Difficulties Regulating Positive Emotions and Substance Misuse: The Influence of Sociodemographic Factors

Abstract

Background:

Alcohol and drug misuse present significant public health concerns due to their high prevalence and deleterious outcomes. A growing body of research provides support for the role of difficulties regulating positive emotions in alcohol and drug misuse. However, research is needed to better understand for whom difficulties regulating positive emotions are most strongly associated with alcohol and drug misuse to inform assessment and treatment efforts.

Objectives:

The goal of the present study was to examine potential sociodemographic moderators (i.e., age, gender, ethnicity, race, income, and educational attainment) in the relations between difficulties regulating positive emotions and alcohol and drug misuse.

Methods:

Participants were 373 trauma-exposed adults (57.1% female, 75.8% White) recruited from the community.

Results:

Significant differences were identified across sociodemographic groups regarding difficulties regulating positive emotions (i.e., gender, ethnicity, race, and income) and alcohol use (i.e., gender). Moderation analyses revealed a significant interaction between difficulties regulating positive emotions and gender on drug misuse (b = 0.08, p < .001), such that the association was significant for females (b = 0.11, p < .001) but not males (b = .03, p = .05).

Conclusions:

Results suggest the importance of developing gender-sensitive recommendations for the assessment and treatment of substance misuse, and of incorporating techniques focused on addressing difficulties regulating positive emotions.

Introduction

Estimates from the National Survey on Drug Use and Health (Substance Abuse and Mental Health Services Administration, 2017) indicate that more than half of individuals ages 12 years or older report past month alcohol use, and approximately 10.6% report past month illicit drug use. Further, approximately 5.6% of individuals 12 years or older meet criteria for alcohol use disorder, while 2.7% meet criteria for another illicit drug use disorder (Substance Abuse and Mental Health Services Administration, 2017). This presents a significant public health concern given the deleterious outcomes associated with substance use (Hull & Bond, 1986; Newcomb & Locke, 2005). For instance, individuals who use substances are at increased risk for negative health outcomes (e.g., illness and injury; Newcomb & Locke, 2005; Whiteford et al., 2013), risky behaviors (e.g., risky sexual behavior, polysubstance use; Bohnert, Roeder, & Ilgen, 2010; Celentano, Latimore, & Mehta, 2008), and legal consequences (e.g., arrests, incarceration; Iguchi et al., 2002). The Global Burden of Disease Study has reported that substance use disorders are among the leading contributor to years lived with disability (Whiteford et al., 2013). Thus, it is of great public health importance that attention be paid to risk factors associated with substance use.

Difficulties regulating emotions have been identified as a key mechanism in the etiology, maintenance, and treatment of alcohol and drug use (for a review, see Weiss, Sullivan, & Tull, 2015c; Weiss, Tull, & Sullivan, 2015d). Difficulties regulating emotions represents a multi-faceted construct including: (a) lack of awareness, understanding, and acceptance of emotions; (b) inability to control behaviors when experiencing emotional distress; (c) lack of access to situationally appropriate strategies for modulating the duration and/or intensity of emotional responses to meet individual goals and situational demands; and (d) an unwillingness to experience emotional distress as part of pursuing meaningful activities in life (Gratz & Roemer, 2004; Gratz & Tull, 2010). Correlational studies have found that individuals who misuse alcohol (Fox, Hong, & Sinha, 2008) and drugs (Fox, Axelrod, Paliwal, Sleeper, & Sinha, 2007) report greater difficulties regulating emotions. Moreover, increased difficulties regulating emotions have been found to be associated with worse alcohol and drug use outcomes, including increased alcohol consumption (Dvorak et al., 2014; Messman-Moore & Ward, 2014), more problematic patterns of substance use (Tull, Bardeen, DiLillo, Messman-Moore, & Gratz, 2015), and increased negative alcohol-related consequences (Dvorak et al., 2014; Messman-Moore & Ward, 2014). Further, some work has conceptualized difficulties regulating emotions as being a driving factor underlying initiation of substance use (Kober & Bolling, 2014; Sher & Grekin, 2007), and empirical research has found difficulties regulating emotions to be prospectively associated with increased risk of developing a substance use disorder (Moffitt et al., 2011). Finally, difficulties regulating emotions have been identified as an important factor in influencing treatment outcomes, including alcohol and drug use during and after cognitive behavioral treatment (Berking et al., 2011) and dialectical behavior therapy (Axelrod, Perepletchikova, Holtzman, & Sinha, 2011).

One key limitation of extant literature in this area has been its near-exclusive focus on difficulties regulating negative emotional experiences. Positive emotions have been found to play a significant role in the development of alcohol and drug misuse (Baker, Piper, McCarthy, Majeskie, & Fiore, 2004), underscoring the importance of assessing for and intervening on difficulties regulating positive emotions prior to an individual developing alcohol or drug misuse. Indeed, preliminary work suggests that some individuals experience difficulties regulating positive emotions in addition to difficulties regulating negative emotions (Weiss, Darosh, Contractor, Schick, & Dixon-Gordon, 2019; Weiss, Gratz, & Lavender, 2015a). For instance, Weiss, Gratz, and Lavender (2015) developed the Difficulties in Emotion Regulation Scale–Positive (DERS-P), and found evidence for three dimensions of difficulties regulating positive emotions: nonacceptance of positive emotions (Nonacceptance; e.g., “When I’m happy, I become scared and fearful of those feelings”), difficulties engaging in goal-directed behaviors in the context of positive emotions (Goals; e.g., “When I’m happy, I have difficulty focusing on other things”), and difficulties controlling impulsive behaviors when experiencing positive emotions (Impulse; e.g., “When I’m happy, I have difficulty controlling my behaviors”). A follow up psychometric study confirmed the presence of these three factors and provided evidence of their internal reliability and convergent/discriminant validity (Weiss et al., 2019).

