a bit tired and headachey; Low energy house stuff, napping |
| Follow-up Past-Week Cannabis Use Occasions | 48 | 42 |
Participant 3 provided details about the context of cannabis use. One particular incident mentioned was that on Tuesday during Week 1: “Attended a social function for my partner’s birthday, celebrated at a number of different locations throughout the evening”, and on Monday during Week 4: “Used after a fight with partner to relax and help refocus myself, then at bedtime for relaxing and sleep”. This seemed to be associated with the shift toward moderate depression symptoms. Relaxation and sleep aid motives also became more prominent in Week 4 compared to Week 1, while enjoyment and social fun motives were lacking.
Participant 4 reported using cannabis on Thursday in Week 1 to “aid with chemo-induced neuropathy”. During Week 4, the participant used cannabis to cope with anxiety, specifically on Tuesday for mood stabilization, and on Thursday “to help get through a very difficult day.” This might be associated with the shift toward mild depression symptoms, moderate anxiety symptoms, and feeling lonely, while the participant did not mention these motives during Week 1. Although the participant did not provide more details in Week 4, the emphasis on these mental health-related motives suggested that their impact on cannabis use could have gone beyond the relaxation motive that was frequently mentioned during both weeks.
Participant 8 provided some details during Week 1 but not in Week 4. In Week 1, the participant “wanted to enjoy some lone time”, “was happy”, and “got stuck on problem, used it to enlighten myself and eventually pulled through”. However, the participant felt lonely, more anxious and depressed in Week 4 and used more cannabis. It was hard to figure out potential contextual or other precipitating factors, because the only thing we knew was that the participant watched anime on Tuesday in Week 4. Participant 9 did not provide too many details either. The shift toward mild depression symptoms might be associated with work that was mentioned on Tuesday and Wednesday in Week 4 but not in Week 1.
Participant 12 reported using cannabis while planning activities, working outside, and talking with friends on the phone or in person during Week 1, specifically mentioning that it helped to focus on lots of tasks. However, during Week 4, the participant spent most of the time at home doing “low energy house stuff” and “napping”. Although it did not appear that the mental health of the participant was turning negative, on Wednesday and Thursday in Week 4, the participant experienced tiredness and headaches due to delayed healing from postchemotherapy complications, as well as feeling low and “jittery from steroid drugs”.
4.Discussion
The present study performed descriptive analyses of short-term transitions in time-varying motives for cannabis use and associated distal health outcomes in a small but socio-demographically diverse sample. We identified two latent classes of cannabis use motives on weekdays in Week 1—the multiple motives class and the motive class defined by enjoyment, relaxation, sleep aid, low risk, and availability-—and two on weekdays in Week 4—the multiple motives class and the motive class defined by sleep aid and availability. We found that three types of transitions were associated with frequent past-week cannabis use as assessed in the follow-up survey—remaining in the multiple motives class (Transition 1), transitioning from the multiple motives class in Week 1 to the sleep aid and availability class in Week 4 (Transition 2), and transitioning from the enjoyment, relaxation, sleep aid, low risk, and availability class in Week 1 to the sleep aid and availability class in Week 4 (Transition 4). However, these associations should not be interpreted causally.
Previous studies have suggested that remaining in or transitioning to the multiple motives class from one where fewer motives were endorsed generally had more frequent cannabis use in the future (Espinosa et al., 2023). Our analyses of Transition 1 also suggested that remaining in the multiple motives class was associated with frequent cannabis use. However, in our analyses of Transition 3, which was experienced by the majority of participants, transitioning to the multiple motives class was not associated with frequent cannabis use. Even though the endorsement of multiple motives was associated with adverse health outcomes, it may be impractical for clinicians to implement interventions that target all motives for cannabis use, and certain motives may be more influential. Bonn-Miller et al. (2014) and Drazdowski et al. (2021) found that the sleep aid motive was associated with increased and problematic cannabis use. Our analyses of Transition 2 and Transition 4 also suggested that specific motives such as sleep aid and availability appeared to have a greater impact on participants who experienced these transitions. These contradictory findings underscored the importance of understanding the context of cannabis use before acting to address the motives for its use.
