Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Feb 28.
Published in final edited form as: Subst Use Misuse. 2020 Feb 28;55(7):1146–1154. doi: 10.1080/10826084.2020.1729202

Cannabis Use and Emotional Awareness Difficulties in Adolescents with Co-occurring Substance Use and Psychiatric Disorders

Lauren Micalizzi a,b,*, Leslie A Brick c, Sarah A Thomas c, Jennifer Wolff c, Christianne Esposito-Smythers d, Anthony Spirito c
PMCID: PMC7189414  NIHMSID: NIHMS1578927  PMID: 32107955

Abstract

Background

Assessing predictors of cannabis use following adolescent substance use treatment may inform essential treatment elements to be emphasized before discharge. Adolescents with low emotional awareness may have limited resources for identifying and overcoming negative emotions and therefore use cannabis to regulate emotions.

Purpose/objectives

The purpose of this study was to test the hypothesis that emotional awareness difficulties are associated with increased cannabis use across the transition out of substance use treatment. This hypothesis was investigated by applying an autoregressive random-intercept cross-lagged panel-modeling framework to test the fit of alternative models and inform hypotheses about directional associations between cannabis use and emotional awareness difficulties over time.

Methods

Participants were 110 adolescents with co-occurring disorders and their families participating in an intensive home-based treatment trial. Adolescents reported on past 7-day cannabis use and difficulties in emotional awareness at baseline and three follow-up assessments across 12 months.

Results

At baseline, 54% of the sample reported past-week cannabis use. A directional effect was supported such that difficulties with emotional awareness at 3 months’ post-baseline, which corresponded to the approximate end of the treatment program, were associated with increased cannabis use at 6 months’ post-baseline, controlling for the stability of cannabis use and emotional awareness over time. Cannabis use, however, was not associated with subsequent difficulties in emotional awareness (i.e., effects in the opposite direction were not supported).

Conclusions/Importance

Emotional awareness difficulties towards the end of a course of intensive outpatient treatment may be associated with increased cannabis use after the completion of treatment.

Keywords: cannabis use, emotional awareness, adolescents, substance use, treatment, co-occurring disorders

Introduction

The rapid legalization and liberalization of cannabis laws for adults may have implications for adolescent cannabis use, such as increased availability, greater social acceptance, lower cost, the introduction of new formulations of cannabis (Hopfer, 2014), and lower perceived harm of cannabis use (Peters & Foust, 2019). These shifts are concerning because heavy cannabis use in adolescence has been linked to developmental and social problems, such as poorer memory, attentional problems (Dougherty et al., 2013), lower academic performance, truancy (McCaffrey, Liccardo Pacula, Han, & Ellickson, 2010), delinquency (Green, Doherty, Stuart, & Ensminger, 2010), and mental health problems (Arseneault et al., 2002). Rates of substance misuse are particularly high among adolescents with psychiatric diagnoses (Abrantes, Brown, & Tomlinson, 2004; Deas, 2006; Deas-Nesmith et al., 1998). Unaddressed substance use problems among adolescents with primary mental health disorders may contribute to poor response to treatment and recurrent hospitalization (Deas-Nesmith et al., 1998) and identifying common problems that underlie both cannabis misuse and other psychiatric conditions in adolescents may help elucidate processes and refine treatment approaches.

A transdiagnostic factor that is implicated in many forms of psychopathology is emotion regulation—how individuals manage their emotional experiences (Aldao, Nolen-Hoeksema, & Schweizer, 2010). Emotion dysregulation is associated with contemporaneous cannabis use in adolescence (Dorard, Berthoz, Phan, Corcos, & Bungener, 2008) as well as comorbidities that frequently co-occur with cannabis use (e.g., ADHD; Mitchell et al., 2016). The liability conferred by emotion dysregulation to cannabis use and associated comorbidities may be particularly noteworthy during adolescence, which is characterized by rapid development and imbalance in the emotion regulation system, the onset of substance use, and high rates of comorbid substance use/misuse and psychiatric diagnoses (Poon, Turpyn, Hansen, Jacangelo, & Chaplin, 2016).

Emotional awareness, a precursor to emotion regulation, encompasses the abilities to attend to and identify emotions, as well as to understand the type and source of emotions (Boden & Berenbaum, 2011; Dorard et al., 2008; Gottman, Katz, & Hooven, 1997). This facet of emotion regulation may have a robust influence on adolescent substance use because developing the skills to accurately attend to and identify emotions provides adolescents with the opportunity to regulate their emotions through safe and healthy means, and potentially decrease the likelihood that they use substances to regulate emotions that they struggle to identify. There is strong theoretical justification for evaluating emotional awareness, specifically, in the study of emotion-related difficulties among individuals who misuse substances. The cognitive-developmental perspective speculates mechanistic hypotheses involving origins of substance misuse in emotional awareness difficulties. For example, alexithymia, a specific emotional disturbance characterized by a lack of emotional awareness (Samur et al., 2013) has been identified in adolescent and adult substance abusers (Dorard et al., 2008; Troisi, Pasini, Saracco, & Spalletta, 1998).

Individuals with substance dependence have lower levels of emotional awareness relative to individuals without substance dependence (Carton et al., 2010; Thorberg, Young, Sullivan, & Lyvers, 2009; Torrado, Ouakinin, & Bacelar-Nicolau, 2013). Similarly, low levels of emotional clarity (i.e., the extent to which you know, understand and are clear about which emotions you are feeling and why you are feeling them) are associated with higher levels of cannabis consumption and abuse among adolescents and college students (Dorard et al., 2008; Limonero, Tomás-Sábado, & Fernández-Castro, 2006). Difficulties with emotional awareness are considered so central to substance misuse that the National Institute on Drug Abuse (2003) recommends that drug prevention programs target emotional awareness in young children.

