Negative Affect Regulation and Marijuana Use in College Students: Evaluating the Mediating Roles of Coping and Sleep Motives

Abstract

Negative affect regulation models suggest that marijuana may be used to reduce negative affect. Extant research has provided support for these models, indicating that specific motives for marijuana use, particularly coping motives (i.e., using to alleviate negative affect), mediate relations between affective vulnerabilities and marijuana outcomes. However, sleep motives (i.e., using to promote sleep) have been neglected from such models, despite theoretical relevance. The present study tested two multiple mediation models in a large sample of marijuana-using college students (N = 1,453) to evaluate the indirect effects of coping and sleep motives in paths from depressive and anxiety symptoms to marijuana outcomes (use, consequences, and cannabis use disorder [CUD] symptoms). Both coping and sleep motives mediated the effects of depressive/anxiety symptoms on each marijuana variable. Moreover, significant double mediated effects were found, such that higher affective symptoms were associated with greater motives; which were associated with more marijuana use; which was related to more negative consequences and CUD symptoms. Findings provide support for sleep motives as a relevant pathway between affective vulnerabilities and marijuana outcomes. Additional research is needed to evaluate the potential benefits of interventions targeting specific marijuana motives.

Introduction

According to negative affect regulation models, individuals may be motivated to use marijuana and other substances to avoid negative affect (i.e., experiences of emotional states involving anger, sadness, and/or fear) and manage distressing emotions (Baker et al. 2004; Cooper et al. 2016; Glodosky and Cuttler 2020; Simons et al. 2005). Research providing support for negative affect regulation models has examined marijuana use in response to both temporary negative affective states (Wycoff et al. 2018) and general psychopathology, particularly mood and anxiety disorders (e.g., Bahorik et al. 2017). Indeed, mood and anxiety disorders are characterized by high negative affectivity, and marijuana users report self-medicating to alleviate depressive and anxiety symptoms (Osborn et al. 2015; Kosiba et al. 2019). Additionally, high comorbidity between cannabis use disorder and mood and anxiety disorders has been reported across several studies (Hasin et al. 2016; Kedzoir et al. 2014). Thus, negative affect regulation models appear to be especially salient to marijuana users who experience symptoms of depression and anxiety.

One such high-risk group includes college students, given their high prevalence rates of depression (Auerbach et al. 2016; Bravo, Villarosa-Hurlocker, and Pearson 2018; Eisenberg et al. 2007; Ibrahim et al. 2013 Wang et al. 2020), anxiety (Auerbach et al. 2016; Beiter et al. 2015; Bravo, Villarosa-Hurlocker, and Pearson 2018; Eisenberg et al. 2007; Wang et al. 2020), and marijuana use (Bravo, Villarosa-Hurlocker, and Pearson 2018; Schulenberg et al. 2020; Suerken et al. 2014). Indeed, multiple studies provide support for negative affect regulation models and their applicability to college student populations. For example, college students with depressive and anxiety symptoms report engaging in marijuana use as a strategy to cope with negative affect (Aselton 2012; Buckner et al. 2007; Lee et al. 2009). Further, negative affect has been shown to increase prior to marijuana use (Shrier, Ross, and Blood 2014). Although marijuana users report a variety of reasons for their use (Lee et al. 2009), coping motives (i.e., using to alleviate negative affect) have been emphasized among models of negative affect regulation (Cooper et al. 2016). Specifically, those who endorse greater coping motives tend to report higher levels of depressive symptoms (Bravo et al. 2019; Moitra et al. 2015), anxiety symptoms (Bonn-Miller et al. 2008; Johnson et al. 2009), frequency of use (Bonn-Miller et al. 2008; Bresin and Mekawi 2019), cannabis use disorder symptoms (CUD; Haug et al. 2017; Moitra et al. 2015), and marijuana-related consequences (Bravo et al. 2019; Bresin and Mekawi 2019; Simons et al. 2005). Notably, coping motives have also been found to mediate associations between negative affectivity and marijuana use outcomes in college student (Bravo et al. 2019) and young adult populations (Farris et al. 2016; Johnson et al. 2009). Although there is good support for the role of coping motives in models of negative affect regulation for marijuana use, empirical study has neglected the integration of other motives that may hold relevance.