To understand these transitions and associated cannabis use behaviors, we examined mental health as assessed in the baseline and followup surveys before and after transitions, as well as open-ended, self-reported contexts of cannabis use in the daily surveys. While depression, anxiety, and loneliness did not appear to be potential factors in cannabis use behavior for participants who experienced Transition 1 and Transition 4, we found that some participants who experienced Transition 2 experienced a shift to more severe depression symptoms, anxiety symptoms, and loneliness, which might have been associated with more frequent cannabis use. While we were working with a small sample given the intensive data collection protocol, this qualitative analysis provided interesting insights into the unique context associated with these types of transitions.
What types of ADIs might these results suggest? These interventions might focus on behavioral activation strategies to address depressive symptoms, anxiety management techniques, and exploring healthier alternatives for managing sleep disturbances. The contextual data collected from participants (e.g., stressors related to chemotherapy-induced neuropathy or acute anxiety triggered by specific daily challenges) further highlighted the need for context-sensitive interventions. Real-time interventions could proactively address identified contexts by delivering targeted support precisely when these contextual triggers arise. In current survey design, participants were instructed not to report a second use episode if it occurred within one hour of a previous episode, which may have reduced the amount of contextual information captured—particularly among individuals who segment a single smoking session into multiple parts or who consume edibles with prolonged effects. This constraint likely produced conservative estimates of contextual variability and should be relaxed or redesigned in future studies.
As Lee et al. (2010) emphasized in an intervention study targeting the transitional period before and after matriculation into college, the timing of interventions may have important implications, as the context of cannabis use can be influential during the first few months of college. In the present study, we detected transitional periods with the RI-LTA and attempted to understand these transitions from self-reported contexts that indicated timing of intervention. Addressing contextual factors of cannabis use may facilitate readiness to change and protective behavioral strategies that mediate substance use treatment effects, such as text-delivered brief motivational interviewing-informed treatment (Mason et al., 2025). Although the sample size was small in the present study, we see this as a promising approach for combining statistical analysis and qualitative assessment in the development of ADIs. Future research could replicate this approach with larger samples and encourage participants to provide detailed contextual information on cannabis use. With larger samples, clinicians may be able to associate particular cannabis use contexts with certain types of transitions and develop tailored interventions and treatment strategies accordingly.
4.1.Limitations
The present study employed self-administered web surveys that allowed participants to disclose more information about the contexts of cannabis use relative to prior studies (e.g., Williams et al., 2022). In this way, we were able to conduct more frequent or regular assessments at lower costs. However, several limitations should be considered.
First, the sample size restricted the generalizability of the findings and limited statistical power. The quota-based sampling approach limited our ability to generalize the findings to broader populations. We intended to demonstrate the feasibility of this approach in a sample that included representatives from all groups, including adults. The results reported in the present study should be treated as descriptive and specific to this small pilot study, and could be used to motivate testable hypotheses for future intervention studies. Future research should replicate the present study with larger, more representative samples that are reflective of cannabis use distributions across various age groups, ultimately producing more generalizable and informative findings that can robustly inform the designs of future interventions. Larger samples would also facilitate the identification of more potential motive classes with clear distinctions, as found in a previous study (West et al., 2024), and allow for the use of multi-step methods such as the BCH method to avoid the undesirable shift of motive class membership (Asparouhov and Muthén, 2021), the analyses of transitions between more than two time points, and the inclusion of confounders such as baseline cannabis use frequency.
Second, additional contextual variables, such as religious affiliation, ethnicity, disability status, veteran status, or social acceptance of cannabis use within the community were not assessed in the present study. These variables may impact motives and contexts of cannabis use, suggesting that future studies should incorporate these factors for a more comprehensive understanding. In addition, the present study did not distinguish between medical and non-medical cannabis use cases. Exploring behavioral differences between these contextual groups could provide valuable insights for tailoring ADIs; for example, given the changes that participants may experience, the ADI may require a revision in response to changes in the motive class throughout the observation period.
Finally, although the present study implemented daily web surveys, the retrospective self-reporting of cannabis use motives and contexts may be subject to memory bias, potentially impacting data accuracy. Future research could employ ecological momentary assessment (EMA) techniques to minimize recall bias. Also, because motives were collected once per day, our ability to characterize within-day context was constrained when multiple use episodes occurred in close succession. This may have reduced contextual granularity—particularly among participants who segment a single smoking session into multiple parts or who consume edibles with prolonged effects—yielding conservative estimates of contextual variability. Future work should relax this constraint and use EMA or event-based, time-stamped logging to capture closely spaced, overlapping, or extended episodes so that intervention timing more accurately reflects lived patterns of use.