Cross-sectional studies of self-reported motives for cannabis use highlight the use of cannabis for emotion-related purposes, such as to boost mood or cope with negative emotions (e.g., Buckner, Keough, & Schmidt, 2007; Moitra, Christopher, Anderson, & Stein, 2015). The prevalence of affect-related cannabis use is increasing. The use of cannabis to cope with negative affect (e.g., anger/frustration, escape problems, relieve stress), resulting in part from poor emotion regulation skills, nearly doubled from 1976–2016 (Patrick, Evans-Polce, Kloska, & Maggs, 2019). One plausible explanation for affect-related cannabis use is that low emotional awareness may interfere with the capacity to deploy emotion regulation skills when needed to cope with stress (Gross & Jazaieri, 2014), which may result in heightened urges to use cannabis (Roos & Witkiewitz, 2017). There have been calls within this field of study to test theoretical models aimed at improving understanding of the association between emotional dysregulation and coping-oriented cannabis use (Bonn-Miller, Vujanovic, & Zvolensky, 2008; Zvolensky, Bernstein, Marshall, & Feldner, 2006). The present study aims to shed light on this issue by evaluating the longitudinal associations between adolescent cannabis use and one facet of emotion regulation, difficulties in emotional awareness, in adolescents with co-occurring substance use and psychiatric disorders.

Due to the lack of prospective work, the directions of effects as well as the mechanisms underlying the association between cannabis use and emotional awareness difficulties among adolescents remains unclear, despite their co-occurrence. As stated above, if the direction of effects is from emotional awareness difficulties to cannabis use, adolescents who are low in emotional awareness may have limited resources for identifying and overcoming negative feelings and thus use substances to regulate emotions. The pharmacological effect of substance use may provide a way for adolescents to regulate their affect or block out feelings (Hessler & Katz, 2010). Alternatively, if the direction of effects is from cannabis use to emotional awareness difficulties, cannabis use may result in poorer emotional awareness through cognitive impairment secondary to cannabis use (e.g., Fontes, Bolla, Cunha, & Almeida, 2018). An equally plausible explanation is that there are reciprocal relations between cannabis use and emotional awareness difficulties that interact over time, such that poor emotional awareness exacerbates later cannabis use and vice versa. Understanding the nature of the relationship between cannabis use and emotional awareness difficulties over the course of time is important as such information may be used to inform and improve interventions for adolescent cannabis use as well as emotional/behavioral problems.

While prior research points to a potential link between cannabis use and emotional awareness, the paucity of longitudinal research hinders determination of directions of effects, a crucial step for identifying modifiable treatment components during intervention development. Further, very few studies focus on early- or mid-adolescence, a critical time period for the development of several emotional capacities, including self-awareness of inner states (Brackett & Mayer, 2003), as well as the onset and escalation of cannabis use (Chadwick, Miller, & Hurd, 2013). As such, the objective of this study is to test the directions of effects between cannabis use and emotional awareness difficulties in a sample of adolescents with co-occurring substance use and psychiatric problems. Although all possible temporal pathways will be systematically tested, based on prior research demonstrating the high prevalence of self-reported affect related motives for cannabis use, it was hypothesized that the primary direction of effects would be from difficulties in emotional awareness to increased cannabis use over time.

Materials and Methods

Participants and Procedure

The sample comprised 110 adolescents (ages 13–18; mean age=15.72; 43% female) and their families recruited from a community clinic program in the northeast region of the United States, that delivered intensive home-based services to adolescents with co-occurring substance use and psychiatric disorders. To be included in the study, participants had to: (a) be between 12–18 years old; (b) report alcohol and/or other substance use in the prior three months; (c) have an Axis I psychiatric disorder based on DSM-IV criteria assessed via a structured diagnostic interview; and (c) be English speaking. Participants were excluded from the study if they: (a) had serious psychotic symptoms (e.g., hallucinations) or a primary diagnosis of an eating disorder or Obsessive-Compulsive Disorder; or (b) were acutely suicidal or homicidal. Follow-up assessments were administered at 3-, 6-, and 12-months post-baseline. The study was approved by the hospital human subjects’ protection committee.

Data for the current study was collected as part of a clinical trial comparing enhanced treatment-as-usual (TAU) to a Cognitive Behavioral Therapy (CBT) protocol designed to simultaneously address both substance use and psychiatric symptoms (Wolff et al., under review). Therapists in the program received a workshop outlining a CBT protocol for adolescents with co-occurring substance use and psychiatric problems. Therapists were then randomly selected to conduct the structured protocol and attend weekly supervision on the implementation of the protocol or conduct TAU with typical clinic supervision procedures. In both conditions, assessment of mood and affect regulation was addressed in treatment, although the emphasis could vary based on the adolescent’s presenting problems. Participants in both conditions received, on average, nine sessions of treatment (Wolff et al., under review).

Diagnostic Categorization of the Sample

The Kiddie Schedule for Affective Disorders and Schizophrenia-Present and Lifetime Version (K-SADS-PL; Kaufman et al., 1997), a semi-structured diagnostic interview, was administered separately to parents and adolescents by a trained master’s level clinician at baseline and a consensus diagnosis was derived in order to characterize the sample. A second clinician rated 20% of all K-SADS interviews and reliability was excellent (k=0.89–1.00 across diagnoses). Cannabis abuse (24.77%) and dependence (61.47%), as diagnosed using DSM-IV criteria on the K-SADS-P were the most frequent substance use disorders in the sample. Diagnoses of alcohol abuse (19.27%) and dependence (15.60%) and other substance use abuse (3.67%) and dependence (13.76%) were less frequent. Approximately half of the sample (58.56%) met criteria for a current Major Depressive Episode, while approximately one-third (33.22%) of the sample was diagnosed with Generalized Anxiety Disorder. Externalizing diagnoses were frequent; the most commonly observed diagnoses were Oppositional Defiant Disorder (65.74%), Conduct Disorder (56.88%), and Attention Deficit/Hyperactivity Disorder (54.55%).

Measures

Cannabis use

Cannabis use was assessed by asking adolescents the number of days they used cannabis in the 7 days prior to each assessment. This measure has been used in prior studies (Spirito et al., 2004; 2011).

