Emotion regulation in substance use disorders: a systematic review and meta‐analysis

Jordan Stellern; Ke Bin Xiao; Erin Grennell; Marcos Sanches; Joshua L Gowin; Matthew E Sloan

doi:10.1111/add.16001

. 2022 Aug 11;118(1):30–47. doi: 10.1111/add.16001

Emotion regulation in substance use disorders: a systematic review and meta‐analysis

Jordan Stellern ¹, Ke Bin Xiao ^2,³, Erin Grennell ², Marcos Sanches ⁴, Joshua L Gowin ^1,^✉, Matthew E Sloan ^2,^3,^5,^6,^7,^✉

PMCID: PMC10087816 PMID: 35851975

Abstract

Background and aims

The ability to regulate emotions effectively has been associated with resilience to psychopathology. Individuals with substance use disorders (SUDs) have been shown to have higher levels of negative emotionality, with some evidence suggesting impairment in emotion regulation compared with individuals without SUDs. However, no previous attempt has been made to systematically review the literature to assess the magnitude of this difference. We aimed to assess the association between SUD diagnosis and emotion regulation as measured by the Difficulties in Emotion Regulation Scale (DERS) and Emotion Regulation Questionnaire (ERQ) through a systematic review and meta‐analysis of existing findings.

Methods

The systematic review was conducted using PubMed, PsycINFO and Embase. We examined cross‐sectional studies that compared a SUD group with a control group and measured emotion regulation using the DERS or the ERQ. The primary analysis focused on papers using the DERS, as this was the predominant instrument in the literature.

Results

Twenty‐two studies met our primary analysis criteria, representing 1936 individuals with a SUD and 1567 controls. Individuals with SUDs relative to controls had significantly greater DERS scores, with a mean difference of 21.44 [95% confidence interval (CI) = 16.49–26.40, P < 0.001] and Hedges’ g = 1.05 (95% CI = 0.86–1.24, P < 0.001). The difference was robust, remaining significant after removing outliers and studies with high risk of bias. Individuals with SUDs demonstrated poorer emotion regulation on each subscale of the DERS, with the largest deficits in the Strategies and Impulse subscales. The ERQ analysis revealed greater use of expressive suppression in those with SUDs relative to controls (Hedges’ g = 0.76, 95% CI = 0.25–1.28, P = 0.004).

Conclusions

People with substance use disorders appear to have greater difficulties in emotion regulation than people without substance use disorders.

Keywords: Alcohol use disorder, cocaine addiction, emotion regulation, methamphetamine, opioid use disorder, substance use disorders

INTRODUCTION

Recent estimates suggest that 5.1% of the global population have an alcohol use disorder (AUD) [1] and approximately 35 million people world‐wide suffer from drug use disorders [2]. The etiology and correlates of these widespread disorders remain incompletely understood. Some evidence suggests that negative emotionality may be related to the development and maintenance of addictive behavior. For example, adolescents with depressive symptoms display higher levels of alcohol consumption 3 months later and may be more likely to develop frequent binge drinking patterns in young adulthood [3, 4]. Conversely, there is evidence that individuals with substance use disorders (SUDs) are more than twice as likely to develop mood disorders than those without SUDs [5]. This increased liability to mood disorders may be due to chronic drug‐induced alterations in the brain’s stress‐ and emotion‐related circuits [6]. Individuals with SUDs also demonstrate a reduced ability to regulate negative moods compared to healthy adults [7]. Developing a clearer understanding of impairments in emotion regulation in individuals with addiction may improve our understanding of the etiology and treatment of SUDs.

Emotion regulation refers to any process or action by which an individual influences their emotions or their emotional expression [8]. An individual can regulate emotions at multiple points, including the situations that they seek or avoid, how they think about their experience and how they express their feelings. Some forms of regulation are associated with greater wellbeing, such as cognitive reappraisal, mindfulness and acceptance [9, 10], whereas other strategies, such as suppression, are associated with poorer psychological outcomes [9, 11] Emotion regulation difficulties have been proposed as a component of clinical disorders and are a primary target of a form of psychotherapy known as dialectical behavior therapy [12]. According to this framework, when an individual experiences an intense emotion the arousal may need an outlet, especially if the person struggles to diminish the emotion’s intensity [12]. Some may then choose to use substances to regulate the negative emotion [12]. These theories may inform SUD treatment. For example, dialectical behavior therapy skills training not only improves emotion regulation but also increases rates of abstinence and decreases substance use severity in individuals with AUD [13]. Poor emotion regulation may therefore be common in those with SUDs and may represent a viable target for treatment. However, to date, no study has attempted to estimate the magnitude of the difference in emotion regulation capacity between those with and without SUDs by reviewing the existing literature.