The present study drew on a previous study in which motives were assessed at different frequencies (West et al., 2024) to select measures for daily motives. Future research should validate the measures and ensure that the measures have favorable psychometric properties and measurement invariance across time in daily surveys.
In conclusion, the approach demonstrated in the present study can be expanded to identify and understand transitional periods in motives for cannabis use that are associated with emerging mental health symptoms and problematic cannabis use in the future. The open-ended self-reported contextual information could be used to further inform the design of ADIs that effectively address active symptoms and curb problematic cannabis use over time, ideally leading to improvements in health.
Supplementary Material
Funding
This work was supported by the National Institutes of Health (NIH) under grant P50DA05403901 from the National Institute on Drug Abuse, supporting Brady T. West and Yongchao Ma; the National Institute on Drug Abuse under grant R01DA031160, supporting Sean Esteban McCabe. The funders played no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the online version at doi:10.1016/j.drugalcdep.2025.112974
Footnotes
Footnote Group
Data availability
The data used in this research are not currently available. Steps are actively being taken to make the data available through openICPSR.
References
Untitled section
References
- Arterberry BJ, Goldstick JE, Walton MA, Cunningham RM, Blow FC, Bonar EE, 2021. Alcohol and cannabis motives: differences in daily motive endorsement on alcohol, cannabis, and alcohol/cannabis co-use days in a cannabis-using sample. Addict. Res. Theory 29 (2), 111–116. 10.1080/16066359.2020.1787390.
- Asparouhov T, & Muthén B 2021. Auxiliary variables in mixture modeling: Using the BCH method in Mplus to estimate a distal outcome model and an arbitrary secondary model. 〈https://www.statmodel.com/examples/webnotes/webnote21.pdf〉
- Bass B, Padwa H, Khurana D, Urada D, Boustead A, 2024. Adult use cannabis legalization and cannabis use disorder treatment in California, 2010–2021. J. Subst. Use Addict. Treat 162. 10.1016/j.josat.2024.209345.
- Blevins CE, Banes KE, Stephens RS, Walker DD, Roffman RA, 2016. Motives for marijuana use among heavy-using high school students: an analysis of structure and utility of the Comprehensive Marijuana Motives Questionnaire. Addict. Behav 57, 42–47. 10.1016/j.addbeh.2016.02.005.
- Bohnert KM, Bonar EE, Arnedt JT, Conroy DA, Walton MA, Ilgen MA, 2018. Utility of the comprehensive marijuana motives questionnaire among medical cannabis patients. Addict. Behav 76, 139–144. 10.1016/j.addbeh.2017.08.001.
- Bonar EE, Cunningham RM, Sweezea EC, Blow FC, Drislane LE, Walton MA, 2021. Piloting a brief intervention plus mobile boosters for drug use among emerging adults receiving emergency department care. Drug Alcohol Depend. 221. 10.1016/j.drugalcdep.2021.108625.
- Bonar EE, Goldstick JE, Chapman L, Bauermeister JA, Young SD, McAfee J, Walton MA, 2022. A social media intervention for cannabis use among emerging adults: randomized controlled trial. Drug Alcohol Depend. 232. 10.1016/j.drugalcdep.2022.109345.
- Bonar EE, Goldstick JE, Collins RL, Cranford JA, Cunningham RM, Chermack ST, Blow FC, Walton MA, 2017. Daily associations between cannabis motives and consumption in emerging adults. Drug Alcohol Depend. 178, 136–142. 10.1016/j.drugalcdep.2017.05.006.
- Bonar EE, Goldstick JE, Tan CY, Bourque C, Carter PM, Duval ER, McAfee J, Walton MA, 2023. A remote brief intervention plus social media messaging for cannabis use among emerging adults: a pilot randomized controlled trial in emergency department patients. Addict. Behav 147. 10.1016/j.addbeh.2023.107829.
- Bonn-Miller MO, Babson KA, Vandrey R, 2014. Using cannabis to help you sleep: heightened frequency of medical cannabis use among those with PTSD. Drug Alcohol Depend. 136, 162–165. 10.1016/j.drugalcdep.2013.12.008.
- Brammer WA, Conn BM, Iverson E, Lankenau SE, Dodson C, Wong CF, 2022. Coping motives mediate the association of trauma history with problematic cannabis use in young adult medical cannabis patients and non-patient cannabis users. Subst. Use Misuse 57 (5), 684–697. 10.1080/10826084.2022.2026970.