Emotional awareness

Difficulties in emotional awareness were evaluated with the 6-item Emotional Awareness subscale of the Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004). The DERS operationalizes and measures emotion dysregulation as a higher-order construct involving multiple, internally consistent lower order dimensions, including difficulties relevant to an individual’s awareness and acceptance of emotional responses and ability to regulate behaviors when under increased affective distress. Items on this subscale include: “I am attentive to my feelings”, “When I’m upset, I take time to figure out what I’m really feeling”, and, “When I’m upset, I acknowledge my emotions.” Higher scores reflect more difficulties in emotional awareness. The DERS has high levels of internal inconsistency (α =. 93; Gratz & Roemer, 2004) and adequate test–retest reliability. In this study, the Emotional Awareness subscale had α =.84, .84, .83, .84 at baseline, 3-, 6-, and 12-months, respectively.

Data Analysis Plan

Random Intercept Cross-Lagged Panel Model (RI-CLPM)

We utilized a random-intercepts cross-lagged panel model (RI-CLPM) to examine cross-lagged effects between cannabis use and difficulties in emotional awareness over the course of the study (Hamaker, Kuiper, & Grasman, 2015). The RI-CLPM is a multilevel extension of the traditional CLPM that separates the within-person variance from between-person variance as a means to address a common error in inference known as the ecological fallacy (i.e., making inferences about individual behavior based on group-level data). Thus, it accounts for trait-level (between-level) differences across people while providing insight into state-level (within-level) differences within people, whereas traditional CLPM confounds these two sources of variation. See Figure 1 for an overview of the model. The RI-CLPM utilizes two random-intercepts to represent the trait-level variation in each domain (i.e., the level of emotional awareness one person has relative to others, RIEA, and the level of cannabis use one person endorses relative to others, RICU) as well as latent factors representing the temporal deviation from expected scores to represent the state-level variation (e.g., the degree to which an individual differs relative to their own expected score for emotional awareness or cannabis use at each assessment point). A correlation between the random intercepts represents the relationship between trait-level cannabis use and trait-level emotional awareness. The autoregressive paths (α, δ) represent the degree to which changes in state-level cannabis use and emotional awareness (within-person effects) can be predicted by deviations from their expected scores. The cross-lagged effects (β, γ) represent the degree to which changes in state-level emotional awareness at one time can predict state-level deviations in cannabis use at the following time point (and vice versa). Finally, residual variances/covariances (u, v), which represent the unexplained within-person variance at the second, third, and fourth occasions, are all fixed to zero such that all variation in observed scores are captured by the state-level and trait-level factor structure. The traditional CLPM is nested within the RI-CPLM when the random intercepts and random intercept correlations are fixed to zero, allowing for a direct model comparison test using the chi-square to determine the optimal model.

Figure 1.

Figure 1.

Open in a new tab

Random intercept cross-lagged panel model depicting the relations between cannabis use (CU) and emotional awareness (EA) over four time points (BL=baseline; 3=3 months’ post-baseline; 6=6 months’ post-baseline; 12=12 months’ post-baseline. RICU=random intercept for cannabis use; RIEA=random intercept for emotional awareness; α, δ = autoregressive paths; β, γ=cross-lagged paths; u, v=residual variances/covariances. Curved, double-headed arrows represent correlations among the variables they connect.

Tests of directional effects

The present study utilized a RI-CLPM to test the fit of four competing models of directional effects (Arnett et al., 2012) designed to inform hypotheses about the longitudinal associations between emotional awareness and cannabis use over time, controlling for within-construct stability and between-construct covariance at previous assessments. The interplay between cannabis use and emotional awareness over time may include the following: (1) Bidirectional (reciprocal) effects: cannabis use and emotional awareness difficulties may mutually exacerbate one another over time (i.e., cannabis use ↔ emotional awareness difficulties); (2) Cannabis-driven effects: increased cannabis use may result in difficulties with emotional awareness over time (i.e., cannabis use → emotional awareness difficulties); (3) Emotionally driven effects: difficulties with emotional awareness may result in higher cannabis use over time (i.e., emotional awareness difficulties → cannabis use); or (4) No reciprocal transactions: cannabis use and emotional awareness difficulties may not mutually exacerbate one another over time.

Analyses were conducted using MPlus structural equation modeling software Version 8 (Muthén & Muthén, 1998–2012). Model testing began by testing the full, bidirectional RI-CLPM with reciprocal effects and comparing it to the traditional CLPM. Next, in Models 2–4 above, we systematically constrained the cross-lagged parameters to test if constraining parameters significantly impacted model fit, relative to the full bidirectional model. The relative fit of each reduced model was determined by the chi-square difference (Δχ2) between the full model and the reduced (nested) model and corresponding change in degrees of freedom (Δdf). A nonsignificant change in chi-square between the full and reduced model indicates that the nonsignificant parameters can be dropped from the model without a significant decrement in overall model fit. The most parsimonious model that did not result in a detriment to model fit was selected.

Model fit was evaluated using model chi-square (χ2), Comparative Fit Index (CFI; values above .9 indicate good fit), Root Mean Square Error of Approximation (RMSEA; values below .05 indicate good fit) and Standardized Root Mean Square Residual (SRMR; values below .05 indicate good fit). Comparisons of non-nested models were conducted using Akaike’s Information Criterion (AIC) and Bayesian Information Criterion (BIC), where lower values indicate more optimal fit. MPlus uses full information maximum likelihood (FIML) estimation to account for missing data. Analyses controlled for treatment condition (CBT versus TAU), sex, and age at all assessment points.

Results

Preliminary Analyses

Missing data were most prevalent at the 3-month assessment (17% missing for emotional awareness and 13% missing for cannabis use). Adolescents in grades 11 and 12 had lower rates of missing data compared to those in grades 8 – 10 (X2 [1, n = 111] = 4.10, p =.043). Attrition was as follows: 12% of participants missed the 3-month follow-up, 14% missed the 6-month follow-up, and 15% missed the 12-month follow-up. There were no statistically significant differences in number of follow-up visits across TAU and CBT groups. TAU and CBT groups did not differ on any socio-demographic variable. Means (standard deviations) for CU were 1.92 (2.53), 2.17 (2.70), 2.63 (2.83) and 3.13 (2.95) at baseline, 3-, 6-, and 12-month follow ups, respectively, and 17.27 (6.09), 16.95 (5.67), 18.34 (5.82), and 17.84 (5.78) for difficulties in emotional awareness at baseline, 3-, 6-, and 12-month follow ups, respectively. The percentages of adolescents reporting any past-week CU at each assessment were: 54% at baseline, 52% at 3 months, 58% at 6 months, and 63% at 12 months. CU and emotional awareness were not statistically different across TAU and CBT groups.