To determine whether adults with SUDs differ from healthy adults in their emotion regulation capacity we reviewed cross‐sectional studies that evaluated emotion regulation using two validated and commonly used self‐report questionnaires: the Difficulties in Emotion Regulation Scale (DERS) [14] and the Emotion Regulation Questionnaire (ERQ) [11]. We hypothesized that individuals with SUDs would have significantly more difficulties with emotion regulation overall than those without SUDs. As a secondary aim, we explored differences in the subscales of the DERS to determine if specific elements of emotion regulation (e.g. awareness of emotions, acceptance of emotions, ability to use effective strategies to regulate emotions) showed greater differentiation between individuals with and without SUDs. Lastly, we explored differences in emotion regulation between individuals with specific SUDs (e.g. alcohol use disorder, opioid use disorder) and individuals without SUDs.

METHODS

This systematic review and meta‐analysis was performed and reported according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta Analyses) 2020 statement and checklist [15]. The study was registered on Prospero on 18 July 2021 (ID #251811, https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=251811).

Information sources

An initial literature search was conducted through the electronic databases PubMed and PsycINFO from inception to 1 May 2021. This initial search strategy was aimed primarily at finding studies that utilized the Difficulties in Emotion Regulation Scale (DERS), which was the original goal of our review. An additional literature search was conducted on 23 February 2022 through these databases as well as Embase to expand our review to include studies using the Emotion Regulation Questionnaire (ERQ).

Search strategy

Studies were identified using a range of search terms in both literature searches, including: ‘emotion regulation’, ‘Difficulties Emotion Regulation Scale’, ‘DERS’, ‘Emotion Regulation Questionnaire (ERQ)’, ‘substance use disorder’, ‘drug addiction’, ‘drug dependence’, ‘drug abuse’, ‘alcohol’, ‘cocaine’, ‘opioid’, ‘methamphetamine’, ‘nicotine’, ‘cannabis’, ‘case–control studies’, ‘matched controls’ and ‘healthy controls’. Details of our search strategy are available in Supporting information, Appendix S1.2–S1.6.

Eligibility criteria

Our original analysis focused on the DERS, as we found this to be the most used measure of emotion regulation in populations with SUDs. Inclusion criteria for study selection included: (1) human studies with full text available in English; (2) administration of the full version of the DERS (36‐item questionnaire) to both the SUD group and the control group and (3) reporting means and standard deviations (SDs) of the total DERS score or all subscale scores for both groups (studies were also included if total scores were not reported but were obtained by e‐mailing the authors). The SUD group needed to be composed of participants either enrolled into a treatment or recovery program for SUD or diagnosed with SUD, substance dependence or substance abuse using clinically structured tools or diagnostic interviews [e.g. the Structured Clinical Interview for DSM‐IV (SCID)]. Studies were excluded if only screening tools were used to show evidence for a disorder [e.g. the Alcohol Use Disorder Identification Test (AUDIT)]. Studies were also excluded if the entire SUD group was diagnosed with a comorbid psychiatric disorder. For example, one study’s sample consisted entirely of individuals with borderline personality disorder with or without a comorbid SUD [16]. In another study, we excluded one of the SUD groups in which all individuals were experiencing methamphetamine‐induced psychosis [17]. To be included, the study’s control group had to be composed of participants without a known SUD diagnosis and who were not recruited from a psychiatric treatment facility. The control group could have no more than 20% of its members diagnosed with prevalent psychiatric disorders, such as a mood or anxiety disorder, if these data were reported. If a study contained more than two groups (e.g. a SUD group, a borderline personality disorder group and a control group), data were only extracted from the SUD group and the control group. Additional details regarding the included studies can be found in Supporting information, Appendix S1.1. Our secondary analysis used additional emotion regulation measures, including the ERQ and versions of the DERS that were shortened or adapted (several studies only used a subset of the DERS subscales). Two other self‐report measures of emotion regulation were identified in our review, the Emotion Regulation Interview and Emotion Regulation Skills Questionnaire, but each was only used in one study, so no meta‐analysis of these scales was conducted.

The DERS

The DERS is a 36‐item self‐report measure that aims to assess emotional dysregulation using six subscales [14]: non‐acceptance of negative emotions (Non‐Acceptance), difficulties engaging in goal‐directed behaviors when distressed (Goals), belief that there is little that one can do to regulate emotions effectively (Strategies), difficulties controlling impulsive behaviors when distressed (Impulse), lack of emotional awareness (Awareness) and lack of emotional clarity (Clarity). Each item is rated on a scale from 1 (almost never) to 5 (almost always). The total DERS score is calculated from the sum of all six subscales, resulting in a score range of 36 to 180. Higher scores suggest more difficulties in emotion regulation. DERS scores have shown good internal consistency with Cronbach’s alphas ranging from 0.77 to 0.96 in the studies we reviewed. Subscales have shown adequate internal consistency with Cronbach’s α > 0.70 for all subscales. DERS scores have also been shown to have good test–retest reliability (ρI = 0.88, P < 0.01) [14]. The DERS has also been reported to have adequate predictive and constructive validity [14], with scores being correlated with depression and anxiety symptoms [18]. Studies have shown no difference in DERS scores by sex [19] or race [20], but age has been associated with DERS score in some studies, with younger individuals tending to have higher scores [21].