Since the development of the DERS-P, a growing number of studies have begun to examine the role of difficulties regulating positive emotions in alcohol and drug misuse. Weiss, Forkus, Contractor and Schick (2018) modeled the associations between the three dimensions of difficulties regulating positive emotions and alcohol and drug misuse; results provided support for significant positive relationships, wherein greater difficulties regulating positive emotions were associated with greater alcohol and drug misuse (Weiss, Forkus, Contractor, & Schick, 2018b). Relatedly, Weiss, Darosh, Contractor, Schick, and Dixon-Gordon (2019) found that difficulties regulating positive emotions differentiated between individuals with (versus without) alcohol and drug misuse. Weiss, Risi, Bold, Sullivan, and Dixon-Gordon (2018) expanded upon this, using daily diary methods to examine difficulties regulating positive emotions as a moderator of the association between positive affect and alcohol use. They found that level of positive affect was associated with alcohol use for those reporting greater, but not fewer, levels of difficulties regulating positive emotions (Weiss, Risi, Bold, Sullivan, & Dixon-Gordon, 2018c). Finally, Weiss et al. (2018) found that those with difficulties regulating both negative and positive emotions reported the highest levels of alcohol and drug misuse, while a class defined by greater difficulties regulating negative emotions alone demonstrated alcohol and drug misuse comparable to a class defined by low difficulties regulating both negative and positive emotions (Weiss et al., 2018a). Taken together, these results suggest the need for further examination of the role of difficulties regulating positive emotions in alcohol and drug misuse.

One important avenue for future research in this area is to identify factors that moderate (i.e., influence the strength and/or directionality of) the relations between difficulties regulating positive emotions and alcohol or drug misuse. Such research may clarify for whom difficulties regulating positive emotions may be more strongly associated with alcohol and drug misuse. Identification of moderating variables has been recognized as an important research goal given the relevance of these types of questions to understanding the generalizability and specificity of research findings, and for gaining a fuller understanding of the complexity of the interrelations among constructs of interest (MacKinnon, 2011). Investigations in this area have key implications for practice, including helping to identify those at greatest risk for substance use, and thus who would benefit most from preventative intervention. Alternatively, such findings may lead to better understanding of for whom interventions targeting difficulties regulating positive emotions would be most effective, thereby informing the tailoring of such interventions (MacKinnon, 2011). Further, this line of research also has important implications for theory: moderating variables may influence the etiology and maintenance of alcohol and drug misuse, such as by influencing one’s motives for using alcohol (LaBrie, Lac, Kenney, & Mirza, 2011) or drugs (Gerrard et al., 2012; Terry-McElrath, O’Malley, & Johnston, 2009).

Sociodemographic variables (i.e., age, gender, race/ethnicity, income, education) have been previously found to influence substance use, with findings suggesting that there exist age, gender, racial/ethnic, income, and education-related differences in risk for experiencing negative consequences related to substance use (Acevedo et al., 2012; Gillmore et al., 1990; MacArthur et al., 2012; McCabe et al., 2007). Risk of experiencing negative consequences related to substance use increases as age of initiation of substance use decreases (Henry et al., 2011; King & Chassin, 2007). Further, while White individuals are more likely to report drug use (McCabe et al., 2007), non-White individuals are significantly more likely to experience problems related to substance use (Iguchi, Bell, Ramchand, & Fain, 2005; Schmidt, Ye, Greenfield, & Bond, 2007). Regarding gender, males tend to report greater substance use and to experience higher rates of substance use disorders (though this trend has narrowed over recent years; McHugh, Votaw, Sugarman, & Greenfield, 2018), including alcohol use disorders (SAMHSA, 2015), and experience greater physical and legal consequences of alcohol use (Perkins, 2002), while females tend to experience some psychosocial consequences of substance use (e.g., interpersonal violence) at higher rates than do males (McHugh et al., 2018; Zilberman, Tavares, & El-Guebaly, 2004). Research examining the role of education has found substance use to be linked to educational attainment, such that individuals who use substances are at decreased likelihood of accessing higher education (King, Meehan, Trim, & Chassin, 2006). Finally, although mixed, there is some evidence that those with higher personal income have greater economic access to substances and therefore are more likely to engage in substance use (Hanson & Chen, 2007; Kar, Haynie, Luk, & Simons-Morton, 2018; Patrick, Wightman, Schoeni, & Schulenberg, 2012).