Model-fitting

Model 1: Test of random effects in the bidirectional model

In Model 1, we examined the bidirectional association between cannabis use and emotional awareness difficulties across four time points, while accounting for the between-person effects on cannabis use and emotional awareness and covariates. Adequate model fit was obtained (CFI=.968; RMSEA=.063 [95% CI=.000-.134]; SRMR=.037; χ2=12.971[9], p=.163). Unstandardized estimates for the state variables for both autoregressive and cross-lagged paths are reported in Table 1. Cannabis use at 3 months’ post-baseline predicted cannabis use at 6 months’ post-baseline and earlier emotional awareness was associated with later emotional awareness from 3- to 6-months’ post baseline and 6- to 12-months’ post-baseline. No significant cross-lagged effects were observed at any time point. The correlation between the random intercepts was not significant (i.e., r=.05, p=.971), indicating that there are no stable between-person differences.

Table 1.

Unstandardized parameter estimates (standard errors) from the random-intercept, traditional, and nested cross-lagged panel models.

RI-CLPM CLPM Model 2 Model 3 Model 3b Model 4
Autoregressive
Cannabis Use
 Bl → 3mo 0.21 (0.18) 0.38 (0.10)*** 0.40 (0.10)*** 0.38 (0.10)*** 0.41 (0.10)*** 0.41 (0.10)***
 3mo → 6mo 0.45 (0.14) ** 0.57 (0.09)*** 0.56 (0.09)*** 0.59 (0.09)*** 0.59 (0.09)*** 0.58 (0.09)***
 6mo → 12mo 0.25 (0.16) 0.39 (0.11)*** 0.38 (0.11)*** 0.39 (0.11)*** 0.39 (0.11)*** 0.39 (0.11)***
Emotional Awareness
 Bl → 3mo 0.20 (0.15) 0.31 (0.09)** 0.33 (0.09)*** 0.32 (0.09)*** 0.34 (.009)*** 0.34 (0.93)***
 3mo → 6mo 0.32 (0.18) 0.44 (0.11)*** 0.41 (0.11)*** 0.44 (0.11)*** 0.45 (0.11)*** 0.42 (0.11)***
 6mo → 12mo 0.36 (0.15) * 0.44 (0.10)*** 0.44 (0.10)*** 0.43 (0.10)*** 0.43 (0.10)*** 0.43 (0.10)***
Cross-lagged
Cannabis Use →Emotional Awareness
 Bl → 3mo −0.36 (0.33) −0.25 (0.23) −0.21 (0.23) — — —
 3mo → 6mo −0.53 (0.35) −0.40 (0.22) −0.40 (0.22) — — —
 6mo → 12mo −0.09 (0.32) −0.15 (0.22) −0.15 (0.22) — — —
Emotional Awareness → Cannabis Use
 Bl → 3mo −0.09 (0.06) −0.07 (0.04) — −0.08 (0.04) — —
 3mo → 6mo 0.09 (0.06) 0.09 (0.04)* — 0.089 (0.04)* 0.09 (.043)* —
 6mo → 12mo −0.04 (0.06) −0.04 (0.05) — −0.03 (0.05) — —
Covariance
Cannabis Use/ Emotional Awareness
 Bl −1.44 (1.90) −1.31 (1.46) −1.34 (1.46) −1.28 (1.46) −1.32 (1.46) −1.31 (1.46)
 3mo 0.99 (1.58) 1.38 (1.34) 1.36 (1.39) 1.17 (1.33) 1.21 (1.39) 1.21 (1.39)
 6mo 1.48 (1.32) 1.83 (1.23) 2.05 (1.29) 1.75 (1.28) 1.78 (1.28) 1.95 (1.34)
 12mo 1.00 (1.53) 0.73 (1.42) 0.74 (1.43) 0.69 (1.43) 0.68 (1.43) 0.68 (1.44)
Random Intercepts 0.05 (1.39) — — — — —
Open in a new tab

Note. RI-CLPM=Random Intercept Cross-lagged Panel Model; CLPM=Cross-lagged Panel Model; Model 2: cannabis use to emotional awareness difficulties; Model 3: emotional awareness difficulties to cannabis use; Model 3b= emotional awareness difficulties to cannabis use →the only cross-lagged path estimated was emotional awareness difficulties at 3 months’ post-baseline→cannabis use at 6 months’ post-baseline); Model 4: no reciprocal effects modeled. Assessments: Bl=baseline; 3mo=3 months’ post-baseline; 6mo=6 months’ post-baseline; 12mo=12 months’ post-baseline. Dashes represent constrained paths.

Given that none of the random intercepts (trait variables) were significant, we tested a more parsimonious model corresponding to the traditional CLPM in which the variances and covariance of the random intercept were fixed to zero. The chi-square difference among the two models (change in χ2 [3] =2.432, p=.488) was not significant, indicating that constraining these parameters did not result in significantly worse model fit. Consequently, we proceeded with model testing using the more parsimonious model with no random intercepts for the tests of subsequent models regarding the direction of effects between cannabis use and emotional awareness difficulties.

Models 2–4: Tests of directions of effects

Given that Model 1 included all cross-lagged parameters (i.e., all bidirectional effects), it served as the reference model for all nested comparisons. Unstandardized parameter estimates for the RI-CLPM and the four models (i.e., [1] bidirectional effects [traditional CLPM]; [2] cannabis use→emotional awareness difficulties; [3] emotional awareness difficulties→cannabis use; [4] no reciprocal effects) are presented in Table 1 and model fit statistics are presented in Table 2. Because the DERS subscale reported on here assesses difficulties in emotional awareness, positive estimates for the cross-lagged paths indicate that higher cannabis use is associated with greater difficulties in emotional awareness.

