Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Aug 1.
Published in final edited form as: Exp Clin Psychopharmacol. 2020 Mar 5;28(4):426–437. doi: 10.1037/pha0000357

Longitudinal associations between negative urgency, symptoms of depression, cannabis and alcohol use in veterans

Rachel L Gunn 1, Angela K Stevens 1, Lauren Micalizzi 2,1, Kristina M Jackson 1, Brian Borsari 3,4, Jane Metrik 1,5
PMCID: PMC7390669  NIHMSID: NIHMS1566062  PMID: 32134284

Abstract

There is a high comorbidity between symptoms of depression and cannabis and alcohol use in civilian and veteran populations. Prospective studies attempting to clarify the directionality of these comorbidities have yielded mixed results. Further, the relations between these constructs and impulsive personality, particularly negative urgency (NU, the tendency to act rashly when experiencing emotional distress) warrants further attention, as NU relates to symptoms of depression and alcohol and cannabis use. Importantly, NU partially accounts for the association between symptoms of depression and cannabis and alcohol problems in cross-sectional studies. This study examined alternative theories of directionality in order to better understand the longitudinal associations between symptoms of depression, NU, and cannabis or alcohol use. Three semi-annual waves of data (baseline, 6-month, and 12-month) were collected in parallel assessments from a sample of OEF/OIF/OND veterans (N=361). Autoregressive cross-lagged panel models were used to test four alternative theory-driven models about the longitudinal associations between the interaction of symptoms of depression and NU and cannabis or alcohol use. Models revealed unique direction of effects specific to each substance, such that the interaction between symptoms of depression and NU at 6-months post-baseline predicted more alcohol use at 12-months post-baseline, whereas more cannabis use at 6-months post-baseline predicted more severe symptoms of depression at 12-months post-baseline. Results suggests alternate directions of effect for cannabis and alcohol use. Future research should examine these patterns over wider assessment periods in order to see more variability and change over time.

Keywords: alcohol, cannabis, depression, negative urgency, veterans

Introduction

Aside from tobacco, cannabis and alcohol are the most commonly used substances nationwide (Substance Abuse and Mental Health Services Administration, 2019). Both alcohol and cannabis are often used to cope with the negative affect experienced with depression (Holahan, Holahan, Moos, Cronkite, & Randall, 2003; Simons, Gaher, Correia, Hansen, & Christopher, 2005). However, the use of substances such as cannabis and alcohol also predicts increased symptoms of depression and poorer clinical and health care outcomes (Lev-Ran et al., 2014; Sullivan, Fiellin, & O’Connor, 2005; Sullivan, Goulet, Justice, & Fiellin, 2011). These independent directional findings suggest that these constructs may be reciprocally related over time (i.e., bi-directionally associated). Consistent with this bi-directional association, individuals with cannabis and alcohol use disorders (CUD/AUD) have higher rates of major depressive disorder (MDD), relative to those without CUD or AUD (Chen, Wagner, & Anthony, 2002; Grant et al., 2016; Metrik et al., 2016). Rates of comorbidity between substance use disorders and mood disorders are also common among military veterans (Seal et al., 2009; Teeters, Lancaster, Brown, & Back, 2017). By testing parallel models of cannabis and alcohol use, we were able to examine whether or not similar patterns of associations with symptoms of depression exist across cannabis use and alcohol use.

Although little research has compared the association between depression symptoms and cannabis versus alcohol use within the same sample, extant cross-sectional and prospective studies have independently suggested use of both substances is strongly linked with depression symptoms. Cross-sectional research has shown that the link between symptoms of depression and cannabis use or CUD is robust both in the general population (Chen et al., 2002; Degenhardt, Hall, & Lynskey, 2003; Dierker, Selya, Lanza, Li, & Rose, 2018; Farris, Metrik, Bonn-Miller, Kahler, & Zvolensky, 2016; Feingold, Fox, Rehm, & Lev-Ran, 2015; Grant, 2004) and among veteran samples (Goldman et al., 2010; Metrik et al., 2016). A similar pattern has emerged for alcohol use. Numerous cross-sectional studies have found a link between AUD or alcohol use and symptoms of depression in the general population (e.g. Buckner, Keough, & Schmidt, 2007; Grant et al., 2015) and among veterans (Brooks Holliday, Pedersen, & Leventhal, 2016; Seal et al., 2011; Yoon, Petrakis, & Rosenheck, 2015).

The affective-motivational theory sheds light on the comorbidity among substance use and negative emotionality, positing that substances are used to regulate negative emotions. This theory underscores the central role of negative affect in motivating substance use (Baker, Piper, McCarthy, Majeskie, & Fiore, 2004; Cox & Klinger, 1988; Simons et al., 2005). Consistent with this theory, both cross-sectional (Bonn-Miller, Boden, Bucossi, & Babson, 2014; Brooks Holliday et al., 2016; Gonzalez, Reynolds, & Skewes, 2011; Metrik et al., 2016) and prospective (Swendsen et al., 2000) data suggest that greater intensity of negative affect leads to increased cannabis use or alcohol use as a short-term means of coping with negative emotions. Importantly, this pattern of negative reinforcement from substances becomes increasingly salient with greater progression to alcohol or drug dependence when long-term neuroadaptation occurs in the withdrawal/negative affect stage of the addiction cycle (Koob & Volkow, 2010). Thus, coping-oriented use of substances can lead to increased depression symptoms over time (Martens et al., 2008; Swendsen & Merikangas, 2000), suggesting a bi-directional association among the constructs.

However, longitudinal studies of directions of effects specific to depression symptoms and MDD and cannabis use and CUD yield mixed findings. In both adolescent girls (Patton et al., 2002) and adults (Bovasso, 2001), increased cannabis use predicted increased symptoms of depression, but increased depression symptoms did not predict increased use of cannabis. When cannabis use and symptoms of depression were evaluated simultaneously to examine bi-directional effects from adolescence to adulthood, reciprocal associations were found, suggesting that they mutually exacerbate each other over time (Womack, Shaw, Weaver, & Forbes, 2016). In one systematic review, Lev-Ran and colleagues (2014) found that, controlling for baseline depression, cannabis use (and especially heavy use) was associated with increased risk for developing MDD. However, large epidemiological studies have found stronger evidence that MDD predicts CUD, rather than the reverse (Agosti, Nunes, & Levin, 2002; Blanco et al., 2016; Conway, Compton, Stinson, & Grant, 2006).

The prospective literature on alcohol use and depression yields a similarly mixed set of findings. Studies have found evidence that MDD precedes AUD (Kuo, Gardner, Kendler, & Prescott, 2006; McCarty et al., 2009), that AUD precedes MDD (Fergusson, Boden, & Horwood, 2009; Rohde, Lewinsohn, Kahler, Seeley, & Brown, 2001), as well as bi-directional associations (Bulloch, Lavorato, Williams, & Patten, 2012; Gilman & Abraham, 2001). Further, one large prospective study found that within the transition to adulthood (compared to other developmental transitions), MDD was more likely to precede AUD than the converse (Brière, Rohde, Seeley, Klein, & Lewinsohn, 2014). Similarly, a prospective study of individuals transitioning from adolescence to young adulthood found that symptoms of depression predicted persistence of alcohol problems (Copeland et al., 2012). Finally, veterans with depression pre-deployment were at increased risk of alcohol-related problems post-deployment (Jacobson et al., 2008). These mixed findings signal the associations between depression symptoms and cannabis or alcohol use may differ at the trait level, such as individual responses to the experience of negative affect.