Although less studied, research has also found support for differences in difficulties regulating positive emotions across sociodemographic groups, such that greater difficulties regulating positive emotions were found among individuals who were younger, male, Hispanic, non-White, working full-time, and making less than $50,000 a year (Weiss et al., 2019). Further, the relation of other affective processes and alcohol and drug misuse has been found to vary as a function of sociodemographic factors (Nolen-Hoeksema, 2012). For instance, Bornovalova, Ouimette, Crawford, and Levy (2009) found that, for women, difficulties controlling impulsive behaviors were more strongly related to substance use, while for men, lack of awareness and clarity of emotions were more strongly associated with substance use (Bornovalova, Ouimette, Crawford, & Levy, 2009).

To extend existing research in this area, the goal of this study was to examine potential sociodemographic moderators of the relations between difficulties regulating positive emotions and alcohol and drug misuse. Specifically, we examined the role of age, gender, ethnicity, race, income, and educational attainment in the relations between difficulties regulating positive emotions and alcohol and drug misuse. We examined these associations in a sample of trauma-exposed adults given previous literature supporting the notion that difficulties regulating positive emotions are prevalent among individuals who have experienced a traumatic event (Weiss, Dixon-Gordon, Peasant, & Sullivan, 2018c) and underlie the relationship between traumatic stress symptoms and alcohol and drug misuse (Weiss, Schick, Contractor, & Dixon-Gordon, 2019b). In particular, this prior research suggests that trauma-exposed individuals may be motivated to avoid positive emotions through substance misuse (Weiss, Forkus, Raudales, Schick, & Contractor, in press) because some positive emotions are experienced as aversive (e.g., elicit distressing physiological arousal [Weiss, Dixon-Gordon, et al., 2018] or competing negative cognitions [Frewen, Dozois, & Lanius, 2012]). In line with previous literature, we hypothesized that greater difficulties regulating positive emotions would be associated with greater alcohol and drug misuse. Due to their nature, no specific hypotheses were generated regarding the moderation analyses.

Methodology

Procedure/Participants

Data were collected as part of a larger study aimed at developing a novel measure for assessing risky behaviors among individuals with traumatic life experiences. Participants were recruited from Amazon’s Mechanical Turk (MTurk) platform. Beyond generating reliable data (Buhrmester, Kwang, & Gosling, 2011; Shapiro, Chandler, & Mueller, 2013), MTurk’s subject pool is diverse (Buhrmester et al., 2011) and represents the general population in terms of demographics (Mischra & Carleton, 2017) and prevalence of certain mental health problems (Shapiro et al., 2013).

Participants 18 years of age and older were screened for the larger study based upon three inclusionary criteria: (1) living in North America; (2) working knowledge of the English language; and (3) endorsed experiencing a traumatic experience (TE) screened with the Criterion A question of the Primary Care PTSD Screen (Prins et al., 2015). Participants who met eligibility criteria provided informed consent and completed the survey on Qualtrics. Participants were compensated $1.25 for study participation. All procedures were approved by the Institutional Review Board at a U.S. university.

Exclusions and Missing Data

Of the obtained 891 responses, duplicate responses were excluded for 18 participants who attempted to answer the questionnaire multiple times (47 responses; remaining n = 844). We then excluded 150 participants not meeting one or more inclusionary criteria (remaining n = 694), 122 participants (remaining n = 572) who failed to pass any of four validity checks interspersed in the study to ensure attentive responding (three items; e.g., participants being asked to rate “I have never brushed my teeth” on a 6-point scale ranging from “strongly disagree” to “strongly agree”) and comprehension (one item, asking participants to click on a little blue circle rather than on the scale with items labelled from 1 to 5; Meade & Craig, 2012; Oppenheimer, Meyvis, & Davidenko, 2009; Thomas & Clifford, 2017), and 97 participants for missing data on all measures (remaining n = 475). Using data obtained from the Life Event Checklist for DSM-5 (LEC-5; Weathers et al., 2013), we excluded 11 participants who either did not endorse a TE, or did not identify their index (i.e., most distressing) TE (remaining n = 464). Finally, we excluded 92 participants missing more than 30% item-level data on any primary variable of interest (see Measures).

The final MTurk sample included 373 participants. Average age of participants was 35.74 years (SD = 11.10), 213 were female (57.1%), and 283 were White (75.8%). Approximately one-quarter of participants reported alcohol use consistent with a diagnosis of alcohol use disorder (n = 91, 24.4%), based on a cutoff score of 5 or greater on the Alcohol Use Disorder Identification Test Consumption items (Bush et al., 1998). Approximately one-sixth of participants reported drug use consistent with a diagnosis of drug use disorder (n = 58, 15.5%), based on a cutoff score of 3 or greater on the Drug Abuse Screening Test (Skinner, 1982). The most common index TEs reported on the LEC-5 were: transportation accidents (17.7%), natural disasters (14.2%), and sexual assault (13.1%). See Table 1 for further sample characteristics and prevalence rates of index traumatic events.

Table 1.