Table 2.

Model fit statistics.

Model χ2 df CFI AIC BIC SRMR RMSEA Δχ2 Δdf p
1: Bidirectional Effects 15.403 12 0.973 4032.716 4183.943 0.040 0.051
(0.000, 0.116)
2: Cannabis→Awareness 23.453 15 0.933 4034.766 4177.892 0.051 0.072
(0.000, 0.125)		 8.05 3 .045
3: Awareness→Cannabis 20.110 15 0.959 4031.424 4174.550 0.057 0.056
(0.000, 0.113)		 4.707 3 .195
4: No Bidirectional Effects 27.920 18 0.921 4033.234 4168.258 0.064 0.071
(0.000, 0.120)		 12.517 6 .051
3b Awareness→Cannabis 23.984 17 0.944 4031.298 4169.023 0.062 0.061
(0.000, 0.113)		 3.9361 1 .047
Open in a new tab

Note. χ2= chi-square; df=degrees of freedom; Δ= change; AIC=Akaike’s Information Criterion; CFI=Comparative Fit Index; RMSEA=Root Mean Square Error of Approximation; SRMR= Standardized Root Mean Square Residual.

1

Δχ2 reported for Model 3b is the difference between Model 4 and Model 3b.

Models 3 and 4 both fit the data well. In Model 3, there was a significant, modest cross-lagged effect from month 3 emotional awareness difficulties to month 6 cannabis use, indicating that difficulties with emotional awareness at 3 months’ post-baseline were associated with increased cannabis use at 6 months’ post-baseline. All other cross-lagged paths were not significant. To determine if Model 3 or 4 was a more optimal representation of the data, a reduced Model 3 was fit in which all cross-lagged parameters were fixed to 0 and the one significant path from Model 3 was freely estimated (Model 3b in Table 1). Because these models are non-nested, the fit of reduced Model 3 was subsequently compared to the fit of Model 4 using the AIC and BIC. Results indicated that reduced Model 3 had lower AIC/BIC values, suggesting that this model is more optimal than Model 4. Thus, the data supports emotionally driven effects such that difficulties with emotional awareness were associated with subsequent cannabis use; however, this effect was only statistically significant during the transition from month 3 to month 6. Figure 2 presents the unstandardized parameter estimates for this final model. We re-ran this final model excluding adolescents without a diagnosis of cannabis dependence and/or abuse at baseline (i.e., 15 participants) and the pattern of findings did not change with the inclusion of these participants.

Figure 2.

Figure 2.

Open in a new tab

Unstandardized parameter estimates from the best fitting cross-lagged panel model.

Note. ***p<.001; *p<.05. CU=cannabis use; EA=emotional awareness problems. Curved, double-headed arrows represent correlations among the variables that they connect. Dashed paths were fixed to 0. Model fit: χ2(17)= 23.984, p=.120; AIC= 4031.298; BIC= 4169.023; CFI= 0.944; RMSEA= 0.061 (95% CI=0.000, 0.113); SRMR= .062.

Discussion

This study examined the directions of effects between cannabis use and emotional awareness difficulties in a sample of treatment-involved adolescents with comorbid psychiatric conditions and substance use problems. Consistent with the hypothesized direction of effects between the two constructs, difficulties in emotional awareness at 3 months’ post-baseline were related to increased levels of subsequent cannabis use at 6 months’ post baseline. All other cross-lagged effects were non-significant. In addition, while the initial model included random effects to account for both state- and trait-level differences in cannabis use and emotional awareness, model testing supported a model in which all random effects were fixed to zero, indicating that variation was predominantly attributable to within-person, rather than between-person, differences.

The model-implied mean (i.e., the intercept) for emotional awareness difficulties at 3 months’ post-baseline was lower relative to the other assessment points, suggesting that teens exhibited fewer problems with emotional awareness at this assessment point than other assessment points, presumably due to the effects of treatment, regardless of which treatment condition was received. The mean level of emotional awareness difficulties at 3 months in this study (16.95) is similar in magnitude to means observed in other studies of non-clinical adolescent samples (e.g., 15.5 [Weinberg & Konsky, 2009] and 16.84 [Broderick & Metz, 2009] and is lower in magnitude relative to another study of adolescents with internalizing problems from our group (19.01; [Wolff et al., 2018]).

The only significant cross-lagged effect between emotional awareness difficulties and cannabis use was observed from 3- to 6- months post-baseline. The 3-month assessment point roughly corresponded to the end of the course of intensive treatment, as most participants were enrolled in the home-based program for three months. It is possible that upon completion of treatment some participants continued to struggle with emotional awareness, putting them at risk to turn to cannabis use during the transition from intensive care to either reduced or no treatment. Given that the 3- to 6-month transition was the only time period to evidence this effect, these findings might highlight a particularly high risk window of time where youth turn to cannabis use as a coping strategy. This observed timeline is aligned with findings that most adolescents relapse to substance use within 6 months of the completion of treatment (e.g., Cornelius et al., 2003). This may especially be true when adolescents terminate an intensive therapy program based on insurance limitations, as occurred in this study, rather than based on symptom improvement. Therefore, identification of emotions as a precursor to the regulation of emotion may be an especially relevant treatment target among adolescents with comorbid psychiatric diagnoses and substance use problems, given that continued substance use may maintain psychiatric problems over time.

Findings for links between the regulation of sadness or depression and substance use are mixed (e.g., Feingold, Weiser, Rehm, & Lev-Ran, 2015), whereas more consistent evidence has been found for an association between the regulation of anger/aggressive behavior and substance use (Banderali et al., 2015; Swaim, Oetting, Edwards, & Beauvais, 1989). Further, there is evidence that using hard drugs (e.g., heroin/other narcotics, methamphetamines) is associated with awareness deficits surrounding anger but not sadness (Hessler & Katz, 2010). Because the DERS assesses difficulties with awareness across a range of emotions, it is not possible to identify whether lack of awareness of a specific emotion(s) drives the relationship with cannabis use.