Impulsive personality is trait that contributes significantly to the etiology of alcohol and cannabis use (Coskunpinar, Dir, & Cyders, 2013; Dawe & Loxton, 2004; De Wit, 2009; Gunn, Finn, Endres, Gerst, & Spinola, 2013; Settles et al., 2012; Verdejo-GarcƔá, Lawrence, & Clark, 2008). Impulsivity is a multi-faceted trait, for which multiple models and conceptualizations exist. All facets of impulsivity have been linked to both alcohol and cannabis use in various populations (Coskunpinar et al., 2013; De Wit, 2009; Stamates & Lau-Barraco, 2017). One model of impulsivity classifies impulsivity into five unique domains: negative urgency (NU), positive urgency, sensation seeking, lack of perseverance, and lack of premeditation (Lynam, Smith, Whiteside, & Cyders, 2006; Smith et al., 2007; Whiteside & Lynam, 2001). When examined together, each of these constructs are related to specific risky behaviors (Adams, Kaiser, Lynam, Charnigo, & Milich, 2012; Birthrong & Latzman, 2014; Cyders, Flory, Rainer, & Smith, 2009; Derefinko, Dewall, Metze, Walsh, & Lynam, 2011; Smith et al., 2007). Specifically, trait negative urgency (NU), which is characterized as the tendency to engage in rash action while experiencing emotional distress (Whiteside & Lynam, 2001), may be particularly relevant to the comorbidity between symptoms of depression and cannabis or alcohol use (Dvorak & Day, 2014; Kaiser, Milich, Lynam, & Charnigo, 2012; Simons et al., 2005; Smith & Cyders, 2016).

In the context of the affective-motivational theory (Cox & Klinger, 1988; Simons & Carey, 2006), those who are high in trait NU and who also experience increased negative affect associated with symptoms of depression may be more likely to drink or use cannabis relative to individuals with low NU traits. This may be due to the reduced ability to inhibit substance use when experiencing negative affect. In fact, the link between symptoms of depression and increased NU has been observed in cross-sectional studies of undergraduate college students (Anestis, Selby, & Joiner, 2007), community samples (Novak, Novak, Lynam, & Foti, 2016), and this sample of veterans (Gunn et al., 2018). In line with this theory, NU partially accounted for high rates of comorbid depression and alcohol problems in a cross-sectional study of college students (Gonzalez et al., 2011). Another cross-sectional study found that higher levels of NU partially accounted for the positive association between symptoms of depression and alcohol use initiation in adolescents (Pang, Farrahi, Glazier, Sussman, & Leventhal, 2014). However, no studies to date have tested competing theories of directionality between symptoms of depression and cannabis or alcohol use while also considering individual differences in NU as a moderator of these associations longitudinally. This is particularly relevant among veterans, who experience higher levels of negative affect and symptoms of depression, compared to civilian samples (Kessler, Chiu, Demler, & Walters, 2005; Kessler et al., 2014). It may be that NU is an additional critical factor that could help to explain mixed findings regarding the directionality between depression and cannabis or alcohol use.

The Current Study

The purpose of this longitudinal study is to examine temporal relationships by probing alternative directions of effects between symptoms of depression and substance use (cannabis and alcohol use), while considering whether NU interacts with symptoms of depression. We utilized three semi-annual waves of observational data from a sample of veterans deployed post-9/11/2001. Our primary aim was to determine the direction of effects among depression symptoms and cannabis use and alcohol use (independently) over time by comparing the fit of alternative models. Additionally, based on prior research and theory (Anestis et al., 2007; Gonzalez et al., 2011; Guillot, Pang, & Leventhal, 2014; Gunn, Jackson, Borsari, & Metrik, 2018; Smith & Cyders, 2016), we hypothesized that higher trait NU paired with more severe symptoms of depression would lead to more frequent cannabis and alcohol use at later assessment points.

Methods

Sample and Procedure

Participants (N=361) completed a prospective study examining cannabis use and affective disorders in returning Operation Enduring Freedom, Operation Iraqi Freedom, and Operation New Dawn (OEF/OIF/OND) veterans who were deployed post 9/11/2001 and who used cannabis at least once in their lifetime. Participants were recruited from a Veteran’s Health Administration (VHA) facility in the Northeast region of the United States by utilizing the VHA OEF/OIF/OND Roster, an accruing database of combat veterans who recently returned from military service in Iraq and Afghanistan and were enrolled in VHA (see Metrik et al., 2016, for details of recruitment procedures). Veterans were screened for eligibility by telephone and were invited for a baseline visit, at which time they signed informed consent and completed a battery of interview and self-report assessments. The baseline visit was followed by two additional visits with parallel assessments at 6-months (N=312; 86.4% retention) and 12 months (N=310; 85.9% retention). The study was approved by the university and local VHA Institutional Review Boards. Participants were compensated $50 per visit and an additional $50 bonus payment for completing all three visits.

Measures

Time-Line Follow-Back

The Time-Line Follow-Back Interview (TLFB; (Dennis, Funk, Harrington Godley, Godley, & Waldron, 2004; Sobell & Sobell, 1992)) was used to assess past-six-month patterns of cannabis and alcohol use. The percent of cannabis and alcohol use days in the 30 days prior to each assessment was used in the present analyses to ensure that the variable captured cannabis and alcohol use across a time period that more closely matched our assessment of depression symptoms.

UPPS-P Impulsive Behavior Scale

Negative urgency was assessed using the Short UPPS-P Impulsive Behavior Scale (Cyders, Littlefield, Coffey, & Karyadi, 2014). The UPPS-P is a 20-item self-report inventory which uses a 4-point Likert scale to assesses five subscales of impulsive personality (negative urgency, positive urgency, sensation seeking, lack of premeditation, and lack of perseverance), each demonstrating high levels of internal consistency in previous studies (Cyders et al., 2014). The NU subscale demonstrated good internal consistency (baseline α = .77, 6-months α = .77, 12-months α = .80).

Depression Symptoms

Depressive symptoms over the past week were assessed at each study wave via the Center for Epidemiological Studies-Depression Scale (CES-D; Radloff, 1997), which is a 20-item self-report inventory using a 4-point Likert scale. The CES-D demonstrated good internal consistency (baseline α = .87, 6-months α = .84, 12-months α = .86).

Data Analytic Strategy

All analyses were conducted in Mplus version 8.2 using maximum likelihood estimation for missing data (Muthén & Muthén, 1998–2018). Missing data were minimal and largely due to attrition in the study, with 49 (14%) and 51(14%) data points missing at 6- and 12-months, respectively, for CES-D and 47 (13%) and 49 (14%) data points at 6- and 12-months for cannabis and alcohol use. The present study utilized cross-lagged panel modeling (CLPM) which is well-suited for testing prospective associations with regard to directionality. We estimated four alternative models that systematically evaluate directional effects to inform hypotheses about the prospective association between the interaction of depression symptoms and NU (i.e., depression X NU) with cannabis and alcohol use, separately, over time (see Figure 1; Arnett et al., 2012). For example, in cannabis use models these prospective associations may include: (1) bi-directional effects of depression symptoms X NU on cannabis use and cannabis use on depression symptoms; (2) cannabis-driven effects, such that cannabis use results in more depression symptoms over time (cannabis use ➔ depression symptoms) but depression X NU does not predict later cannabis use; (3) depression X NU driven effects, such that the interaction between depression symptoms and NU may result in more cannabis use over time (depression X NU ➔ cannabis use) but cannabis use does not predict depression symptoms; or (4) no cross-lagged effects between depression X NU and cannabis use. All models examined time over and above autoregressive (i.e., stability) parameters. Parallel models were run for alcohol use. As outlined in the introduction, a theory-driven approach was taken to only examine the interaction between depression symptoms and NU on cannabis or alcohol use, as opposed to also examining the interaction between cannabis or alcohol use and NU on depression symptoms.

Figure 1. Analytic approach of alternative autoregressive cross-lagged panel path models.

Figure 1.

Open in a new tab

Autoregressive cross-lagged panel path model depicting associations between cannabis use, alcohol use, symptoms of depression, and the interaction with NU across three time points. Curved, double-headed arrows represent correlations among the variables, diagonal arrows represent cross-lagged paths, straight arrows represent autoregressive paths. NU box represents path with interaction of depression symptoms and NU. (A) Bidirectional effects of depression symptoms X NU on cannabis/alcohol use and cannabis/alcohol use on depression symptoms; (B) Cannabis/alcohol-driven effects, such that cannabis/alcohol use results in more depression symptoms over time (d paths dropped); (C) Depression symptoms X NU driven effects, such that the interaction between depression symptoms and NU may result in more cannabis/alcohol use over time (c paths dropped); or (D) No cross-lagged effects between depression symptoms X NU and cannabis/alcohol use over time (c and d paths dropped).