Sample Characteristics

	M (SD)	n (%)
Age	35.74 (11.10)
Gender
Male		154 (41.3%)
Female		213 (57.1%)
Ethnicity
Hispanic/Latinx		46 (12.3%)
Not Hispanic/Latinx		321 (86.1%)
Race
White		283 (75.9%)
African American/Black		39 (10.5%)
Asian		41 (11.0%)
American Indian/Alaskan Native		19 (5.1%)
Native Hawaiian/Pacific Islander		2 (0.5%)
Employment Status
Full-time		271 (72.7%)
Part-time		54 (14.5%)
Unemployed		29 (7.8%)
Retired		12 (3.2%)
Unemployed Student		7 (1.9%)
Income
Less than $15,000		35 (9.4%)
$15,000 to $24,999		52 (13.9%)
$25,000 to $34,999		56 (15.0%)
$35,000 to $49,999		48 (12.9%)
$50,000 to $64,999		71 (19.0%)
$65,000 to $79,999		33 (8.8%)
$80,000 or higher		78 (20.9%)
Study Variables of Interest
DERS-P	19.47 (10.55)
AUDIT-C Continuous Score	3.08 (2.53)
AUDIT-C Probable Alcohol Use Disorder Score*		91 (24.4%)
DAST-10 Continuous Score	1.25 (1.96)
DAST-10 Probable Drug Use Disorder Score**		58 (15.5%)
Index Traumatic Events
Natural disaster		53 (14.2%)
Fire or explosion		19 (5.1%)
Transportation accident		66 (17.7%)
Serious accident at work, home, or during recreational activity		10 (2.7%)
Exposure to toxic substance		2 (0.5%)
Physical assault		32 (8.6%)
Assault with a weapon		11 (2.9%)
Sexual assault		49 (13.1%)
Other unwanted or uncomfortable sexual experience		10 (2.7%)
Combat or exposure to a war-zone		3 (0.8%)
Captivity		2 (0.5%)
Life-threatening illness or injury		30 (8.0%)
Severe human suffering		4 (1.1%)
Sudden violent death		26 (7.0%)
Sudden accidental death		25 (6.7%)
Serious injury, harm, or death you caused to someone else		7 (1.9%)
Any other very stressful event or experience		21 (5.6%)

Note: DERS-P = Difficulties in Emotion Regulation Scale – Positive. AUDIT-C = Alcohol Use Disorder Identification Test – Consumption Questions. DAST-10 = Drug Abuse Screening Test.

^*Based on cutoff score of 5 or greater on AUDIT-C.

^**Based on cutoff score of 3 or greater on DAST-10.

Measures

The Life Events Checklist for DSM-5 (LEC-5; Weathers et al., 2013) is a 17-item self-report measure assessing lifetime experiences of traumatic events. Participants indicate their exposure to each event on a 6-point scale: happened to me, witnessed it, learned about it, part of my job, not sure, and does not apply. Endorsement of any of the first four response options is considered a positive endorsement of a traumatic event consistent with the DSM-5 Criterion A for PTSD (American Psychiatric Association, 2013). The LEC was used in the present study to exclude individuals who had not experienced a traumatic event.

The Difficulties in Emotion Regulation Scale – Positive (DERS-P; Weiss, Gratz, & Lavender, 2015b) is a 13-item self-report measure that assesses difficulties regulating positive emotions across three subscales: nonacceptance of positive emotions (DERS-P Accept), difficulties engaging in goal-directed behaviors when experiencing positive emotions (DERS-P Goals), and difficulties controlling impulsive behaviors when experiencing positive emotions (DERS-P Impulse). Higher scores indicate greater difficulties regulating positive emotions. Participants rate each item using a 5-point Likert-type scale (1 = almost never, 5 = almost always). The DERS-P has good psychometric properties (Weiss et al., 2019; Weiss et al., 2015b), and Cronbach’s α in the current sample was .96.

The Alcohol Use Disorder Identification Test Consumption Questions (AUDIT-C; Bush et al., 1998) is a 3-item self-report measure assessing heavy drinking. Participants rate each item using a 5-point Likert-type scale. Higher scores indicate greater alcohol misuse, with possible scores ranging from 0 to 12. The AUDIT-C has good psychometric properties (Bush et al., 1998), and Cronbach’s α in the current sample was .75.

The Drug Abuse Screening Test (DAST-10; Skinner, 1982) is a 10-item self-report measure that assesses the presence of problems related to drug use, such as occupational or relational problems, illegal activities, or regret. Responses to each item have 1 (yes) and 0 (no) options. Higher scores indicate greater drug misuse, with possible scores ranging from 0 to 10. The DAST-10 demonstrates good reliability and validity (Skinner, 1982), and Cronbach’s α in the present sample was .84.

Participants self-reported on sociodemographic characteristics, including age, gender, race, ethnicity, income, and years of education completed.

Analytic Plan

As recommended by Tabachnick and Fidell (2007), all study variables were assessed for adherence to assumptions of generalized linear models. Sociodemographic variables were recoded due to the small size of some subgroups making it infeasible to conduct analyses on each subgroup (i.e., gender [male (n = 154) versus female (n = 219)], ethnicity [Hispanic (n = 46) versus non-Hispanic (n = 327)], race [White (n = 283) versus non-White (n = 90], income [less than $50k (n = 191) versus $50k and greater (n = 182)], and education attainment [high school degree or less (n = 50) versus more than high school degree (n = 323)]). Income was split at the median for the present sample. To examine the effect of age on difficulties regulating positive emotions and drug and alcohol misuse, Pearson correlation coefficients were calculated. Then, a series of t tests were conducted to explore other sociodemographic differences in difficulties regulating positive emotions and drug and alcohol misuse. Pearson product-moment correlations were calculated to explore the bivariate associations of difficulties regulating positive emotions and drug and alcohol misuse.