Multiple factors interact to confer risk for problematic substance use in adolescents (Siegel, 2015). Studying the interaction of different components of emotion (e.g., lower- and higher-order constituent processes across neurobiological and behavioral domains; Thompson, Lewis & Calkins, 2008) may result in better prediction of substance use onset in adolescence. Future research should examine emotional awareness more thoroughly than simply via self-report (e.g., interview, performance tasks) and evaluate emotion-specific awareness to facilitate a more complete understanding of these processes. Greater understanding of these processes may enable more fine-grained understanding of emotional processes and ultimately, how they might best be addressed in the treatment of adolescents with co-occurring disorders. Indeed, there are several behavioral treatments for adult substance disorders that target emotional awareness and regulation (Roos & Witkiewitz, 2017), such as Mindfulness-based relapse prevention (Witkiewitz & Bowen, 2010) and Mindfulness-based Recovery Enhancement (Garland, 2013).

This study is, to our knowledge, the first to reveal a longitudinal association between emotional awareness difficulties and cannabis use in a clinical adolescent sample. Although they require replication, these findings may have implications for an increased focus on emotional awareness throughout treatment, but particularly prior to the discontinuation of treatment. In light of these contributions, the following limitations should be considered. First, although the DERS has acceptable psychometric properties, participants self-reported emotion awareness on a brief subscale and these results should be replicated. Second, participants retrospectively recalled past week cannabis use. Future research may consider evaluating the relationship between emotional awareness and substance use over a longer period, and in a more fine-grained fashion, e.g. via ecological momentary assessment. Third, we were not powered to explore how the relationship between cannabis use and emotional awareness varies across comorbid diagnoses, which is a critical next step. There were high rates of comorbid internalizing and externalizing (particularly ADHD) diagnoses in this sample. It is possible that the observed pattern could present differently among different combinations of comorbid diagnoses and combining the sample suppressed potential findings that might be evident within a larger sample. Future work should strive to tease apart differential patterns in the relationship between emotional awareness and cannabis use depending on psychopathology.

Additionally, the significant cross-lagged effect from emotional awareness difficulties at 3-months’ post baseline to increased cannabis use 6-months’ post baseline was not significant in the RI-CLPM, it became significant once we fixed the random intercept to zero. Several criticisms of the traditional CLPM surround the point that the model does not separate trait and state-level effects (Hamaker et al., 2015); however, when we constrained the random intercept of the RI-CLPM to zero, the model fit did not significantly worsen relative to the traditional CLPM. In the spirit of parsimony, and given that we observed a small decreased in the standard error surrounding the effect (i.e., a decrease from 0.06 in the RI-CLPM to 0.04 in the traditional CLPM), we proceeded model testing with the constrained model, which suggested that there was minimal trait level variability supported in this data. However, given the theoretical strengths of using the RI-CLPM, we see these findings as preliminary and we emphasize that findings from this study should be replicated in a larger, independent sample. Another limitation of this study is that we examined if emotional awareness was related to cannabis use, not negative consequences of cannabis use. Future research should also examine if emotional awareness is related to negative consequences of marijuana as well as frequency of use. Finally, the current study evaluated a sample of adolescents with co-occurring substance use and psychiatric symptoms. It is possible that a different pattern of findings may emerge in adolescents who abuse substances but do not meet criteria for co-occurring psychiatric disorders.

This study utilized a cross-lagged panel modeling technique to evaluate alternative directions of effects between cannabis use and emotional awareness difficulties over time in a sample of adolescents with comorbid psychiatric diagnoses and substance use problems. Overall, findings suggest that adolescents who had difficulty identifying emotions, presumably negative emotions, after completing treatment tended to use cannabis at increased rates three months later, potentially as a maladaptive coping strategy. Increasing awareness of emotions may be an important treatment strategy for adolescents who misuse cannabis. Identifying and employing alternative coping strategies, thereby buffering against continued high rates of cannabis use, may also have positive effects on psychiatric symptom severity in adolescents with co-occurring disorders.

Acknowledgements

This work was supported by the National Institutes of Health under grants R01AA020705 and T32DA016184. Sarah A. Thomas is partially supported by Institutional Development Award Number U54GM115677 from the National Institute of General Medical Sciences of the National Institutes of Health, which funds Advance Clinical and Translational Research (Advance-CTR). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Disclosure of Interest: The authors report no conflicts of interest. Grant numbers: R01AA020705, T32DA016184, U54GM115677.

Footnotes

Data Sharing Policy: The data that support the findings of this study are available from the corresponding author, [LM], upon reasonable request.