We first tested a saturated bi-directional model where all autoregressive and cross-lagged parameters were freely estimated for depression X NU and cannabis or alcohol use (Model 1, reference model). In Models 2–4, we systematically constrained cross-lagged parameters to zero (i.e., c and/or d cross-lagged parameters; see Figure 1) and compared these constrained models to the unconstrained model (Model 1) using a chi-square difference test. A statistically nonsignificant change in chi-square (p<.05) between the unconstrained (Model 1) and constrained models (Models 2–4) indicates the paths constrained to zero can be removed from the model without a significant decrement in model fit. For models where the depression-NU interaction was significant, we proceeded with that model. When this interaction was not statistically significant, we then tested a model with only main effects (e.g., depression ➔ cannabis use vs. cannabis use ➔ depression). The most parsimonious model that did not result in a significant decrement in model fit when compared to Model 1 was selected.

Model fit was evaluated using chi-square (χ2), comparative fit index (CFI), and root mean square error of approximation (RMSEA), with CFI values above .90 indicating good model fit (Hu & Bentler, 1999) and with RMSEA values of .06 (or below) suggesting a closer fit to the data and values of .08 and .10 values representing fair and marginal fit, respectively (Browne & Cudeck, 1993). Akaike’s Information Criteria (AIC) and Bayesian Information Criterion (BIC) were also inspected to guide model selection, where lower values indicate closer fit to the data (Vrieze, 2012). All variables were centered and all models adjusted for age and percent cannabis use days (for alcohol models) and percent alcohol use days (for cannabis models) at each assessment point. Race, ethnicity, sex, and marital status were also considered as covariates but these were not statistically significant and were subsequently removed from all analyses.

Results

Demographic and descriptive statistics are presented in Table 1.

Table 1.

Sample Demographics and descriptive statistics

Variable n %
Gendera
 Male 337 93
Racea
 White 289 80
 Black/African American 16 4
 Asian 6 2
 Native Hawaiian/Pacific Islander 2 .8
 American Indian/Alaska Native 2 2.8
 Multiracial 17 4
 Other 29 8
Ethnicitya
 Hispanic/Latino(a) 43 27
Marital Statusa
 Single/Never married 116 32
 Married/Living with partner 173 48
 Divorced/Separated 72 20
Employment Statusb
 Employed 283 78
 Unemployed/Homemaker 67 19
 Student 87 24
 Military Service 101 28
Combat Operation (s) served inab
 Operation Enduring Freedom (OEF) 269 75
 Operation Iraqi Freedom (OIF) 191 53
 Operation New Dawn (OND) 71 20
Most recent branch of service
 Army 249 69
 Marines 30 8
 Air Force 32 9
 Navy 46 13
 Coast Guard/Other 4 1
Current Cannabis Use Disorder (Bl) 53 15
Current Alcohol Use Disorder (Bl) 113 31
Current Major Depressive Disorder (Bl) 53 15
Any alcohol usec 340 94
Any cannabis usec 163 45
M SD
Agea 33.56 9.44
Years since last deployment 3.93 2.80
Number of deployments 1.87 1.15
Bl CES-D .72 .73
6mo CES-D .65 .65
12mo CES-D .60 .64
Bl NU 2.16 .72
6mo NU 2.12 .71
Bl % Alcohol use days 25.90 29.87
6mo % Alcohol use days 27.82 31.39
12mo % Alcohol use days 26.46 31.45
Bl % Cannabis use days 16.81 34.93
6mo % Cannabis use days 16.49 33.61
12mo % Cannabis use days 16.31 33.72
Open in a new tab

Note. N=361, BL=baseline assessment, 6mo= 6-month follow-up; 12mo=12-month follow-up assessment, NU = Negative urgency, CES-D = symptoms of depression from the Center for Epidemiological Studies-Depression Scale

a

Baseline assessment reported

b

Multiple responses permitted

c

Defined as reporting use on at least one day on the Timeline Follow-back across all three time points.

Cannabis Model 1: Bi-directional model

Fit statistics for all models are presented in Table 2. Model 1 examined the most-saturated model by estimating the bi-directional associations between depression symptoms X NU on cannabis use and cannabis use on depression symptoms (see Figure 1) across three time points. Less than adequate model fit was obtained (CFI=.897; RMSEA=.115 [90% CI=.097-.135]; χ2 =138.25(28), p<.001). Regarding cross-lagged paths, more cannabis use at 6-months post baseline significantly predicted increased depressive symptoms at 12-months post baseline (β=.11; 95% CI=.01, .21). However, there were no significant cross-lagged paths for depression X NU on later cannabis use at either time point. We subsequently examined a model without the interaction between depression and NU. Model fit indices were suboptimal (CFI=.85; RMSEA=.124 [90% CI=.11, .14]; χ2 =207.73(37), p<.001). This model served as the reference model for all nested comparisons; all model fit statistics are presented in Table 2, panel A. All autoregressive paths for cannabis use and depression symptoms were positive and significant. The cross-lagged path from cannabis use at 6-months to depressive symptoms at 12-months remained significant (β=.11; 95% CI=.01, .21), whereas the reverse effect was not statistically significant (see Table 3).

Table 2.

Model fit statistics for four alternative models

Model χ2 df CFI RMSEA BIC AIC Δχ2 Δdf p
A. Cannabis Use Models
1 207.73 37 .850 .124 10829.58 10685.39
2 209.75 39 .850 .121 10820.21 10683.42 2.03 2 .36
3 214.35 39 .846 .123 10824.80 10688.01 6.62 2 .04
4 216.44 41 .845 .120 10815.51 10686.12 8.72 4 .07
B. Alcohol Use Models
1 154.51 36 .877 .105 9794.56 9650.63
2 169.18 42 .868 .101 9775.08 9653.30 14.67 6 .03
3 155.75 38 .878 .102 9784.41 9647.87 1.23 2 .54
4 170.41 44 .869 .099 9764.93 9650.53 15.90 8 .04
Open in a new tab

Note. CFI = comparative fit index; RMSEA = root mean square error of approximation; BIC = Bayesian Information Criteria; AIC = Akaike’s Information Criteria. Model 1(comparison for all nested models): Bidirectional associations (i.e., depression X NU ↔ cannabis/alcohol use); Model 2: cannabis/alcohol use to depression symptoms; Model 3: depression symptoms x NU to cannabis/alcohol use; Model 4: no reciprocal transactions modeled. Model comparisons reflect a chi-square difference test comparing Models 2–4 to Model 1 for both outcomes. A p-value <.05 reflects a statistically significant difference in model fit.

Table 3.

Standardized parameter estimates (standard errors) from the traditional and nested cross-lagged panel models examining cannabis use

Model 1 Model 2 Model 3 Model 4
Autoregressive
Cannabis Use
 Bl → 6mo 0.75 (0.03)*** 0.77 (0.03)*** 0.75 (0.03)*** 0.77 (0.03)***
 6mo → 12mo 0.82 (0.02)*** 0.82 (0.02)*** 0.82 (0.02)*** 0.82 (0.02)***
Depression Symptoms
 Bl → 6mo 0.64 (0.04)*** 0.64 (0.04)*** 0.66 (0.03)*** 0.65 (0.03)***
 6mo → 12mo 0.60 (0.04)*** 0.60 (0.04)*** 0.63 (0.04)*** 0.63 (0.04)***
Cross-lagged
Cannabis Use → Depression Symptoms
 Bl → 6mo 0.04 (0.05) 0.04 (0.05) — —
 6mo → 12mo 0.11 (0.05)** 0.11 (0.05)** — —
Depression Symptoms → Cannabis Use
 Bl → 6mo 0.06 (0.04) — 0.06 (0.04) —
 6mo → 12mo 0.01 (0.03) — 0.01 (0.03) —
Covariance
Cannabis Use/ Depression Symptoms
 Bl 0.30 (0.05)*** 0.30 (0.05)*** 0.30 (0.05)*** 0.30 (0.05)***
 6mo 0.10 (0.06) 0.09 (0.06) 0.09 (0.06) 0.09 (0.06)
 12mo 0.03 (0.06) 0.03 (0.06) 0.03 (0.06) 0.03 (0.06)
Open in a new tab

Note. Model 1: Bidirectional associations from depression symptoms to cannabis use vs. cannabis use to depression symptoms; Model 2: cannabis use to depression symptoms; Model 3: depression symptoms to cannabis use; Model 4: no cross-lagged parameters modeled. Assessments: Bl=baseline; 6mo=6 months post-baseline; 12mo=12 months post-baseline.