Beyond coping motives for marijuana use, endorsement of sleep motives (i.e., using to promote sleep) has also been associated with increased depressive symptoms, anxiety symptoms, marijuana use frequency, and negative marijuana outcomes (Blevins et al. 2016; Bonn-Miller, Babson, and Vandrey 2014; Lee et al. 2009; Metrik et al. 2016). Further, college students report elevated rates of sleep disturbance (Gaultney 2010; Lund et al. 2010), and as many as 44% of college student marijuana users report using for sleep promotion (Drazdowski, Kliewer, and Marzell 2021). Importantly, sleep disturbance is a common feature of depressive and anxiety disorders (Nyer et al. 2013; Taylor et al. 2013); as such, sleep motives may present an additional pathway by which college students are motivated to use marijuana to avoid negative affect resulting from and precipitating sleep disturbances. Yet, sleep motives for using marijuana have seldom been incorporated into models of negative affect regulation, especially among college students. Research in this area has instead focused on civilian and veteran populations with posttraumatic stress disorder (PTSD). Specifically, civilians with elevated PTSD symptoms were more likely to use marijuana to promote sleep (Bonn-Miller et al. 2010; Bonn-Miller, Babson, and Vandrey 2014). In veterans, sleep motives mediated associations from affective vulnerabilities (i.e., PTSD and depression) to marijuana use, while coping motives did not (Metrik et al. 2016). Thus, it is reasonable to expect that using marijuana for sleep promotion may play a similar role for avoiding negative affect among other populations at risk for affective disorders and sleep problems, including college students.

In an effort to examine whether sleep motives similarly function to mediate associations between affective vulnerabilities and marijuana outcomes in college students, the present study examined two multiple mediation models in a large sample of marijuana-using college students. Specifically, we simultaneously evaluated the mediating roles of coping and sleep motives in associations between depressive/anxiety symptoms and marijuana outcomes (i.e., frequency of use, negative consequences, CUD symptoms). Given that prior research examining relations between affective disorder symptoms and marijuana use in college student populations has documented differential associations (Walters et al. 2018), as well as the noted high rates of comorbidity between depression and anxiety (Kessler et al. 2005), we evaluated two independent models (one for each affective disorder). For both models, we expected that the relationship between depressive/anxiety symptoms and marijuana outcomes would be mediated by coping and sleep motives such that higher depressive/anxiety symptoms would relate to elevated endorsement of coping and sleep motives, which would be associated with greater marijuana use, which in turn would relate to more negative consequences and CUD symptoms.

Method

Participants and Procedure

College students (N = 4,764) were recruited from Psychology Department Participant Pools from seven universities located in six U.S. states (for more information, see Looby et al. 2021). Based on study aims, the analytic sample for the present study was limited to 1,453 college students who reported past 30-day marijuana use. Within the selected sample, students predominantly reported being White non-Hispanic (47.5%) and female at birth (70.2%), with a mean age of 19.61 years (SD = 2.55). Participants provided informed consent prior to initiating the study and were offered research participation credit at their respective institution for completing the study. The University of Wyoming Institutional Review Board approved all recruitment, consent, and study procedures.

Measures

For all measures (unless noted), composite scores were created by averaging or summing items (reverse-coding items when appropriate) such that higher scores indicate higher levels of the construct.