- Bray BC, Berglund PA, Evans-Polce RJ, Patrick ME, 2021. A latent transition analysis of self-reported reasons for marijuana use during young adulthood. Eval. Health Prof 44 (1), 9–24. 10.1177/0163278720984514.
- Bresin K, Mekawi Y, 2019. Do marijuana use motives matter? Meta-analytic associations with marijuana use frequency and problems. Addict. Behav 99. 10.1016/j.addbeh.2019.106102.
- Caulkins JP, 2024. Changes in self-reported cannabis use in the United States from 1979 to 2022. Addiction. 10.1111/add.16519.
- Choi NG, Moore J, Choi BY, 2024. Cannabis use disorder and substance use treatment among U.S. adults. J. Subst. Use Addict. Treat 167. 10.1016/j.josat.2024.209486.
- Cooper ML, 1994. Motivations for alcohol use among adolescents: development and validation of a four-factor model. Psychol. Assess 6 (2), 117–128. 10.1037/1040-3590.6.2.117.
- Cooper ML, Kuntsche E, Levitt A, Barber LL, Wolf S, 2016. Motivational models of substance use: A review of theory and research on motives for using alcohol, marijuana, and tobacco. In: Sher KJ (Ed.), The Oxford handbook of substance use and substance use disorders, 1. Oxford University Press, pp. 375–421.
- Cox WM, Klinger E, 1988. A motivational model of alcohol use. J. Abnorm. Psychol 97 (2), 168–180. 10.1037/0021-843X.97.2.168.
- Cox WM, Klinger E, 1990. Incentive motivation, affective change, and alcohol use: A model. In: Cox In.W.M. (Ed.), Why people drink. Gardner Press, pp. 291–311.
- Cox WM, Klinger E, 2011. A motivational model of alcohol use: Determinants of use and change. In: Cox In.W.M., Klinger E (Eds.), Handbook of Motivational Counseling, 1st ed. Wiley, pp. 131–158. 10.1002/9780470979952.ch6.
- Di Forti M, Quattrone D, Freeman TP, Tripoli G, Gayer-Anderson C, Quigley H, Rodriguez V, Jongsma HE, Ferraro L, La Cascia C, La Barbera D, Tarricone I, Berardi D, Szöke A, Arango C, Tortelli A, Velthorst E, Bernardo M, Del-Ben CM, Van Der Ven E, 2019. The contribution of cannabis use to variation in the incidence of psychotic disorder across Europe (EU-GEI): a multicentre casecontrol study. Lancet Psychiatry 6 (5), 427–436. 10.1016/S2215-0366(19)30048-3.
- Diener E, Wirtz D, Tov W, Kim-Prieto C, Choi D, Oishi S, Biswas-Diener R, 2010. New well-being measures: short scales to assess flourishing and positive and negative feelings. Soc. Indic. Res 97 (2), 143–156. 10.1007/s11205-009-9493-y.
- Drazdowski TK, Kliewer WL, Marzell M, 2021. College students’ using marijuana to sleep relates to frequency, problematic use, and sleep problems. J. Am. Coll. Health 69 (1), 103–112. 10.1080/07448481.2019.1656634.
- Espinosa A, Ruglass LM, Conway FN, Jackson KM, White HR, 2023. Motives, frequency, and consequences of cannabis use among college students. J. Drug Issues 53 (1), 61–78. 10.1177/00220426221093608.
- Graupensperger S, Fleming CB, Jaffe AE, Rhew IC, Patrick ME, Lee CM, 2021. Changes in young adults’ alcohol and marijuana use, norms, and motives from before to during the COVID-19 pandemic. J. Adolesc. Health. Off. Publ. Soc. Adolesc. Med 68 (4), 658–665. 10.1016/j.jadohealth.2021.01.008.
- Hughes ME, Waite LJ, Hawkley LC, Cacioppo JT, 2004. A short scale for measuring loneliness in large surveys: results from two population-based studies. Res. Aging 26 (6), 655–672. 10.1177/0164027504268574.
- Kroenke K, Spitzer RL, Williams JB, 2001. The PHQ-9: validity of a brief depression severity measure. J. Gen. Intern. Med 16 (9), 606–613. 10.1046/j.1525-1497.2001.016009606.x.