Cannabis Model 2–4: Test of direction of effects

Next, we compared Models 2–4 to Model 1. Models 2, 3, and 4 also revealed marginally acceptable fit. Whereas there was a significant chi-square difference test for Model 3 versus Model 1 (Table 2, panel A), the chi-square difference test for Model 2 (i.e., cannabis-driven effects) versus Model 1 (Δχ2(2) =2.03, p=.36) and Model 4 (i.e., no cross-lagged effects) versus Model 1 (Δχ2(4) =8.72, p=.07) were not significant, indicating that constraining these parameters did not result in significantly worse model fit. To examine comparative model fit between these two models (2 and 4), we conducted another chi-square difference test (i.e., Model 2 versus Model 4 because Model 4 is nested under Model 2) and found that Model 4 differed significantly from Model 2 (Δχ2(2) =6.70, p=.04), indicating that Model 2 retained the best fit to the data; thus, it was selected. Results from this model again revealed the cross-lagged effect from cannabis use at 6-months post baseline to depression symptoms at 12-months post baseline (β=.11; 95% CI=.01, .21; see Figure 2, Panel A), but not baseline cannabis use to depression symptoms at 6-months post baseline. Therefore, evidence from the comparative analytic models supports the prospective association between more frequent cannabis use at 6-months post baseline and subsequent increase in severity of depression symptoms 12-months post baseline.

Figure 2. Results from the best fitting cross-lagged panel model for cannabis use (panel A) and alcohol use (panel B).

Figure 2.

Open in a new tab

***p<.001; **p<.01. Cannabis use =past month % cannabis use days from TLFB; Alcohol use= past month % alcohol use days from TLFB; Dep sxs=CES-D depression symptoms. Curved, double-headed arrows represent correlations among the variables that they connect. Dashed paths were not significant. All variables centered.

Alcohol Model 1: Bi-directional model

Model fit statistics are presented in Table 1, panel B. Good model fit was obtained in Model 1 (CFI=.877; RMSEA=.105 [90% CI=.09-.125]; χ2 =154.51(36), p=.00). Autoregressive and cross-lagged path standardized estimates are reported in Table 4. All autoregressive paths for alcohol use and depression symptoms were positive and significant. Regarding cross-lagged paths, the interaction between depression symptoms and NU at 6-months significantly predicted alcohol use at 12-months (β=.12; 95% CI=.04, .19; see Figure 1) but not between baseline and 6-months post baseline. There were no significant cross-lagged paths for alcohol use to later depression symptoms at either time point. The covariances (baseline) and residual covariances between alcohol use and depression symptoms were not significant at any time point.

Table 4.

Standardized parameter estimates (standard errors) from the traditional and nested cross-lagged panel models examining alcohol use

Model 1 Model 2 Model 3 Model 4
Autoregressive
Alcohol Use
 Bl → 6mo 0.72 (0.03)*** 0.72 (0.03)*** 0.72 (0.03)*** 0.72 (0.03)***
 6mo → 12mo 0.75 (0.03)*** 0.75 (0.03)*** 0.75 (0.03)*** 0.75 (0.03)***
Depression Symptoms
 Bl → 6mo 0.64 (0.04)*** 0.64 (0.04)*** 0.64 (0.04)*** 0.64 (0.04)***
 6mo → 12mo 0.61 (0.04)*** 0.61 (0.04)*** 0.61 (0.04)*** 0.61 (0.04)***
Cross-lagged
Alcohol Use → Depression Symptoms
 Bl → 6mo 0.05 (0.05) 0.05 (0.05) — —
 6mo → 12mo −0.02 (0.05) −0.02 (0.05) — —
Depression symptoms (Dep) X NU → Alcohol Use
 Bl Dep x NU → 6mo 0.01 (0.04) — 0.01 (0.04) —
 Bl Dep −0.00 (0.04) — −0.00 (0.04) —
 Bl NU 0.03 (0.04) — 0.03 (0.04) —
 6mo Dep x NU → 12mo 0.11 (0.04)* — 0.11 (0.04)* —
 6mo Dep 0.07 (0.04) — 0.07 (0.04) —
 6mo NU −0.07 (0.04) — −0.07 (0.04) —
Residual Covariance
Alcohol Use/ Depression Symptoms
 Bl 0.02 (0.06) 0.02 (0.06) 0.02 (0.06) 0.02 (0.06)
 6mo 0.05 (0.06) 0.05 (0.06) 0.05 (0.06) 0.05 (0.06)
 12mo 0.09 (0.06) 0.07 (0.06) 0.09 (0.06) 0.07 (0.06)
Open in a new tab

Note. Model 1: Bidirectional from depression symptoms X NU to alcohol use vs. alcohol use to depression symptoms; Model 2: alcohol use to depression symptoms; Model 3: depression symptoms x NU to alcohol use; Model 4: no cross-lagged parameters modeled. Assessments: Bl=baseline; 6mo=6 months post-baseline; 12mo=12 months post-baseline.

*

p < .05

**

p < .01

***

p < .001.

Alcohol Models 2–4: Test of direction of effects

Models 2, 3, and 4 also fit the data well. Comparisons between these models and Model 1 showed that the chi-square difference test for comparing Models 2 and 4 were significant (Table 2, panel B), suggesting worse fit for these models relative to Model 1. The chi-square difference test for Model 3 versus Model 1 (Δχ2(2) =1.23, p=.54) was not significant, indicating that constraining cross-lagged parameters from symptoms of depression to cannabis and alcohol use to zero did not result in significantly worse model fit; thus, Model 3 was selected. Results revealed a significant cross-lagged effect from the interaction of depression symptoms and NU at 6-months predicting alcohol use at 12-months (β=.11; 95% CI=.04, .19; see Figure 3). Simple slope analyses revealed unstandardized estimates of −3.75 (p= .30) those 1 SD below the mean of NU, 3.42 (p= .09) at the mean, and 10.59 (p<.001) for 1 SD above the mean. In other words, veterans with higher symptoms depression and higher NU at 6-months reported more alcohol use one year later. This path was not significant from baseline depression symptoms X NU to alcohol use at 6-months. Therefore, results from the comparative analytic models support prospective associations between greater severity of depression symptoms in combination with higher levels of NU 6-months post baseline with subsequent alcohol use 12-months post baseline.

Discussion

This study examined directions of effects between symptoms of depression alone and in conjunction with NU and cannabis or alcohol use in a sample of OEF/OIF/OND veterans. Directions of effects was tested rigorously and systematically using a series of nested autoregressive cross-lagged panel models, adjusting for age and cannabis or alcohol use (in respective models) at each time point. There was an interaction between symptoms of depression and NU at 6-months post baseline that predicted more frequent alcohol use 12-months post baseline, whereas cannabis use at 6-months post baseline predicted greater severity of depression symptoms 12-months post baseline. All other cross-lagged paths (from baseline to 6-months post baseline) were non-significant. As expected, autoregressive paths for symptoms of depression, alcohol use, and cannabis use were all highly significant and positive, suggesting stability over time.