To address the question of whether sociodemographic characteristics, difficulties regulating positive emotions and their interactions are associated with alcohol and drug misuse, a series of moderation analyses were conducted with the PROCESS SPSS macro as recommended by Hayes (2012). The PROCESS procedures use ordinary least squares regression and bootstrapping methodology, which confers more statistical power than do standard approaches to statistical inference and does not rely on distributional assumptions. Bootstrapping was done with 1,000 random samples generated from the observed covariance matrix to estimate bias-corrected 95% confidence intervals (CIs) and significance values. For interactions found to be significant, following the methods described by Aiken and West (1991), we plotted regression slopes of differences in alcohol and drug misuse and conducted follow-up analyses to examine whether the slopes of the regression lines differed significantly from zero. A Benjamini-Hochberg adjustment was utilized to minimize both Type I and Type II error for the moderation analyses (Benjamini & Hochberg, 1995). Specifically, for each model, the p values for these effects were rank ordered by size. Then, each individual p-value’s Benjamini-Hochberg critical value was calculated using the formula (i/m)Q, where i = the individual p-value’s rank, m = the total number of tests, and Q = the p value (.05). Original p values are then compared to respective Benjamini-Hochberg critical values. This method preserves an overall Type I error without increasing the risk for Type II error and unnecessarily reducing statistical power.

Results

Sociodemographic Differences in Difficulties Regulating Positive Emotions and Alcohol and Drug Misuse

See Table 2 for independent sample t tests examining sociodemographic group differences in difficulties regulating positive emotions and, alcohol and drug misuse. Findings revealed significant differences in difficulties regulating positive emotions; men, Hispanic individuals, non-White individuals, and individuals making less than $50,000 a year endorsed greater difficulties regulating positive emotions. Further, age was significantly and negatively related to difficulties regulating positive emotions (r = −.29, p < .001). Alcohol misuse appeared to differ only as a function of gender, with men reporting greater alcohol misuse than women. No significant differences were found between groups with regards to drug misuse.

Table 2.

Sociodemographic Differences in Variables of Interest

Scale	Groups	n	M	SD	Test Statistic
Gender (Male versus Female)
DERS-P	Male	154	21.37	11.42	t(361) = 3.61, p = .001, d = 0.35
	Female	209	17.80	9.12
AUDIT-C	Male	160	3.63	2.72	t(371) = 3.81, p < .001, d = 0.40
	Female	213	2.63	2.32
DAST-10	Male	145	1.19	1.71	t(338) = −0.39, p = .70, d = 0.05
	Female	195	1.28	2.15
Ethnicity (Hispanic versus non-Hispanic)
DERS-P	Hispanic	44	22.61	12.62	t(361) = −2.28, p = .02, d = 0.33
	Non-Hispanic	319	18.86	9.87
AUDIT-C	Hispanic	46	2.80	1.93	t(371) = 0.73, p = .47, d = 0.13
	Non-Hispanic	327	3.10	2.62
DAST-10	Hispanic	39	1.79	2.46	t(338) = −1.87, p = .06, d = 0.28
	Non-Hispanic	301	1.17	1.89
Race (White versus non-White)
DERS-P	White	279	18.41	9.41	t(361) = 3.09, p = .002, d = 0.36
	Non-White	84	22.32	12.40
AUDIT-C	White	283	3.03	2.55	t(371) = 0.45, p = .65, d = 0.06
	Non-White	90	3.17	2.54
DAST-10	White	259	1.22	1.92	t(338) = 0.29, p = .77, d = 0.04
	Non-White	81	1.30	2.15
Income (<$50,000 versus ≥$50,000)
DERS-P	<$50,000	182	20.75	11.14	t(361) = 2.69, p = .008, d = 0.28
	≥$50,000	181	17.87	9.18
AUDIT-C	<$50,000	191	3.08	2.69	t(371) = 0.17, p = .86, d = 0.02
	≥$50,000	182	3.04	2.40
DAST-10	<$50,000	174	1.32	2.01	t(338) = 0.72, p = .47, d = 0.08
	≥$50,000	166	1.16	1.94
Education (≤High School degree versus >High School degree)
DERS-P	≤High School degree	49	16.98	8.59	t(361) = −1.71, p = .09, d = 0.28
	>High School degree	314	19.68	10.50
AUDIT-C	≤High School degree	50	2.62	2.79	t(371) = −1.32, p = .19, d = 0.19
	>High School degree	323	3.13	2.51
DAST-10	≤High School degree	48	1.40	1.87	t(338) = 0.59, p = .56, d = 0.09
	>High School degree	292	1.22	1.99

Note: DERS-P = Difficulties in Emotion Regulation Scale – Positive, AUDIT-C = Alcohol Use Disorder Identification Test – Consumption Questions, DAST-10 = Drug Abuse Screening Test

Correlations between Difficulties Regulating Positive Emotions and Alcohol and Drug Misuse

See Table 3 for bivariate relations among difficulties regulating positive emotions and alcohol and drug misuse in the overall sample and across sociodemographic groups. Correlations in the overall sample showed that difficulties regulating positive emotions were significantly and positively associated with both alcohol and drug misuse. These relations held across racial and income groups. However, among women, individuals identifying as Hispanic, and individuals with a high school degree or less, the association between difficulties regulating positive emotions and alcohol misuse was nonsignificant.