- Lapham GT, Matson TE, Bobb JF, Luce C, Oliver MM, Hamilton LK, Bradley KA, 2023. Prevalence of cannabis use disorder and reasons for use among adults in a US state where recreational cannabis use is legal. JAMA Netw. Open 6 (8). 10.1001/jamanetworkopen.2023.28934.
- Lee CM, Neighbors C, Woods BA, 2007. Marijuana motives: young adults’ reasons for using marijuana. Addict. Behav 32 (7), 1384–1394. 10.1016/j.addbeh.2006.09.010.
- Lee CM, Neighbors C, Hendershot CS, Grossbard JR, 2009. Development and preliminary validation of a comprehensive marijuana motives questionnaire. J. Stud. Alcohol Drugs 70 (2), 279–287. 10.15288/jsad.2009.70.279.
- Lee CM, Neighbors C, Kilmer JR, Larimer ME, 2010. A brief, web-based personalized feedback selective intervention for college student marijuana use: A randomized clinical trial. Psychol. Addict. Behav 24 (2), 265–273. 10.1037/a0018859.
- Mason MJ, Coatsworth JD, Riggs NR, Russell M, Mennis J, Zaharakis N, Brown A, 2025. Text message-delivered cannabis use disorder treatment with young adults: a large randomized clinical trial. J. Subst. Use Addict. Treat 170. 10.1016/j.josat.2024.209611.
- McCabe SE, Arterberry BJ, Dickinson K, Evans-Polce RJ, Ford JA, Ryan JE, Schepis TS, 2021. Assessment of changes in alcohol and marijuana abstinence, co-use, and use disorders among US young adults from 2002 to 2018. JAMA Pediatr. 175 (1), 64–72. 10.1001/jamapediatrics.2020.3352.
- Muthen LK, & Muthen BO 2017. Mplus User’s Guide (Eighth). Muthén & Muthén. 〈https://books.google.nl/books?id=dgDlAQAACAAJ〉
- Muthén B, Asparouhov T, 2022. Latent transition analysis with random intercepts (RI-LTA). Psychol. Methods 27 (1), 1–16. 10.1037/met0000370.
- Muthén B, & Asparouhov T 2025. Mplus Web Talk No. 10 - Using Mplus To Do Multistep Mixture Modeling: Latent Transition Analysis. 〈https://www.statmodel.com/Webtalk10.shtml〉
- Nahum-Shani I, Dziak JJ, Walton MA, Dempsey W, 2022. Hybrid experimental designs for intervention development: what, why, and how. Adv. Methods Pract. Psychol. Sci 5 (3). 10.1177/25152459221114279.
- Nahum-Shani I, Qian M, Almirall D, Pelham WE, Gnagy B, Fabiano GA, Waxmonsky JG, Yu J, Murphy SA, 2012. Experimental design and primary data analysis methods for comparing adaptive interventions. Psychol. Methods 17 (4), 457–477. 10.1037/a0029372.
- Nahum-Shani I, Smith SN, Spring BJ, Collins LM, Witkiewitz K, Tewari A, Murphy SA, 2018. Just-in-Time Adaptive Interventions (JITAIs) in mobile health: key components and design principles for ongoing health behavior support. Ann. Behav. Med 52 (6), 446–462. 10.1007/s12160-016-9830-8.
- Nylund-Gibson K, Garber AC, Carter DB, Chan M, Arch DAN, Simon O, Whaling K, Tartt E, Lawrie SI, 2023. Ten frequently asked questions about latent transition analysis. Psychol. Methods 28 (2), 284–300. 10.1037/met0000486.
- Nylund-Gibson K, Grimm RP, Masyn KE, 2019. Prediction from latent classes: a demonstration of different approaches to include distal outcomes in mixture models. Struct. Equ. Model. A Multidiscip. J 26 (6), 967–985. 10.1080/10705511.2019.1590146.
- Patrick ME, Bray BC, Berglund PA, 2016. Reasons for marijuana use among young adults and long-term associations with marijuana use and problems. J. Stud. Alcohol Drugs 77 (6), 881–888. 10.15288/jsad.2016.77.881.
- Patrick ME, Evans-Polce RJ, Kloska DD, Maggs JL, 2019. Reasons high school students use marijuana: prevalence and correlations with use across four decades. J. Stud. Alcohol Drugs 80 (1), 15–25. 10.15288/jsad.2019.80.15.