As hypothesized, and consistent with prior studies of alcohol problems (Anestis et al., 2007; Gonzalez et al., 2011; Pang et al., 2014), there was a significant interaction between symptoms of depression and NU predicting more frequent alcohol use. This is the first study to find an association between depression symptoms and NU in the prediction of alcohol use, and the first study to examine these processes in veterans. These findings are consistent with an affective-motivational model which posits that symptoms of depression predict more frequent alcohol use. Further, the association between depression symptoms and alcohol use may be exacerbated by high levels of NU. In other words, veterans with difficulty inhibiting appetitive behaviors who experience more severe symptoms of depression may be even more likely to drink in order to regulate frequent negative affect, leading to more frequent alcohol consumption. Consistent with other studies (e.g. Brière et al., 2014; Copeland et al., 2012), results did not suggest a bi-directional association (from alcohol use to increased symptoms of depression) in this sample.

However, it is important to note that some studies have found bi-directional associations (Bulloch et al., 2012). Heterogeneity in populations, controls, measures, and severity of depression symptoms may explain these associations. For example, this population of veterans who were predominately male may yield a different set of results compared to a population with a greater proportion of females (see Limitations). Taken together with this prior literature, results suggest that the longitudinal association between depression symptoms (especially when coupled with high levels of NU) and increased alcohol use is strong, but in some samples, alcohol use may also relate to increased symptoms of depression over time.

In contrast to the alcohol use findings, more cannabis use at 6-months post baseline predicted more severe symptoms of depression 12-months post baseline. These results are consistent with other studies suggesting cannabis use may be a risk factor for increased symptoms of depression (Bovasso, 2001; Degenhardt et al., 2003; Lev-Ran et al., 2014; Patton et al., 2002). Although it was hypothesized that the interaction between depression and NU would predict cannabis use, it may be that this moderator plays a greater role in accounting for the association with affect-related alcohol use, rather than cannabis use, over time. Consistent with this, a study of young adults found that anxiety symptoms, but not depression symptoms, were associated with frequency of cannabis use (Johnson, Bonn-Miller, Leyro, & Zvolensky, 2009), suggesting cannabis may not be used as frequently to manage depressive symptoms. Rather, use of cannabis may exacerbate symptoms of depression. This possible link between cannabis use and increased symptoms of depression has important implications for the use of cannabis to self-medicate mental health symptoms, particularly as use of medicinal cannabis is increasing in the general population and in veteran samples (Metrik, Bassett, Aston, Jackson, & Borsari, 2018) who also perceive cannabis as a low-risk substance (Wilkinson, van Schalkwyk, Davidson, & D’Souza, 2016).

Additionally, cross-sectional analysis from our group in this same sample found that NU partially accounted for the association between MDD and cannabis problems, but not cannabis use (Gunn et al., 2018). It is possible that the effect of NU is more consequential when the relationship between depression and cannabis involvement is amplified to higher levels of impairment—that of a major psychiatric disorder in relation to clear negative consequences from cannabis use. Although this and prior studies suggest that cannabis use may lead to later increased symptoms of depression, there is also robust evidence for the opposite direction of effects (Agosti et al., 2002; Conway et al., 2006) as well as absence of a longitudinal relationship between cannabis use and depression (Blanco et al., 2016). Furthermore, there is evidence of shared genetic risk for depression and cannabis use, suggesting that there may be an underlying propensity to experience both (Reviewed in Hasin, 2018). Future research should examine within- and between-person variability over time between symptoms of depression and cannabis use while also considering the role of NU.

Contrasting patterns of effects between cannabis versus alcohol use which may suggest unique etiology of cannabis versus alcohol use with depression symptoms and NU. For alcohol, the only significant directional cross-lagged effect observed was from the interaction between depression symptoms and NU predicting subsequent alcohol use, whereas for cannabis, increased use predicted greater severity of depression six months later. These associations were both limited to the 6-month to 12-month timepoints, however, with no significant effects from baseline to 6-months post baseline. Given the observational nature of the study, it is difficult to interpret potential explanations for these differences between time points, as we would have expected the effects observed to be consistent across time periods. Future research should consider examining these patterns across a longer study window to potentially capture more variability within the constructs, as well as possibly changes in the associations across constructs over time. Although we did not explore additional mechanisms (i.e., mediators) of the effects observed here, these findings suggest that coping-related alcohol use may be particularly prevalent among veterans high in NU. The relationship between cannabis and depression is complex and may vary by the severity of cannabis use and depression, other substance use, and situational factors that may drive these effects (Troup, Andrzejewski, Braunwalder, & Torrence, 2016). It is also possible that an alternative mechanism, such as social avoidance or isolation resulting from cannabis use leads to increased depression symptoms over time (Buckner, Heimberg, & Schmidt, 2011).

Strength and Limitations

To our knowledge, this is the first study to examine competing models of directionality between depression symptoms in conjunction with NU and cannabis or alcohol use. We used a statistically robust approach to examine these associations prospectively. A few limitations should be considered when interpreting the results of this study. First, the cannabis models revealed poorer fit to the data compared to alcohol models, therefore these findings require replication and should be considered preliminary. Second, this sample had relatively low levels of depression symptoms overall. Although we had a relatively high proportion of individuals with a current major depressive disorder (15%) relative to the general population (Kessler et al., 2005). Future research should examine clinical samples with higher rates of depression symptoms in order to fully characterize these associations. The sample also had relatively low proportion of current cannabis users relative to alcohol users (45 and 94%, respectively). Although including a range of cannabis use in the sample strengthens the test of its association with depression symptoms, a sample of more regular cannabis users may have impacted the strength of the relationship between depression symptoms, negative urgency, and frequency of cannabis use. Third, similar to many studies of veterans who are disproportionately male, a small number of females limit the generalizability of findings to both sexes. Gender difference is particularly important to examine in future research given higher prevalence rates of depression in women (Substance Abuse and Mental Health Services Administration., 2018). Finally, the current study focused on substance use frequency (rather than AUD and/or CUD) and depression symptoms rather than MDD. Although the use of categorical diagnoses is crucial for prioritizing individuals most in need of intervention, the utility of symptom counts vs. categorical diagnoses have been demonstrated for dimensional behaviors (Micalizzi, Brick, Marraccini, Benca-Bachman, Palmer, & Knopik, under review). Levels of both depression and substance use are variable within the population and it is important to understand how those varying levels interrelate. In light of this, different patterns of effects may emerge across dimensional and clinical characterization of variables, and this fact should be considered carefully and evaluated empirically in future studies.

Conclusions

The present study utilized data from a sample of veterans assessed across a period of 12-months to evaluate how depression symptoms and NU are related to cannabis or alcohol use over time. For alcohol, the only significant cross-lagged effect observed was from the interaction between depression symptoms and NU at 6-months post baseline predicting alcohol use at 12-months post baseline. Whereas for cannabis, increased use at 6-months post baseline predicted greater severity of depression symptoms 12-months post baseline. These findings suggest unique etiology of cannabis versus alcohol use with depression in that veterans high in depression symptoms and NU are at increased risk for more alcohol use; whereas veterans with higher levels of cannabis use are at increased risk for depression symptoms over time.

Public Significance Statement.

Depression, cannabis use, and alcohol use are significant problems among the general public and military veterans. Results from this study suggest that symptoms of depression are related to cannabis use and alcohol use in different ways among veterans. Veterans with more severe symptoms of depression and high trait negative urgency use alcohol more frequently over time.

Acknowledgements

The authors gratefully acknowledge Cassandra Tardiff, Suzanne Sales, and Timothy Souza for their contribution to the project.

This work was supported by the National Institute on Drug Abuse, R01 DA033425 (PIs Metrik, Borsari), T32 AA007459 (Gunn), T32 DA016184 (Stevens), K08 AA027551 (Gunn).

Footnotes

Disclosures: The authors have no conflicts of interest to declare (None).