Mental health was assessed using the DSM-5 Level 1 Cross-Cutting Symptom Measure–Adult (DSM XC; American Psychiatric Association 2013). This 23-item measure assesses 13 domains of mental health; for the present analyses, only items related to anxiety (three items; α = .88) and depression (two items; Spearman-Brown coefficient = .86) subscales were utilized. Participants were asked: “During the past two (2) weeks, how much (or how often) have you been bothered by the following problems?” and responded to questions using a 5-point scale (0 = none, not at all, 4 = severe, nearly every day). Per the scoring instructions, a score of 2 or higher on the anxiety and depression domains is suggestive of clinically-relevant mental health problems (Narrow et al. 2013).

Motives for using marijuana were assessed using the 35-item Comprehensive Marijuana Motives Questionnaire (CMMQ; Lee et al. 2009). Participants were asked to indicate on a 5-point response scale (1 = almost never/never, 5 = almost always/always) how frequently their personal marijuana use was motivated by each of the items in the past 30 days. The CMMQ assesses 12 distinct dimensions of marijuana motives; however for this study, only the coping (three items; α = .84) and sleep (three items; α = .88) subscales were examined. Items from the coping subscale include “To forget your problems”, “Because you were depressed”, and “To escape from your life”, and items from the sleep subscale include “To help you sleep”, “Because it helps make napping easier and enjoyable”, and “Because you are having problems sleeping.”

Marijuana use frequency was assessed using the Marijuana Use Grid (MUG; Pearson and Marijuana Outcomes Study Team 2021). Participants were asked to report at which times they consumed marijuana during a “typical week” over the past 30 days. Each day of the week was broken down into six 4-hour time blocks (12a-4a, 4a-8a, 8a-12p, 12p-4p, 4p-8p, and 8p-12a). Typical frequency of marijuana use was calculated by summing the total number of time blocks for which participants reported using during a typical week (range of possible answers: 0–42). Additionally, type of marijuana product used (plant, edibles, concentrates, others) was assessed using a constant sum approach. Students were asked to calculate the percentage of time they had used various marijuana products (total had to equal 100%) during the past month.

Marijuana use negative outcomes, in the past 30-days, were assessed with the Brief Marijuana Consequences Questionnaire (B-MACQ; Simons et al. 2012). The B-MACQ is a 21-item questionnaire designed to measure various domains of marijuana-related consequences over the past 30 days (α = .89), including those related to negative affect, such as “I have been unhappy because of my marijuana use” and “I have become very rude, obnoxious, or insulting after using marijuana.” Cannabis misuse was assessed with the Cannabis Use Disorder Identification Test - Revised (CUDIT-R; Adamson et al. 2010). The CUDIT-R is an 8-item screening tool used to identify harmful or problematic marijuana use over the past 6 months (α = .86).

Statistical Analyses

In order to test our hypotheses, two multiple mediation models (see Figure 1) were tested using Mplus 8.6 (Muthén and Muthén 1998–2018). Specifically, mediation models were estimated with depressive symptoms and anxiety symptoms as independent predictors (i.e., two models) of marijuana motives, marijuana use frequency, and marijuana negative outcomes (i.e., consequences and CUD risk) such that double-mediated effects were examined (i.e., affective disorder symptoms → coping/sleep motives → marijuana use frequency → marijuana negative outcomes). Coping and sleep motives were entered simultaneously in each model (sex was a covariate in both models). We examined the total, direct, and indirect effects using bias-corrected bootstrapped estimates based on 10,000 bootstrapped samples. Because of our large sample size and to reduce Type 1 error, statistical significance was determined based on 99% bias-corrected bootstrapped confidence intervals that do not contain zero.

Figure 1.

Depicts the significant standardized effects of the path model. Significant associations were determined by 99% bias-corrected standardized bootstrapped confidence interval (based on 10,000 bootstrapped samples) that does not contain zero. Non-significant paths and effects of sex on variables (i.e., covariate) are not depicted for parsimony but are available upon request.