References

  1. Abrantes AM, Brown SA, & Tomlinson KL (2004). Psychiatric comorbidity among inpatient substance abusing adolescents. Journal of Child & Adolescent Substance Abuse, 13(2), 83–101. [Google Scholar]
  2. Aldao A, Nolen-Hoeksema S, & Schweizer S (2010). Emotion-regulation strategies across psychopathology: A meta-analytic review. Clinical Psychology Review, 30(2), 217–237. doi: 10.1016/j.cpr.2009.11.004 [DOI] [PubMed] [Google Scholar]
  3. Arnett AB, Pennington BF, Willcutt E, Dmitrieva J, Byrne B, Samuelsson S, & Olson RK (2012). A cross-lagged model of the development of ADHD inattention symptoms and rapid naming speed. J Abnorm Child Psychol, 40(8), 1313–1326. doi: 10.1007/s10802-012-9644-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Arseneault L, Cannon M, Poulton R, Murray R, Caspi A, & Moffitt TE (2002). Cannabis use in adolescence and risk for adult psychosis: Longitudinal prospective study. BMJ, 325(7374), 1212. doi: 10.1136/bmj.325.7374.1212 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Banderali G, Martelli A, Landi M, Moretti F, Betti F, Radaelli G,… Verduci E (2015). Short and long term health effects of parental tobacco smoking during pregnancy and lactation: A descriptive review. Journal of Translational Medicine, 13(1), 327. doi: 10.1186/s12967-015-0690-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Boden MT, & Berenbaum H (2011). What you are feeling and why: Two distinct types of emotional clarity. Personality and Individual Differences, 51(5), 652–656. doi: 10.1016/j.paid.2011.06.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bonn-Miller MO, Vujanovic AA, & Zvolensky MJ (2008). Emotional dysregulation: Association with coping-oriented marijuana use motives among current marijuana users. Substance Use & Misuse, 43(11), 1653–1665. doi: 10.1080/10826080802241292 [DOI] [PubMed] [Google Scholar]
  8. Brackett MA, & Mayer JD (2003). Convergent, discriminant, and incremental validity of competing measures of emotional intelligence. Personality and Social Psychology Bulletin, 29(9), 1147–1158. doi: 10.1177/0146167203254596 [DOI] [PubMed] [Google Scholar]
  9. Broderick PC, & Metz S (2009). Learning to BREATHE: A pilot trial of a mindfulness curriculum for adolescents. Advances in School Mental Health Promotion, 2(1), 35–46. doi: 10.1080/1754730X.2009.9715696 [DOI] [Google Scholar]
  10. Buckner JD, Keough ME, & Schmidt NB (2007). Problematic alcohol and cannabis use among young adults: The roles of depression and discomfort and distress tolerance. Addictive Behaviors, 32(9), 1957–1963. doi: 10.1016/j.addbeh.2006.12.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Carton S, Bayard S, Paget V, Jouanne C, Varescon I, Edel Y, & Detilleux M (2010). Emotional awareness in substance-dependent patients. Journal of Clinical Psychology, 66(6), 599–610. doi: 10.1002/jclp.20662 [DOI] [PubMed] [Google Scholar]
  12. Chadwick B, Miller ML, & Hurd YL (2013). Cannabis use during adolescent development: Susceptibility to psychiatric illness. Frontiers in Psychiatry, 4, 129. doi: 10.3389/fpsyt.2013.00129 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cornelius JR, Maisto SA, Pollock NK, Martin CS, Salloum IM, Lynch KG, & Clark DB (2003). Rapid relapse generally follows treatment for substance use disorders among adolescents. Addictive Behaviors, 28(2), 381–386. doi: 10.1016/s0306-4603(01)00247–7 [DOI] [PubMed] [Google Scholar]
  14. Deas D, & Brown ES (2006). Adolescent substance abuse and psychiatric comorbidities. Journal of Clinical Psychiatry, 67, 18–23. [PubMed] [Google Scholar]
  15. Deas-Nesmith D, Campbell S, & Brady KT (1998). Substance use disorders in an adolescent inpatient psychiatric population. Journal of the National Medical Association, 90(4), 233–238. [PMC free article] [PubMed] [Google Scholar]
  16. Dorard G, Berthoz S, Phan O, Corcos M, & Bungener C (2008). Affect dysregulation in cannabis abusers. European Child & Adolescent Psychiatry, 17(5), 274–282. doi: 10.1007/s00787-007-0663-7 [DOI] [PubMed] [Google Scholar]
  17. Dougherty DM, Mathias CW, Dawes MA, Furr RM, Charles NE, Liguori A,… Acheson, A. (2013). Impulsivity, attention, memory, and decision-making among adolescent marijuana users. Psychopharmacology, 226(2), 307–319. doi: 10.1007/s00213-012-2908-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Factor PI, Rosen PJ, & Reyes RA (2016). The relation of poor emotional awareness and externalizing behavior among children with ADHD. Journal of Attention Disorders, 20(2), 168–177. doi: 10.1177/1087054713494005 [DOI] [PubMed] [Google Scholar]
  19. Feingold D, Weiser M, Rehm J, & Lev-Ran S (2015). The association between cannabis use and mood disorders: A longitudinal study. Journal of Affective Disorders, 172, 211–218. doi: 10.1016/j.jad.2014.10.006 [DOI] [PubMed] [Google Scholar]
  20. Fontes MA, Bolla KI, Cunha PJ, Almeida PP, Jungerman F, Laranjeira RR,… & Lacerda AL (2011). Cannabis use before age 15 and subsequent executive functioning. The British Journal of Psychiatry, 198(6), 442–447. doi: 10.1192/bjp.bp.110.077479 [DOI] [PubMed] [Google Scholar]
  21. Garland EL (2013). Mindfulness-oriented recovery enhancement for addiction, stress, and pain. Washington, DC: NASW Press. [Google Scholar]
  22. Gottman JM, Katz LF, & Hooven C (1997). Meta-emotion: How families communicate emotionally. New York: Taylor & Francis. [Google Scholar]
  23. Gratz KL, & Roemer L (2004). Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the Difficulties in Emotion Regulation Scale. Journal of Psychopathology and Behavioral Assessment, 26(1), 41–54. doi: 10.1023/B:JOBA.0000007455.08539.94 [DOI] [Google Scholar]
  24. Green KM, Doherty EE, Stuart EA, & Ensminger ME (2010). Does heavy adolescent marijuana use lead to criminal involvement in adulthood? Evidence from a multiwave longitudinal study of urban African Americans. Drug and Alcohol Dependence, 112(1–2), 117–125. doi: 10.1016/j.drugalcdep.2010.05.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Gross JJ, & Jazaieri H (2014). Emotion, emotion regulation, and psychopathology: An affective science perspective. Clinical Psychological Science, 2(4), 387–401. doi: 10.1177/2167702614536164 [DOI] [Google Scholar]
  26. Hamaker EL, Kuiper RM, & Grasman RPPP (2015). A critique of the cross-lagged panel model. Psychological Methods, 20(1), 102–116. doi: 10.1037/a0038889 [DOI] [PubMed] [Google Scholar]