Table 3.

Intercorrelations Among Variables of Interest

	DERS-P → AUDIT-C	DERS-P → DAST-10
Overall	.22***	.30***
Gender (Male versus Female)
Male	.28**	.18*
Female	.10	.42***
Ethnicity (Hispanic versus non-Hispanic)
Hispanic	.27	.33*
Non-Hispanic	.22***	.29***
Race (White versus non-White)
White	.18**	.29***
Non-White	.32**	.34**
Income (<$50,000 versus ≥$50,000)
<$50,000	.26***	.24**
≥$50,000	.16*	.36***
Education (≤High School degree versus >High School degree)
≤High School degree	.20	.35*
>High School degree	.21***	.30***

Note: DERS-P = Difficulties in Emotion Regulation Scale – Positive, AUDIT-C = Alcohol Use Disorder Identification Test – Consumption Questions, DAST-10 = Drug Abuse Screening Test;

^*p < .05,

^**p < .01,

^***p <.001

Moderation Analyses Examining Sociodemographic Differences in the Relation between Difficulties Regulating Positive Emotions and Alcohol Misuse

A set of moderation analyses examined the main and interactive effects of sociodemographic characteristics (i.e., age, gender, ethnicity, race, income, and educational attainment) and difficulties regulating positive emotions on alcohol misuse. In each model, significant main effects were detected for difficulties regulating positive emotions with alcohol misuse, bs ranging from .04 to .07, ps ranging from <.001 to .001. Age, (b = 0.03, p = .03) and gender, (b = −0.77, p = .004) also demonstrated significant main effects. No significant interaction effects were identified. See Table 4 for a summary of main and interactive effects regarding alcohol misuse.

Table 4.

Main and interactive effects of difficulties regulating positive emotions and sociodemographic variables on alcohol misuse

Construct	b	SE	T	p	95% CI
Age
DERS-P	.07	.02	4.47	<.001	[.04, .10]
Age	.03	.01	2.16	.03	[.01, .05]
DERS-P X Age	.002	.002	1.20	.23	[−.001, .01]
Gender (Male versus Female)
DERS-P	.04	.01	3.24	.001	[.02, .07]
Gender	−.77	.27	−2.90	.004	[−1.29, −.25]
DERS-P x Gender	−.04	.03	−1.60	.11	[−.09, .01]
Ethnicity (Hispanic versus non-Hispanic)
DERS-P	.06	.01	4.35	<.001	[.03, .08]
Ethnicity	−.50	.41	−1.22	.23	[−1.32, .31]
DERS-P x Ethnicity	−.02	.03	−0.54	.59	[−.08, .05]
Race (White versus non-White)
DERS-P	.05	.01	3.93	<.001	[.03, .08]
Race	.14	.32	0.42	.67	[−.49, .76]
DERS-P x Race	−.02	.03	−0.60	.55	[−.07, .04]
Income (<$50,000 versus ≥$50,000)
DERS-P	.05	.01	3.98	<.001	[.03, .08]
Income	−.16	.27	0.61	.54	[−.36, .68]
DERS-P x Income	−.02	.03	−0.89	.38	[−.08, .03]
Education (≤High School degree versus >High School degree)
DERS-P	.05	.01	4.08	<.001	[.03, .08]
Education	.31	.40	0.79	.43	[−.47, 1.09]
DERS-P x Education	−.01	.04	−0.29	.77	[−.10, .07]

Note: Bolded typeface indicates signficance at the level p<.05; DERS-P = Difficulties in Emotion Regulation Scale - Positive

Moderation Analyses Examining Sociodemographic Differences in the Relations between Difficulties Regulating Positive Emotions and Drug Misuse

A second set of moderation analyses examined the main and interactive effects of sociodemographic characteristics (i.e., age, gender, ethnicity, race, income, and educational attainment) and difficulties regulating positive emotions on drug misuse. In all but one model, significant main effects were detected for difficulties regulating positive emotions on drug misuse, bs ranging from .03 to .08, ps ranging from <.001 to .05. A significant main effect was detected for gender, (b = −1.26, p = .006), and the interaction between gender and difficulties regulating positive emotions was significant, (b = 0.08, p = <.001). ¹ As illustrated in Figure 1, analysis of simple slopes revealed that difficulties regulating positive emotions was significantly positively associated with drug misuse when participants were female, (b = 0.11, SE = 0.02, t = 6.79, p < .001, 95% CI [0.08, 0.14]), but not when participants were male, (b = .03, SE = .01, t = 1.94, p = .05, 95% CI [−0.0003, 0.06]). See Table 5 for a summary of main and interactive effects regarding drug misuse.