Results

Within the analytic sample and averaged across items of a specific domain, 53.1% (n = 772) reported clinically meaningful depressive symptoms, 55.2% (n = 802) reported clinically meaningful anxiety symptoms, 41.4% (n = 602) reported both, and 32.6% (n = 473) did not report clinically meaningful depressive or anxiety symptoms (i.e., scores were less than 2 for each scale). In addition, 24.3% (n = 353) exceeded the cut-off for probable CUD (score of 13 or higher on the CUDIT-R; Adamson et al. 2010). On average, participants reported using marijuana on 11.65 (SD = 10.78) days over the past month. Further, college students reported the following average percentages for type of marijuana product used over the past month: plant (65.8%), edibles (12.6%), concentrates (17.4%), and ‘other’ (e.g., vaporizer, tincture; 4.2%). Bivariate correlations between study variables and descriptive statistics are presented in Table 1. The total, indirect, and direct effects of the mediation models are summarized in Table 2 and direct effects are depicted in Figure 1.

Table 1.

Bivariate correlations and descriptive statistics among all study variables

	1	2	3	4	5	6	7	8	M	SD
1. DSM-5 Depression	---								1.45	1.19
2. DSM-5 Anxiety	.68	---							1.30	1.14
3. Coping Motives	.43	.41	---						2.29	1.19
4. Sleep Motives	.20	.25	.52	---					2.63	1.32
5. Marijuana Frequency	.08	.08	.25	.35	---				6.76	7.68
6. Consequences	.28	.28	.40	.32	.39	---			4.27	4.59
7. CUDIT	.21	.21	.44	.42	.56	.66	---		8.94	6.75
8. Sex	.09	.11	.07	.07	−.06	−.09	.14	---	0.70	0.46

Note. Sex was coded 0 = male, 1 = female. Significant correlations are in bold typeface for emphasis and were determined by a 99% bias-corrected confidence interval (based on 10,000 bootstrapped samples) that does not contain zero.

Table 2.

Summary of total, indirect, and direct effects of depressive and anxiety symptoms on marijuana use outcomes and cannabis use disorder via marijuana use motives and marijuana use frequency

Marijuana Outcome Variables:	Consequences		CUD Symptoms
Predictor Variable: DSM-5 Depression	β	99% CI	β	99% CI
Total	0.29	0.23, 0.35	0.23	0.16, 0.29
Total indirecta	0.14	0.10, 0.18	0.17	0.12, 0.22
Coping Motives	0.10	0.07, 0.14	0.11	0.08, 0.15
Sleep Motives	0.01	0.001, 0.03	0.02	0.01, 0.03
Marijuana Frequency	−0.01	−0.03, 0.02	−0.01	−0.05, 0.02
Coping Motives – Marijuana Frequency	0.01	0.004, 0.03	0.02	0.01, 0.04
Sleep Motives – Marijuana Frequency	0.02	0.01, 0.03	0.03	0.02, 0.04
Direct	0.15	0.09, 0.22	0.06	−0.002, 0.12
Predictor Variables: DSM-5 Anxiety	β	99% CI	β	99% CI
Total	0.29	0.23, 0.35	0.22	0.16, 0.29
Total indirecta	0.14	0.10, 0.18	0.17	0.12, 0.22
Coping Motives	0.10	0.07, 0.14	0.11	0.08, 0.14
Sleep Motives	0.01	−0.002, 0.03	0.02	0.01, 0.04
Marijuana Frequency	−0.01	−0.03, 0.01	−0.01	−0.05, 0.02
Coping Motives – Marijuana Frequency	0.01	0.004, 0.03	0.02	0.01, 0.04
Sleep Motives – Marijuana Frequency	0.02	0.01, 0.03	0.03	0.02, 0.05
Direct	0.15	0.08, 0.21	0.06	−0.003, 0.12

Note. Significant associations are in bold typeface for emphasis and were determined by a 99% bootstrapped confidence interval (based on 10,000 bootstrapped samples) that does not contain zero.

^aReflects the combined indirect associations within the model.

CUD = Cannabis Use Disorder.