References

  1. Adams ZWZ, Kaiser AJ., Lynam DR., Charnigo RJ., & Milich R (2012). Drinking motives as mediators of the impulsivity-substance use relation: Pathways for negative urgency, lack of premeditation, and sensation seeking. Addictive Behaviors, 37, 848–855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Agosti V, Nunes E, & Levin F (2002). Rates of Psychiatric Comorbidity Among U.S. Residents with Lifetime Cannabis Dependence. The American Journal of Drug and Alcohol Abuse, 28(4), 643–652. 10.1081/ADA-120015873 [DOI] [PubMed] [Google Scholar]
  3. Anestis MD, Selby EA, & Joiner TE (2007). The role of urgency in maladaptive behaviors. Behaviour Research and Therapy, 45(12), 3018–3029. 10.1016/j.brat.2007.08.012 [DOI] [PubMed] [Google Scholar]
  4. 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. Journal of Abnormal Child Psychology, 40(8), 1313–1326. 10.1007/s10802-012-9644-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Baker TB, Piper ME, McCarthy DE, Majeskie MR, & Fiore MC (2004). Addiction Motivation Reformulated: An Affective Processing Model of Negative Reinforcement. Psychological Review, 111(1), 33–51. 10.1037/0033-295X.111.1.33 [DOI] [PubMed] [Google Scholar]
  6. Birthrong A, & Latzman RD (2014). Aspects of impulsivity are differentially associated with risky sexual behaviors. Personality and Individual Differences, 57, 8–13. 10.1016/j.paid.2013.09.009 [DOI] [Google Scholar]
  7. Blanco C, Hasin DS, Wall MM, Flórez-Salamanca L, Hoertel N, Wang S, … Olfson M (2016). Cannabis use and risk of psychiatric disorders prospective evidence from a US National longitudinal study. JAMA Psychiatry, 73(4), 388–395. 10.1001/jamapsychiatry.2015.3229 [DOI] [PubMed] [Google Scholar]
  8. Bonn-Miller MO, Boden MT, Bucossi MM, & Babson KA (2014). Self-reported cannabis use characteristics, patterns and helpfulness among medical cannabis users Marcel. The American Journal of Drug and Alcohol Abuse, 40(1), 23–30. 10.3109/00952990.2013.821477 [DOI] [PubMed] [Google Scholar]
  9. Bovasso GB (2001). Cannabis Abuse as a Risk Factor for Depressive Symptoms. American Journal of Psychiatry, 158(12), 2033–2037. Retrieved from https://www.researchgate.net/profile/Gregory_Bovasso/publication/11626949_Cannabis_Abuse_as_a_Risk_Factor_for_Depressive_Symptoms/links/0c960535700d1bf72a000000.pdf [DOI] [PubMed] [Google Scholar]
  10. Brière FN, Rohde P, Seeley JR, Klein D, & Lewinsohn PM (2014). Comorbidity Between Major Depression and Alcohol Use Disorder From Adolescence to Adulthood. Comprehensive Pyschiatry, 55(3), 526–533. 10.1016/j.comppsych.2013.10.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Brooks Holliday S, Pedersen ER, & Leventhal AM (2016). Depression, posttraumatic stress, and alcohol misuse in young adult veterans: The transdiagnostic role of distress tolerance. Drug Alcohol Depend, 161, 348–355. 10.1016/j.drugalcdep.2016.02.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Browne MW, & Cudeck R (1993). Alternative ways of assessing model fit. In Bollen KA & Long JS (Eds.), Testing structural equation models (pp. 136–162). 10.1177/0049124192021002005 [DOI] [Google Scholar]
  13. Buckner JD, Heimberg RG, & Schmidt NB (2011). Social anxiety and marijuana-related problems: The role of social avoidance. Addictive Behaviors, 1–2, 129–132. 10.1016/j.addbeh.2010.08.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. 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. 10.1016/j.addbeh.2006.12.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Bulloch A, Lavorato D, Williams J, & Patten S (2012). Alcohol consumption and major depression in the general population: The critical importance of dependence. Depression and Anxiety. 10.1002/da.22001 [DOI] [PubMed] [Google Scholar]
  16. Chen C-Y, Wagner FA, & Anthony JC (2002). Marijuana use and the risk of Major Depressive Episode Epidemiological evidence from the United States National Comorbidity Survey. Social Psychiatry and Psychiatric Epidemiology, 37, 199–206. [DOI] [PubMed] [Google Scholar]
  17. Conway KP, Compton W, Stinson FES, & Grant BF (2006). Lifetime Comorbidity of DSM-IV Mood and Anxiety Disorders and Specific Drug Use Disorders: Results From the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin PsychiatryJ Clin Psychiatry, 6767(2), 247–257. [DOI] [PubMed] [Google Scholar]
  18. Copeland W, Angold A, Shanahan L, Dreyfuss J, Dlamini I, & Jane Costello E (2012). Predicting persistent alcohol problems: A prospective analysis from the Great Smoky Mountain Study. Psychological Medicine, 42(9), 1925–1935. 10.1017/S0033291711002790 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Coskunpinar A, Dir AL, & Cyders MA (2013). Multidimensionality in impulsivity and alcohol use: A meta-analysis using the UPPS model of impulsivity. Alcoholism: Clinical and Experimental Research, 37(9), 1441–1450. 10.1111/acer.12131 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Cox WM, & Klinger E (1988). A motivational model of alcohol use. Journal of Abnormal Psychology, 97, 168–180. 10.1037/0021-843X.97.2.168 [DOI] [PubMed] [Google Scholar]
  21. Cyders MA, Flory K, Rainer S, & Smith GT (2009). The Role of Personality Dispositions to Risky Behavior in Predicting First Year College Drinking. Addiction, 104(2), 193–202. 10.1111/j.1360-0443.2008.02434.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Cyders MA, Littlefield AK, Coffey S, & Karyadi KA (2014). Examination of a short English version of the UPPS-P Impulsive Behavior Scale ☆. Addictive Behaviors, 39, 1372–1376. 10.1016/j.addbeh.2014.02.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Dawe S, & Loxton NJ (2004). The role of impulsivity in the development of substance use and eating disorders. Neuroscience and Biobehavioral Reviews, 28, 343–351. 10.1016/j.neubiorev.2004.03.007 [DOI] [PubMed] [Google Scholar]
  24. De Wit H (2009). Impulsivity as a determinant and consequence of drug use: A review of underlying processes. Addiction Biology, 14(1), 22–31. 10.1111/j.1369-1600.2008.00129.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Degenhardt L, Hall W, & Lynskey M (2003). Exploring the association between cannabis use and depression. Addiction, 98, 1493–1504. Retrieved from https://pdfs.semanticscholar.org/cf3e/289629ee0c026fb25adc4c412bfed091e39e.pdf [DOI] [PubMed] [Google Scholar]
  26. Dennis M, Funk R, Harrington Godley S, Godley MD, & Waldron H (2004). Cross-validation of the alcohol and cannabis use measures in the Global Appraisal of Individual Needs (GAIN) and Timeline Followback (TLFB; Form 90) among adolescents in substance abuse treatment. Addiction, 99(2), 125–133. [DOI] [PubMed] [Google Scholar]
  27. Derefinko K, Dewall CN, Metze AV, Walsh EC, & Lynam DR (2011). Do Different Facets of Impulsivity Predict Different Types of Aggression? Aggressive Behavior, 37(3), 223–233. 10.1002/ab.20387 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Dierker L, Selya A, Lanza S, Li R, & Rose J (2018). Depression and marijuana use disorder symptoms among current marijuana users. Addictive Behaviors, 76, 161–168. 10.1016/j.addbeh.2017.08.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Dvorak RD, & Day AM (2014). Marijuana and self-regulation: Examining likelihood and intensity of use and problems. Addictive Behaviors, 39(3), 709–712. 10.1016/j.addbeh.2013.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Farris SG, Metrik J, Bonn-Miller MO, Kahler CW, & Zvolensky MJ (2016). Anxiety Sensitivity and Distress Intolerance as Predictors of Cannabis Dependence Symptoms, Problems, and Craving: The Mediating Role of Coping Motives. Journal of Studies on Alcohol and Drugs, 77(6), 889–897. Retrieved from http://www.jsad.com/doi/pdf/10.15288/jsad.2016.77.889 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Feingold D, Fox J, Rehm J, & Lev-Ran S (2015). Natural outcome of cannabis use disorder: A 3-year longitudinal follow-up. Addiction, 110(12), 1963–1974. 10.1111/add.13071 [DOI] [PubMed] [Google Scholar]