- Patrick ME, Fairlie AM, Cadigan JM, Abdallah DA, Larimer ME, Lee CM, 2019. Daily motives for alcohol and marijuana use as predictors of simultaneous use among young adults. J. Stud. Alcohol Drugs 80 (4), 454–461. 10.15288/jsad.2019.80.454.
- Patrick ME, Miech R, Johnston L, O’Malley P, 2023. Monitoring the future panel study annual report: National data on substance use among adults ages 19–60, 1976-2022. Institute for Social Research. University of Michigan. 10.7826/ISR-UM.06.585140.002.07.0002.2023.
- Patrick ME, Schulenberg JE, O’Malley PM, Johnston LD, Bachman JG, 2011. Adolescents’ reported reasons for alcohol and marijuana use as predictors of substance use and problems in adulthood. J. Stud. Alcohol Drugs 72 (1), 106–116. 10.15288/jsad.2011.72.106.
- Patrick ME, Schulenberg JE, O’Malley PM, Maggs JL, Kloska DD, Johnston LD, Bachman JG, 2011. Age-related changes in reasons for using alcohol and marijuana from ages 18 to 30 in a national sample. Psychol. Addict. Behav 25 (2), 330–339. 10.1037/a0022445.
- Pearson MR, Bravo AJ, Conner BT, Parnes JE, 2020. A day in the life: a daily diary examination of marijuana motives and protective behavioral strategies among college student marijuana users. J. Drug Issues 50 (2), 142–156. 10.1177/0022042619890837.
- Perski O, Hebert ET, Naughton F, Hekler EB, Brown J, Businelle MS, 2022. Technology-mediated just-in-time adaptive interventions (JITAIs) to reduce harmful substance use: a systematic review. Addiction 117 (5), 1220–1241. 10.1111/add.15687.
- R Core Team (2023). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing. 〈https://www.R-project.org/〉
- Shrier LA, Burke PJ, Kells M, Scherer EA, Sarda V, Jonestrask C, Xuan Z, Harris SK, 2018. Pilot randomized trial of MOMENT, a motivational counseling-plus-ecological momentary intervention to reduce marijuana use in youth. –29 mHealth 4, 29. 10.21037/mhealth.2018.07.04.
- Shrier LA, Rhoads A, Burke P, Walls C, Blood EA, 2014. Real-time, contextual intervention using mobile technology to reduce marijuana use among youth: A pilot study. Addict. Behav 39 (1), 173–180. 10.1016/j.addbeh.2013.09.028.
- Spitzer RL, Kroenke K, Williams JBW, Löwe B, 2006. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch. Intern. Med 166 (10), 1092. 10.1001/archinte.166.10.1092.
- Stanger C, Scherer EA, Vo HT, Babbin SF, Knapp AA, McKay JR, Budney AJ, 2020. Working memory training and high magnitude incentives for youth cannabis use: a SMART pilot trial. Psychol. Addict. Behav 34 (1), 31–39. 10.1037/adb0000480.
- Sultan RS, Zhang AW, Olfson M, Kwizera MH, Levin FR, 2023. Nondisordered cannabis use among US adolescents. JAMA Netw. Open 6 (5). 10.1001/jamanetworkopen.2023.11294.
- Votaw VR, Witkiewitz K, 2021. Motives for substance use in daily life: a systematic review of studies using ecological momentary assessment. Clinic. Psychol. Sci. A J. Assoc. Psychol. Sci 9 (4), 535–562. 10.1177/2167702620978614.
- West BT, Ma Y, Lankenau S, Wong CF, Bonar EE, Patrick ME, Walton MA, McCabe SE, 2024. Latent transition analysis of time-varying cannabis use motives to inform adaptive interventions. Psychol. Addict. Behav 10.1037/adb0001012.
- Williams EC, Fletcher OV, Frost MC, Harris AHS, Washington DL, Hoggatt KJ, 2022. Comparison of substance use disorder diagnosis rates from electronic health record data with substance use disorder prevalence rates reported in surveys across sociodemographic groups in the veterans health administration. JAMA Netw. Open 5 (6). 10.1001/jamanetworkopen.2022.19651.
Associated Data
Supplementary Materials
Data Availability Statement
The data used in this research are not currently available. Steps are actively being taken to make the data available through openICPSR.