Depression Model

Both coping and sleep motives indirectly related depressive symptoms to marijuana use, negative consequences, and CUD symptoms. For marijuana use, higher depressive symptoms were associated with higher coping and sleep motives; which in turn were associated with higher marijuana use (% variance explained could not be calculated as a suppression effect occurred such that the direct effect of depressive symptoms on marijuana use was negative; but a positive indirect effect was found once mediators were included). For negative consequences, 35.0% of the variability in the relationship between depressive symptoms and negative consequences was explained by coping motives and an additional 4.7% was explained by sleep motives. In the relationship between depressive symptoms and CUD symptoms, 49.5% and 7.6% of the variability was explained by coping and sleep motives, respectively. Moreover, statistically significant double mediated effects were found, such that higher depressive symptoms were associated with greater coping and sleep motives; which in turn were associated with more marijuana use; which in turn was related to more negative consequences and CUD symptoms. Notably, double mediated effects of coping motives → use frequency, and sleep motives → use frequency, accounted for 4.7% and 5.8% of the variability in the relation between depressive symptoms and negative consequences, respectively. When evaluating the relation between depressive symptoms and CUD symptoms, 9.9% of the variability was explained by coping motives → use, and 12.0% was explained by sleep motives → use. Of note, when accounting for all predictors in the full model, a positive direct effect of depressive symptoms on consequences was detected, whereas depressive symptoms were not directly associated with CUD symptoms (see Table 2 and Figure 1 for summary of results).

Anxiety Model

Both coping and sleep motives indirectly related anxiety symptoms to marijuana use and CUD symptoms. For marijuana use, higher anxiety symptoms were associated with higher coping and sleep motives; which in turn were associated with higher marijuana use (% variance explained could not be calculated as a suppression effect occurred such that the direct effect of anxiety symptoms on marijuana use was negative; but a positive indirect effect was found once mediators were included). When evaluating the relation between anxiety symptoms and CUD symptoms, 48.1% and 8.8% of the variability was explained by coping and sleep motives, respectively. Notably, only coping motives emerged as a significant mediator in the relation between anxiety symptoms and negative consequences. Moreover, statistically significant double mediated effects were found, such that higher anxiety symptoms were associated with greater coping and sleep motives; which in turn were associated with more marijuana use; which in turn was related to more negative consequences and CUD symptoms. Further, double mediated effects of coping motives → use frequency and sleep motives → use frequency accounted for 4.5% and 7.2% of the variability in the relation between anxiety symptoms and negative consequences, respectively. When evaluating the relation between anxiety symptoms and CUD symptoms, 9.3% of the variability was explained by coping motives → use and 14.9% was explained by sleep motives → use. Of note, when accounting for all predictors in the full model, a statistically significant positive direct effect of anxiety symptoms on consequences was detected, while anxiety symptoms were not directly related to CUD symptoms (see Table 2 and Figure 1 for summary of results).

Discussion

The present study evaluated a negative affect regulation model for marijuana use and associated outcomes in a large sample of college student users. Specifically, marijuana coping and sleep motives were estimated simultaneously as mediators of associations between depressive and anxiety symptoms and marijuana outcomes. Consistent with previous research (Bravo et al. 2019; Farris et al. 2016; Metrik et al. 2016), results indicated that elevated affective disorder symptoms were associated with increased endorsement of coping motives, negative consequences, and CUD symptoms; coping motives were associated with use frequency and consequences; and coping motives mediated relations between affective disorder symptoms and marijuana outcomes through use. Sleep motives were similarly related to all variables of interest and mediated pathways between depressive symptoms and marijuana outcomes, accounting for a unique proportion of the variance among these relations. Sleep motives also mediated pathways from anxiety symptoms to negative consequences and CUD symptoms through use. Further, significant double-mediated effects were found in both models. These results add to growing literature on the mediational role of marijuana coping motives in relations between affect and marijuana use outcomes in college students (Bravo et al. 2019) and young adults (Johnson et al. 2009; Simons et al. 2005; Vilhena-Churchill and Goldstein 2014), while also demonstrating that sleep motives are germane to negative affect regulation.