  32. Fergusson DM, Boden JM, & Horwood LJ (2009). Tests of causal links between alcohol abuse or dependence and major depression. Archives of General Psychiatry. 10.1001/archgenpsychiatry.2008.543 [DOI] [PubMed] [Google Scholar]
  33. Gilman SE, & Abraham HD (2001). A longitudinal study of the order of onset of alcohol dependence and major depression. Drug and Alcohol Dependence. 10.1016/S0376-8716(00)00216-7 [DOI] [PubMed] [Google Scholar]
  34. Goldman M, Suh JJ, Lynch KG, Szucs R, Ross J, Xie H, … Oslin DW (2010). Identifying Risk Factors for Marijuana Use Among Veterans Affairs Patients. Journal of Addiction Medicine, 4(1), 47–51. 10.1097/ADM.0b013e3181b18782 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Gonzalez VM, Reynolds B, & Skewes MC (2011). Role of impulsivity in the relationship between depression and alcohol problems among emerging adult college drinkers. Experimental and Clinical Psychopharmacology, 19(4), 303–313. 10.1037/a0022720 [DOI] [PubMed] [Google Scholar]
  36. Grant BF (2004). Prevalence and Co-Occurrence of Substance Use Disorders and Independent Mood and Anxiety Disorders on Alcohol and Related Conditions. Archives of General Psychiatry, 61(2), 807–816. [DOI] [PubMed] [Google Scholar]
  37. Grant BF, Goldstein RB, Saha TD, Chou SP, Jung J, Zhang H, … Hasin DS (2015). Epidemiology of DSM-5 Alcohol Use Disorder. JAMA Psychiatry, 72(8), 757 10.1001/jamapsychiatry.2015.0584 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Grant BF, Saha TD, Ruan WJ, Goldstein RB, Chou SP, Jung J, … Hasin DS (2016). Epidemiology of DSM-5 Drug Use Disorder: Results From the National Epidemiologic Survey on Alcohol and Related Conditions–III Bridget. JAMA Psychiatry, 73(1), 39 10.1001/jamapsychiatry.2015.2132 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Guillot CR, Pang RD, & Leventhal AM (2014). Anxiety sensitivity and negative urgency: a pathway to negative reinforcement-related smoking expectancies. Journal of Addiction Medicine, 8(3), 189–194. 10.1097/ADM.0000000000000017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Gunn RL, Finn PR, Endres MJ, Gerst KR, & Spinola S (2013). Dimensions of disinhibited personality and their relation with alcohol use and problems. Addictive Behaviors, 38(7), 2352–2360. 10.1016/j.addbeh.2013.02.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Gunn RL, Jackson K, Borsari B, & Metrik J (2018). Negative Urgency Mediates the Relationship Between Depression and Marijuana Problems. Borderline Personality Disorder and Emotion Dysregulation, 5(10). [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Hasin DS (2018). US Epidemiology of Cannabis Use and Associated Problems. Neuropsychopharmacology Reviews, 43(10), 195–212. 10.1038/npp.2017.198 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Holahan CJ, Holahan CK, Moos RH, Cronkite RC, & Randall PK (2003). Drinking to cope and alcohol use and abuse in unipolar depression: A 10-year model. Journal of Abnormal Psychology, 112(1), 159–165. 10.1037/0021-843X.112.1.159 [DOI] [PubMed] [Google Scholar]
  44. Hu LT, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  45. Jacobson IG, Ryan MAK, Hooper TI, Smith TC, Amoroso PJ, Boyko EJ, … Bell NS (2008). Alcohol use and alcohol-related problems before and after military combat deployment. JAMA - Journal of the American Medical Association, 300(6), 663–675. 10.1001/jama.300.6.663 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Johnson KA, Bonn-Miller MO, Leyro TM, & Zvolensky MJ (2009). Anxious Arousal and Anhedonic Depression Symptoms and the Frequency of Current Marijuana Use: Testing the Mediating Role of Marijuana-Use Coping Motives Among Active Users. Journal of Studies on Alcohol and Drugs, 70, 543–550. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2696295/pdf/jsad543.pdf [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Kaiser AJ, Milich R, Lynam DR, & Charnigo RJ (2012). Negative Urgency, Distress Tolerance, and substance abuse among college students. Addictive Behaviors, 37(10), 1075–1083. 10.1016/j.addbeh.2012.04.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Kessler RC, Chiu WT, Demler O, & Walters EE (2005). Prevalence, Severity, and Comorbidity of 12-Month DSM-IV Disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 62(6), 617–627. 10.1001/archpsyc.62.6.617 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Kessler RC, Heeringa SG, Stein MB, Colpe LJ, Fullerton CS, Hwang I, … Ursano RJ (2014). Thirty-Day Prevalence of DSM-IV Mental Disorders Among Nondeployed Soldiers in the US Army. JAMA Psychiatry, 71(5), 504 10.1001/jamapsychiatry.2014.28 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Koob GF, & Volkow ND (2010). Neurocircuitry of addiction. Neuropsychopharmacology, Vol. 35, pp. 217–238. 10.1038/npp.2009.110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Kuo PH, Gardner CO, Kendler KS, & Prescott CA (2006). The temporal relationship of the onsets of alcohol dependence and major depression: Using a genetically informative study design. Psychological Medicine. 10.1017/S0033291706007860 [DOI] [PubMed] [Google Scholar]
  52. Lev-Ran S, Roerecke M, Le Foll B, George TP, Mckenzie K, & Rehm J (2014). The association between cannabis use and depression: a systematic review and meta-analysis of longitudinal studies. Psychological Medicine, 44, 797–810. 10.1017/S0033291713001438 [DOI] [PubMed] [Google Scholar]
  53. Lynam DR, Smith GT, Whiteside SA, & Cyders MA (2006). The UPPS-P: Assessing five personality pathways to impulsive behavior. In West Lafayette, IN: Purdue University; 10.1016/j.rpsp.2016.05.002 [DOI] [Google Scholar]
  54. Martens MP, Neighbors C, Lewis MA, Lee CM, Oster-Aaland L, & Larimer ME (2008). The Roles of Negative Affect and Coping Motives in the Relationship Between Alcohol Use and Alcohol-Related Problems Among College Students. Journal of Studies on Alcohol and Drugs, 69(3), 412–419. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4942848/pdf/nihms799481.pdf [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. McCarty CA, Kosterman R, Mason WA, McCauley E, Hawkins JD, Herrenkohl TI, & Lengua LJ (2009). Longitudinal associations among depression, obesity and alcohol use disorders in young adulthood. General Hospital Psychiatry. 10.1016/j.genhosppsych.2009.05.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Metrik J, Bassett S, Aston ER, Jackson K, & Borsari B (2018). Medicinal versus recreational cannabis use among returning veterans. Translational Issues in Psychological Science, 4(1), 6–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Metrik J, Jackson K, Bassett SS, Zvolensky MJ, Seal K, & Borsari B (2016). The mediating roles of coping, sleep, and anxiety motives in cannabis use and problems among returning veterans with PTSD and MDD. Psychology of Addictive Behaviors, 30(7), 743–754. 10.1037/adb0000210 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Micalizzi L, Brick LA, Marraccini ME, Benca-Bachman C, Palmer RHC, & Knopik VS (n.d.). SNP Heritability and Differential Patterns of Genetic Overlap between Inattention and Four Neurocognitive Factors in Youth. [DOI] [PMC free article] [PubMed]
  59. Muthén LK, & Muthén BO (n.d.). Mplus User’s Guide. (Eighth; Muthen & Muthen, Ed.). 10.1111/j.1600-0447.2011.01711.x [DOI] [Google Scholar]