  27. Hessler DM, & Katz LF (2010). Brief report: Associations between emotional competence and adolescent risky behavior. Journal of Adolescence, 33(1), 241–246. doi: 10.1016/j.adolescence.2009.04.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hopfer C (2014). Implications of marijuana legalization for adolescent substance use. Substance abuse, 35(4), 331–335. doi: 10.1080/08897077.2014.943386 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kassel JD (2010). Substance abuse and emotion. Washington, DC: American Psychological Association. [Google Scholar]
  30. Kaufman J, Birmaher B, Brent D, Rao UMA, Flynn C, Moreci P,… & Ryan N (1997). Schedule for affective disorders and schizophrenia for school-age children-present and lifetime version (K-SADS-PL): Initial reliability and validity data. Journal of the American Academy of Child & Adolescent Psychiatry, 36(7), 980–988. [DOI] [PubMed] [Google Scholar]
  31. Limonero J, Tomás-Sábado J, & Fernández-Castro J (2006). Perceived emotional intelligence and its relation to tobacco and cannabis use among university students. Psicothema, 18, 95–100. [PubMed] [Google Scholar]
  32. McCaffrey DF, Liccardo Pacula R, Han B, & Ellickson P (2010). Marijuana use and high school dropout: The influence of unobservables. Health Economics, 19(11), 1281–1299. doi: 10.1002/hec.1561 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mitchell JT, Sweitzer MM, Tunno AM, Kollins SH, & McClernon FJ (2016). “I use weed for my ADHD”: A qualitative analysis of online forum discussions on cannabis use and ADHD. PloS one, 11(5), e0156614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Moitra E, Christopher PP, Anderson BJ, & Stein MD (2015). Coping-motivated marijuana use correlates with DSM-5 cannabis use disorder and psychological distress among emerging adults. Psychology of Addictive Behaviors, 29(3), 627–632. doi: 10.1037/adb0000083 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Muthén LK, & Muthén BO (1998–2012). MPlus User’s Guide (Eighth Edition.). Los Angeles, CA: Muthén & Muthén. [Google Scholar]
  36. National Institute on Drug Abuse (2003). Preventing Drug Use among Children and Adolescents. A research-based guide for parents, educators, and community leaders. Bethesda, Maryland: U.S. Department of Health and Human Services. [Google Scholar]
  37. Patrick ME, Evans-Polce RJ, Kloska DD, & Maggs JL (2019). Reasons high school students use marijuana: Prevalence and correlations with use across four decades. Journal of Studies on Alcohol and Drugs, 80(1), 15–25. doi: 10.15288/jsad.2019.80.15 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Peters T, & Foust C (2019). High school student cannabis use and perceptions towards cannabis in southcentral Colorado–comparing communities that permit recreational dispensaries and communities that do not. Journal of Cannabis Research, 1(1), 2. doi: 10.1186/s42238-019-0002-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Poon JA, Turpyn CC, Hansen A, Jacangelo J, & Chaplin TM (2016). Adolescent Substance Use & Psychopathology: Interactive Effects of Cortisol Reactivity and Emotion Regulation. Cognitive Therapy and Research, 40(3), 368–380. doi: 10.1007/s10608-015-9729-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Roos CR, & Witkiewitz K (2017). A contextual model of self-regulation change mechanisms among individuals with addictive disorders. Clinical Psychology Review, 57, 117–128. doi: 10.1016/j.cpr.2017.08.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Samur D, Tops M, Schlinkert C, Quirin M, Cuijpers P, & Koole SL (2013). Four decades of research on alexithymia: moving toward clinical applications. Frontiers in psychology, 4, 861 Doi: 10.3389/fpsyg.2013.00861 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Siegel JP (2015). Emotional regulation in adolescent substance use disorders: Rethinking risk. Journal of Child & Adolescent Substance Abuse, 24(2), 67–79. doi: 10.1080/1067828X.2012.761169 [DOI] [Google Scholar]
  43. Spirito A, Monti PM, Barnett NP, Colby SM, Sindelar H, Rohsenow DJ,… Myers M (2004). A randomized clinical trial of a brief motivational intervention for alcohol-positive adolescents treated in an emergency department. Journal of Pediatrics, 145(3), 396–402. doi: 10.1016/j.jpeds.2004.04.057 [DOI] [PubMed] [Google Scholar]
  44. Spirito A, Sindelar-Manning H, Colby SM, Barnett NP, Lewander W, Rohsenow DJ, & Monti PM (2011). Individual and family motivational interventions for alcohol-positive adolescents treated in an emergency department results of a randomized clinical trial. Archives of Pediatrics and Adolescent Medicine, 165(3), 269–274. doi: 10.1001/archpediatrics.2010.296 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Swaim RC, Oetting ER, Edwards RW, & Beauvais F (1989). Links from emotional distress to adolescent drug use: A path model. Journal of Consulting and Clinical Psychology, 57(2), 227–231. doi: 10.1037/0022-006X.57.2.227 [DOI] [PubMed] [Google Scholar]
  46. Thompson RA, Lewis MD, & Calkins SD (2008). Reassessing emotion regulation. Child Development Perspectives, 2(3), 124–131. doi: 10.1111/j.1750-8606.2008.00054.x [DOI] [Google Scholar]
  47. Thorberg FA, Young RM, Sullivan KA, & Lyvers M (2009). Alexithymia and alcohol use disorders: A critical review. Addictive Behaviors, 34(3), 237–245. doi: 10.1016/j.addbeh.2008.10.016 [DOI] [PubMed] [Google Scholar]
  48. Torrado MV, Ouakinin SS, & Bacelar-Nicolau L (2013). Alexithymia, emotional awareness and perceived dysfunctional parental behaviors in heroin dependents. International Journal of Mental Health and Addiction, 11(6), 703–718. doi: 10.1007/s11469-013-9448-z [DOI] [Google Scholar]
  49. Troisi A, Pasini A, Saracco M, Spalletta G (1998). Psychiatric symptoms in male cannabis users not using other illicit drugs. Addiction, 93, 487–492. [DOI] [PubMed] [Google Scholar]
  50. Weinberg A, & Klonsky ED (2009). Measurement of emotion dysregulation in adolescents. Psychological Assessment, 21(4), 616–621. doi: 10.1037/a0016669. [DOI] [PubMed] [Google Scholar]
  51. Zvolensky MJ, Bernstein A, Marshall EC, & Feldner MT (2006). Panic attacks, panic disorder, and agoraphobia: Associations with substance use, abuse, and dependence. Current Psychiatry Reports, 8(4), 279–285. doi: 10.1007/s11920-006-0063-6 [DOI] [PubMed] [Google Scholar]

RESOURCES