As posited by negative affect regulation models of marijuana use (Baker et al. 2004; Simons et al. 2005) and supported by prior research (Bravo et al. 2019; Farris et al. 2016; Metrik et al. 2016), coping motives emerged as a robust mediator between affective vulnerabilities and marijuana outcomes. Notably, coping motives mediated the relation between depressive symptoms and marijuana use, which was not found in prior research among veterans (Metrik et al. 2016). This may be explained by differing sample characteristics, as past year prevalence rates of marijuana use among college students (24.6%; Caldeira et al. 2008) exceed those of veterans (9.0%; Davis et al. 2018).

Expanding upon models of negative affect regulation, we conceptualized sleep motives as another relevant pathway through which affective vulnerability relates to marijuana outcomes, which was supported in the present study. Although this has been demonstrated in past research focusing on civilians and veterans with PTSD (Bonn-Miller et al. 2010; Bonn-Miller, Babson, and Vandrey 2014; Metrik et al. 2016), the present study explored this within a college student population. Our findings are informed by extant literature documenting college students’ proneness to experience poor sleep (Gaultney 2010; Lund et al. 2010) and tendency to engage in deleterious sleep hygiene practices (Forquer et al. 2008). Not only are sleep difficulties common among presentations of depression and anxiety (Lund et al. 2010; Nyer et al. 2013), but poor sleep may also precipitate depression and anxiety (Kahn, Sheppes, and Sadeh 2013). In turn, college students may use marijuana in order to self-medicate sleep problems resulting from negative affect. However, using marijuana to treat sleep problems may confer increased risk for negative marijuana-related consequences, which is supported by previous research documenting that endorsement of sleep motives is associated with greater consequences (Lee et al. 2009). This may be explained by more problematic patterns of cannabis use observed in those with sleep difficulties. Specifically, insomnia symptoms are associated with greater odds of hazardous marijuana use and CUD symptoms (Wong et al. 2019). Additionally, those who use marijuana for sleep promotion may initially experience improved sleep onset, while later developing tolerance to the sleep-enhancing effects of marijuana, which may necessitate increased use in order to receive the same initial benefit (Babson and Bonn-Miller 2014). Taken together, using marijuana for sleep promotion may lead to more hazardous and frequent use over time, thereby increasing risk for marijuana-related negative consequences and CUD. Still, additional research is needed to longitudinally examine whether marijuana use for sleep promotion contributes to problematic patterns of use.

In both models, sleep motives demonstrated a stronger effect on frequency of marijuana use as compared to coping motives. Extant research has demonstrated robust relations between marijuana use frequency and sleep motives (Blevins et al. 2016; Lee et al. 2009), suggesting those who use marijuana for sleep promotion perhaps engage in a more consistent pattern of use. Moreover, increased endorsement of sleep motives may help explain college students’ increased marijuana use relative to other populations, given the prevalence (Gaultney 2010) and chronic nature of sleep disturbances (Taylor et al. 2013) experienced by this population. However, coping motives are more strongly associated with negative consequences (Cooper et al. 2016) and CUD symptoms (Moitra et al. 2015), a finding that was replicated in the present study. In concert, exploring both coping and sleep motives, as was done in the present study, serves to provide a more comprehensive understanding regarding relations between marijuana motives and key marijuana outcomes. Future research may wish to explore the differential effects of marijuana motives on marijuana outcomes, as well as whether combined elevations on specific motives confer increased risk for negative outcomes.