Figure 1. Difficulties Regulating Positive Emotions by Gender Interaction for Drug Misuse.

Note: DERS-P = Difficulties in Emotion Regulation Scale - Positive

Table 5.

Main and interactive effects of difficulties regulating positive emotions and sociodemographic variables on drug misuse

Construct	b	SE	t	p	95% CI
Age
DERS-P	.06	.01	4.53	<.001	[.03, .08]
Age	−.003	.01	−0.33	.74	[−025, .02]
DERS-P X Age	−.001	.001	−0.46	.65	[−.003, .002]
Gender (Male versus Female)
DERS-P	.03	.01	1.94	.05	[−.0003, .06]
Gender	−1.26	.45	−2.78	.006	[−2.15, −.37]
DERS-P x Gender	.08	.02	3.77	<.001	[.04, .12]
Ethnicity (Hispanic versus non-Hispanic)
DERS-P	.06	.01	5.04	<.001	[.04, .08]
Ethnicity	−.09	.70	0.13	.90	[−1.29, 1.47]
DERS-P x Ethnicity	.01	.03	0.47	.64	[−.05, .07]
Race (White versus non-White)
DERS-P	.06	.02	3.28	.001	[.02, .10]
Race	.19	.52	0.37	.71	[−.84, 1.22]
DERS-P x Race	.001	.02	0.02	.98	[−.04, .05]
Income (<$50,000 versus ≥$50,000)
DERS-P	.05	.01	3.33	.001	[.02, .08]
Income	−.56	.45	−1.24	.22	[−1.45, .33]
DERS-P x Income	.03	.02	1.35	.18	[−.01, .07]
Education (≤High School degree versus >High School degree)
DERS-P	.08	.03	2.37	.02	[.01, .14]
Education	−.10	.65	−0.15	.88	[−1.38, 1.19]
DERS-P x Education	−.01	.03	−0.43	.67	[−.08, .05]

Note: Bolded typeface indicates signficance at the level p<.05; DERS-P = Difficulties in Emotion Regulation Scale – Positive

Discussion

The goal of the current study was to examine potential sociodemographic moderators (i.e., age, gender, ethnicity, race, income, and educational attainment) of the relations between difficulties regulating positive emotions and alcohol and drug misuse. Significant differences in difficulties regulating positive emotions were identified across sociodemographic groups, with males reporting greater difficulties than females, Hispanic individuals reporting greater difficulties than non-Hispanic individuals, non-White individuals reporting greater difficulties than White individuals, and individuals making less than $50,000/year reporting greater difficulties than those making more than $50,000/year. Regarding substance misuse, male participants reported greater alcohol misuse than female participants, and no significant differences in drug misuse were detected across demographic groups. As hypothesized, greater difficulties regulating positive emotions was associated with higher levels of both alcohol and drug misuse. Finally, while there were no significant differences between males and females on drug misuse, gender was shown to moderate the link between difficulties regulating positive emotions and drug misuse.

We identified significant differences in difficulties regulating emotions across sociodemographic variables, specifically gender, ethnicity, race, and income; there is some support in previous literature to explain these differences. Regarding gender, consistent with our findings, previous work has also found men to report greater difficulties regulating positive emotions than do females (Weiss et al., 2019; Weiss et al., 2015b). It may be that males are more socialized to control or restrict their expression of positive emotions, and more masculine gender roles (typically associated with males) has been linked to greater fear of positive emotions (Jakupcak, Salters, Gratz, & Roemer, 2003); this may lead to greater nonacceptance of positive emotions. Further, males have been found to display greater impulse control difficulties than do females (Cross, Copping, & Campbell, 2011), which may extend to difficulties controlling impulsive behaviors in the context of positive emotions. Regarding race and ethnicity, our findings are consistent with prior research. For example, Latinx individuals have been found to report more suppression or avoidance of positive emotions compared to White individuals (Butler et al., 2003; Gross & John, 2003). Finally, regarding study findings on income, perhaps individuals with greater disposable income and higher socioeconomic status may experience less psychological distress as well as better coping strategies (Baum, Garofalo &Yali, 1999; Dohrenwend & Dohrenwend, 1969; Roohafza et al., 2009); these in turn may contribute to greater emotion regulation abilities (Côté, Gyurak, & Levenson, 2010).

Regarding the finding that gender was found to moderate the link between difficulties regulating positive emotions and drug misuse, this association was found to be significant for females but not for males. Prior literature has suggested that women are more likely than men to endorse methods of managing emotionality, including emotional avoidance (Nolen-Hoeksema, 2012). It may be that women experience greater affectivity, and therefore are more likely to use drugs to suppress or avoid emotions that they experience as aversive or threatening. Conversely, while males reported greater difficulties regulating positive emotions than did females, the association between these difficulties and drug misuse was nonsignificant. It may be that these difficulties are associated with different outcomes for males that are not reflected in the present study. Future work is needed to elucidate the consequences of difficulties regulating positive emotions for males. These findings suggest the importance of assessing for difficulties regulating positive emotions among women to aid in prevention of drug misuse, and for women who are referred for treatment related to drug misuse, as increasing skills to regulate positive emotions may serve to reduce levels of drug misuse.