  60. Novak BK, Novak KD, Lynam DR, & Foti D (2016). Individual differences in the time course of reward processing: Stage-specific links with depression and impulsivity. Biological Psychology, 119, 79–90. 10.1016/j.biopsycho.2016.07.008 [DOI] [PubMed] [Google Scholar]
  61. Pang RD, Farrahi L, Glazier S, Sussman S, & Leventhal AM (2014). Depressive symptoms, negative urgency and substance use initiation in adolescents. Drug and Alcohol Dependence, 144, 225–230. 10.1016/j.drugalcdep.2014.09.771 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Patton GC, Coffey C, Carlin JB, Degenhardt L, Lynskey M, & Hall W (2002). Cannabis use and mental health in young people: cohort study. British Medical Journal, 325, 1195–1198. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC135489/pdf/1195.pdf [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Rohde P, Lewinsohn PM, Kahler CW, Seeley JR, & Brown RA (2001). Natural course of alcohol use disorders from adolescence to young adulthood. Journal of the American Academy of Child and Adolescent Psychiatry. 10.1097/00004583-200101000-00020 [DOI] [PubMed] [Google Scholar]
  64. Seal KH, Cohen G, Waldrop A, Cohen BE, Maguen S, & Ren L (2011). Substance use disorders in Iraq and Afghanistan veterans in VA healthcare, 2001 −2010: Implications for screening, diagnosis and treatment. Drug and Alcohol Dependence, 116, 93–101. 10.1016/j.drugalcdep.2010.11.027 [DOI] [PubMed] [Google Scholar]
  65. Seal KH, Metzler TJ, Gima KS, Bertenthal D, Maguen S, & Marmar CR (2009). Trends and Risk Factors for Mental Health Diagnoses Among Iraq and Afghanistan Veterans Using Department of Veterans Affairs Health Care, 2002–2008. American Journal of Public Health, 99(9), 1651–1658. 10.2105/AJPH.2008.150284 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Settles RE, Fischer S, Cyders MA, Combs JL, Gunn RL, & Smith GT (2012). Negative urgency: a personality predictor of externalizing behavior characterized by neuroticism, low conscientiousness, and disagreeableness. Journal of Abnormal Psychology, 121(1), 160–172. 10.1037/a0024948 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Simons JS, & Carey KB (2006). An affective and cognitive model of marijuana and alcohol problems. Addictive Behaviors, 31(9), 1578–1592. 10.1016/j.addbeh.2005.12.004 [DOI] [PubMed] [Google Scholar]
  68. Simons JS, Gaher RM, Correia CJ, Hansen CL, & Christopher MS (2005). An affective-motivational model of marijuana and alcohol problems among college students. Psychology of Addictive Behaviors, 19(3), 326–334. 10.1037/0893-164X.19.3.326 [DOI] [PubMed] [Google Scholar]
  69. Smith GT, & Cyders MA (2016). Integrating affect and impulsivity: The role of positive and negative urgency in substance use risk. Drug and Alcohol Dependence, 163, S3–S12. 10.1016/j.drugalcdep.2015.08.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Smith GT, Fischer S, Cyders MA, Annus AM, Spillane NS, & Mccarthy DM (2007). On the Validity and Utility of Discriminating Among Impulsivity-Like Traits. Assessment, 14(2), 155–170. 10.1177/1073191106295527 [DOI] [PubMed] [Google Scholar]
  71. Sobell L, & Sobell M (1992). Timeline Follow-Back: A technique for assessing self-reported alcohol consumption In Psychosocial and biological methods (pp. 41–72). Totowa, NJ: Humana Press. [Google Scholar]
  72. Stamates AL, & Lau-Barraco C (2017). The dimensionality of impulsivity: Perspectives and implications for emerging adult drinking. Experimental and Clinical Psychopharmacology, 25(6), 521–533. 10.1037/pha0000153 [DOI] [PubMed] [Google Scholar]
  73. Substance Abuse and Mental Health Services Administration. (2018). 2017 National Survey on Drug Use and Health: Methodological summary and definitions. Rockville, MD. [Google Scholar]
  74. Substance Abuse and Mental Health Services Administration. (2019). Key Substance Use and Mental Health Indicators Results from the 2018 National in the United States: Survey on Drug Use and Health (HHS Publication No. PEP19–5068, NSDUH Series H-54). (Vol. 7). Retrieved from https://www.samhsa.gov/data/%0AOriginating
  75. Sullivan LE, Fiellin DA, & O’Connor PG (2005). The prevalence and impact of alcohol problems in major depression: A systematic review. American Journal of Medicine, 118(4), 330–341. 10.1016/j.amjmed.2005.01.007 [DOI] [PubMed] [Google Scholar]
  76. Sullivan LE, Goulet JL, Justice AC, & Fiellin DA (2011). Alcohol Consumption and Depressive Symptoms Over Time: A Longitudinal Study of Patients With and Without HIV Infection. Drug and Alcohol Dependence, 117(2–3), 158–163. 10.1038/jid.2014.371 [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Swendsen JD, & Merikangas KR (2000). The comorbidity of depression and substance use disorders. Clinical Psychology Review, 20(2), 173–189. Retrieved from https://ac.els-cdn.com/S0272735899000264/1-s2.0-S0272735899000264-main.pdf?_tid=f37f2c22-c4c2-11e7-924e-00000aacb360&acdnat=1510172905_01d0c9a36ca158c0f0d69a161c0290ed [DOI] [PubMed] [Google Scholar]
  78. Swendsen JD, Tennen H, Carney MA, Affleck G, Willard A, & Hromi A (2000). Mood and alcohol consumption: An experience sampling test of the self-medication hypothesis. Journal of Abnormal Psychology, 109(2), 198–204. 10.1037//0021-843X.109.2.198 [DOI] [PubMed] [Google Scholar]
  79. Teeters JB, Lancaster CL, Brown DG, & Back SE (2017). Substance use disorders in military veterans: prevalence and treatment challenges. Substance Abuse and Rehabilitation, 8, 69–77. 10.2147/SAR.S116720 [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Troup LJ, Andrzejewski JA, Braunwalder JT, & Torrence RD (2016). The relationship between cannabis use and measures of anxiety and depression in a sample of college campus cannabis users and non-users post state legalization in Colorado. PeerJ, 4, e2782 10.7717/peerj.2782 [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Verdejo-GarcƔá A, Lawrence AJ, & Clark L (2008). Impulsivity as a vulnerability marker for substance-use disorders: Review of findings from high-risk research, problem gamblers and genetic association studies. Neuroscience and Biobehavioral Reviews, 32, 777–810. 10.1016/j.neubiorev.2007.11.003 [DOI] [PubMed] [Google Scholar]
  82. Vrieze SI (2012). Model selection and psychological theory: A discussion of the differences between the Akaike information criterion (AIC) and the Bayesian information criterion (BIC). Psychological Methods, 17(2), 228 10.1037/a0027127 [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Whiteside SP, & Lynam DR (2001). The Five Factor Model and impulsivity: using a structural model of personality to understand impulsivity. Personality and Individual Differences, 30, 669–689. Retrieved from https://ac.els-cdn.com/S0191886900000647/1-s2.0-S0191886900000647-main.pdf?_tid=0633060e-c31a-11e7-a480-00000aab0f6b&acdnat=1509990400_ef898c3a74d1ae1e88ef3039a64f7cca [Google Scholar]
  84. Wilkinson ST, van Schalkwyk GI, Davidson L, & D’Souza DC (2016). The Formation of Marijuana Risk Perception in a Population of Substance Abusing Patients. Psychiatric Quarterly, 87, 177–187. 10.1007/s11126-015-9369-z [DOI] [PubMed] [Google Scholar]
  85. Womack SR, Shaw DS, Weaver CM, & Forbes EE (2016). Bidirectional associations between cannabis use and depressive symptoms from adolescence through early adulthood among at-risk young men. Journal of Studies on Alcohol and Drugs, 77, 287–297. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4803661/pdf/jsad.2016.77.287.pdf [DOI] [PMC free article] [PubMed] [Google Scholar]
  86. Yoon G, Petrakis IL, & Rosenheck RA (2015). Correlates of major depressive disorder with and without comorbid alcohol use disorder nationally in the veterans health administration. American Journal on Addictions, 24(5), 419–426. 10.1111/ajad.12219 [DOI] [PubMed] [Google Scholar]

RESOURCES