Importantly, our findings regarding sleep motives as a significant mediator provide a more comprehensive understanding of pathways through which affective symptoms relate to marijuana outcomes among college students. Of note, affective symptoms were still directly associated with marijuana consequences when accounting for all other predictors, indicating potential for other mediators that were not accounted for in the present study. Rumination may be one possibility, as ruminative thinking and coping motives indirectly related depressive symptoms to marijuana outcomes in previous research (Bravo et al. 2019). Moreover, rumination is also associated with longer sleep onset (Pillai et al. 2014) and may be one mechanism which influences motives to use marijuana for sleep promotion. Taken together, future research may benefit from including sleep motives in models of negative affect regulation, while also evaluating other possible pathways (e.g., rumination), which may help increase specificity and provide a more complete understanding of negative affect regulation within the context of marijuana use.

Results should be interpreted in light of several limitations. First, a cross-sectional design was employed, which precludes our ability to make inferences regarding causal relations or temporal ordering among variables. While the noted pathways were consistent with those proposed by models of negative affect regulation, differences in directionality are plausible. Longitudinal and ecological momentary assessment (EMA) designs are necessary in order to delineate proper ordering of relationships between affective disorder symptoms, marijuana motives, and associated outcomes within college student populations. Further, results from the present study are interpreted within the context of a relatively homogenous group of college student marijuana users from a select number of four-year universities in the United States, most of whom identified as White non-Hispanic and female at birth. Notably, prevalence rates of cannabis use disorder, marijuana use frequency, and risk for marijuana-related problems among college students appears to be higher for males, relative to females (Calakos et al. 2017; McCabe et al. 2007; Schulenberg et al. 2020). Additionally, relations between mood and anxiety disorder symptoms and marijuana use may vary by gender. For instance, one study found that anxiety symptoms and marijuana use were mediated by coping motives for males, whereas this relation was mediated by social motives for females (Buckner, Zvolensky, and Schmidt 2012). Thus, effect sizes in the present study may have been attenuated due to larger female representation. Moreover, results may not generalize to college students from other regions and backgrounds, as well as non-college student populations. Future research may benefit from exploring gender differences within the context of negative affect regulation for marijuana use, as well as the applicability of such models to diverse populations. Finally, our study employed a brief, self-report measure of depression and anxiety symptoms to index negative affect. Although previous research has supported the clinical utility of the selected measure among college students (Bravo, Villarosa-Hurlocker, and Pearson 2018), we were unable to infer clinical diagnoses from its use. Accordingly, future work may benefit from more comprehensive assessment of psychopathology and other forms of negative affect (e.g., anger), such as through use of clinical interviewing, in order to evaluate similar pathways among clinical samples.

Findings from the present study highlight the relevance of sleep motives as an additional pathway through which college students may use marijuana to manage negative affectivity. Clinical efforts may benefit from enhanced screening of marijuana motives and targeting individual reasons for use in order to prevent and intervene on problematic marijuana use. Although marijuana is purported to enhance sleep and users report subjective improvements to sleep following use (Altman et al. 2019), continuous use of marijuana for sleep promotion may impair long-term sleep quality (Babson, Sottile, and Morabito 2017). Recent research using objective indices of sleep has concluded that marijuana disrupts sleep architecture (Gates, Albertella, and Copeland 2014), and sleep disturbances are central to marijuana withdrawal syndrome (Gates, Albertella, and Copeland 2016) and are linked to increased risk of relapse (Babson, Sottile, and Morabito 2017; Gates, Albertella, and Copeland 2016). Thus, incorporating psychoeducation regarding the effects of marijuana on sleep and using evidence-based sleep treatments, such as brief behavioral treatment for insomnia (BBTI; Troxel, Germain, and Buysse 2012), may be effective means to reduce problematic marijuana use among those with affective disorder symptoms, as well as those who report using marijuana for sleep promotion. Together, the present study highlights the relevance of coping-oriented motives, including sleep motives, within the context of negative affect regulation models of marijuana use. Additional research is needed to further understand associations between negative affect, relevant marijuana motives, and marijuana outcomes and how evidence-based interventions can be applied to target problematic use.