On the other hand, while there was a significant main effect of difficulties regulating positive emotions on alcohol misuse, no sociodemographic characteristic emerged as significant moderators of this association. This indicates that the association between difficulties regulating positive emotions and alcohol misuse was significant and comparable in both strength and direction across sociodemographic groups. Thus, it may be clinically relevant to assess and treat difficulties regulating positive emotions among all individuals receiving treatment for problematic alcohol use broadly. For instance, dialectical behavior therapy (DBT) has been found effective in decreasing substance use for individuals with co-occurring borderline personality disorder, and has been proposed as a potential tool for treating individuals without co-occurring borderline personality disorder (Dimeff & Linehan, 2008), though little work has been done evaluating the utility of DBT in reducing substance use for other populations (Dimeff & Linehan, 2008). To date, one pilot trial has been conducted examining the feasibility of DBT skills training for individuals with alcohol dependence (both with and without comorbid personality disorder), finding a significant association between improved emotion regulation and alcohol-related outcomes (Maffei, Cavicchioli, Movalli, Cavallaro, & Fossati, 2018), which demonstrates promise for its applications to substance misuse. Indeed, one of the core components of DBT is emotion regulation, which includes skills targeting acceptance of emotions and behavioral control (e.g., impulsivity, goal-directed behavior) in emotional contexts (Linehan, 1993). Research is needed to explore the utility of treatments such as DBT in reducing substance use through an improvement in difficulties regulating positive emotions.

While findings of the present study add to research on the relation between difficulties regulating emotions and substance use, they should be considered within the context of the study’s limitations, explicated further below. First, the cross-sectional, correlational nature of the data presented here precludes the examination of temporal ordering and directionality of relations among difficulties regulating positive emotions and substance use variables; it is possible that certain sociodemographic characteristics may influence the ordering of these associations. Future studies should address this concern through prospective, longitudinal investigations. Further, these findings are based on self-report measures of difficulties regulating positive emotions and alcohol and drug misuse. It may be that individuals have poor insight into their emotional experiences or that retrospective reports of substance use were over- or under-inflated. Previous work has suggested that those with difficulties regulating emotions and those who misuse alcohol and other drugs may have increased difficulty reporting accurately on their internal states (Fox et al., 2008; Tull, Bornovalova, Patterson, Hopko, & Lejuez, 2008). Thus, future studies should incorporate the use of objective behavioral and physiological measures of emotion regulation (Gratz, Rosenthal, Tull, Lejuez, & Gunderson, 2006).

Second, the AUDIT-C and DAST represent qualitatively different forms of substance use assessment in that the AUDIT-C is a measure of alcohol consumption, while the DAST is a measure of drug-related problems. This difference may have contributed to the fact that a significant interaction was detected for drug misuse only. Thus, future studies may benefit from examining these associations using the full AUDIT (Saunders, Aasland, Babor, De la Fuente, & Grant, 1993), which includes questions assessing alcohol-related problems, or using a measure of drug use specifically (e.g., Drug Use Questionnaire; Hien & First, 1991).

Additionally, the conceptualization of the measure of difficulties regulating positive emotions (Gratz & Roemer, 2004; Gratz & Tull, 2010; Weiss et al., 2015a) used in the present study differs from other frameworks of emotion regulation, such as those that a) equate emotion regulation with temperamental characteristics of emotional intensity or reactivity (Livesley, Jang, & Vernon, 1998), b) define emotion regulation as control of emotions and reduction in emotional arousal (Zeman & Garber, 1996), and c) classify specific emotion regulation strategies as either adaptive or maladaptive (Butler et al., 2003; Gross, 1998). Future studies would benefit from exploring the role of sociodemographic characteristics in the associations among various aspects of emotion regulation in relation to substance use behaviors.

Finally, it will be important for future work to be conducted to continue to examine these findings. For instance, while the MTurk recruitment platform is a notable strength of our study (Buhrmester et al., 2011; Mischra & Carleton, 2017; Shapiro et al., 2013), collecting data via the internet using an online format has disadvantages that may limit generalizability of results, such as sample biases (e.g., because of self-selection) and lack of control over the research environment (e.g., no opportunity to clarify questions, distractions; Kraut et al., 2004). Thus, future research that integrates other data collection methods (e.g., interviewing, focus groups) is warranted. Further, the present sample was entirely comprised of trauma-exposed individuals, which may also limit generalizability of these findings. Thus, it will be important for future research to continue to examine these associations among individuals who have not experienced a traumatic event. It is also worth noting that, due to the larger sample size required to adequately power studies aiming to examine interaction effects, findings such as those reported in the present study may be more difficult to replicate (McClelland & Judd, 1993).

Despite these limitations, results of the current study extend our understanding of the role of sociodemographic characteristics in difficulties regulating positive emotions, substance misuse, and their association. While preliminary, our findings provide support for gender differences in the association between difficulties regulating positive emotions and drug misuse, such that this relation was found to be significant for female (but not for male) participants. These results may inform the development of gender-sensitive recommendations for the assessment and treatment of substance misuse, including those targeting difficulties regulating positive emotions.