The ERQ

The ERQ [11] is a 10‐item self‐report measure that aims to measure two factors related to emotion regulation: cognitive reappraisal, which is assessed through six items, and expressive suppression, which is assessed through four items. Each item is rated on a scale from 1 (strongly disagree) to 7 (strongly agree), resulting in a score range of 6–42 for cognitive reappraisal and 4–28 for expressive suppression. Higher scores indicate more frequent use of reappraisal or suppression. Alpha reliabilities averaged 0.79 for cognitive reappraisal and 0.73 for expressive suppression [11]. Test–retest reliability throughout 3 months was 0.69 for both cognitive reappraisal and expressive suppression [11]. Men scored significantly higher than women on expressive suppression, but there were no gender differences with regard to cognitive reappraisal [11].

Selection process

Two independent reviewers (J.S. and K.X.) selected the articles by screening the abstracts against the eligibility criteria using Covidence [22], a web‐based software. Articles that were missing an abstract or did not specify the type of emotion regulation tools used in the study were screened by full‐text review. Duplicate papers were removed. Discrepancies in study selection were resolved by a third author (E.G.). A consensus was reached for remaining disagreements by consulting the rest of the study team.

Data collection process

The data extraction was conducted by two raters (J.S. and E.G.) and the inter‐rater reliability was ICC = 1 for DERS scores in the control and SUD arms, indicating excellent reliability. After extraction, a third author (K.X.) reviewed independent data extractions, highlighted any discrepancies and corrected them. Data on ERQ and shortened versions of the DERS were extracted by two authors (K.X. and E.G.) and the inter‐rater reliability was ICC = 1 for scores in the control and SUD arms, again indicating excellent reliability. Fifty‐four discrepancies were found in the first extraction (DERS) and 10 discrepancies found in the second extraction (ERQ and shortened versions of the DERS). The former discrepancies were corrected by the third author, and the latter were corrected after a discussion between the authors. Finally, data were independently extracted by a senior author (M.E.S.) and any residual errors were corrected after consultation with the study team.

Data extraction

Data extracted from eligible papers included sample size, age, sex, diagnostic tool used, diagnosis of psychiatric comorbidities, recruitment source, years of education, geographical location of study and year of publication (see Table 1 for DERS and Supporting information, Table S2 for ERQ and shortened versions of the DERS). For all the questionnaires, mean and SDs for all subscale scores were extracted. For papers using the full DERS, we also extracted the mean and SDs for the total score. If values were missing from the paper, we first attempted to obtain the values by contacting the paper’s authors; if the authors did not respond, we estimated values based on graphs in the paper (n = 2). Three studies provided DERS subscale scores but not total scores, so means were calculated, and standard deviation (SD) was imputed. Sensitivity analyses were conducted which excluded these studies to ensure that they were not biasing our results.

TABLE 1.

Study characteristics.

Author [year]	Country	Groups	n	Age (mean SD)	Control groups, matched by	Component of DERS	DERS Score (mean, SD)	Recruitment source	Diagnostic tool	NIH quality assessment
Azizi et al. [49] [2017]	Iran	Opioid dependence	88 (0F/88M)	30.90 (4.14)	Age, gender, education level	All six subscale scores	121.18 (7.77)	Outpatient substance abuse treatment center	ND	Fair
Azizi et al. [49] [2017]	Iran	Control ¹	40 (0F/40M)	29.32 (3.89)			102.62 (8.83)	Patients’ relatives/friends	ND	Fair
Bottesi et al. [50] [2021]	Italy	SUD	131 (56F/75M)	39.74 (11.83)	Sex	All six subscale scores	87.42 (13.31)	Patients from community mental health centers and inpatient mental health clinics	DSM‐5 (but no structured diagnostic tool mentioned)	Fair
Bottesi et al. [50] [2021]	Italy	Control	131 (56F/75M)	40.02 (12.34)			58.97 (9.35)	Engaged through advertisements	DSM‐5 (but no structured diagnostic tool mentioned)	Fair
Buckholdt et al. [33] [2015]	United States	Substance abuse	81 (39F/42M)	36.6 (11.4)	ND	Total score, all six subscales	92.6 (24.9)	Patients in residential substance abuse treatment facility	ND	Poor
Buckholdt et al. [33] [2015]	United States	Control	118 (90F/28M) ^c	20.9 (4.8)			78.9 (22.6)	Undergraduate university students	ND	Poor
Di Pierro et al. [51] [2015]	Italy	SUD	58 (16F/42M)	35.14 (9.31)	ND	Total score, all six subscales ^a	82.96 (12.33)	Patients in a residential treatment service for patients with SUDs	SCID‐II for DSM‐IV	Fair
Di Pierro et al. [51] [2015]	Italy	Control	73 (53F/20M)	28.85 (5.95)			77.41 (10.53)	Community participants recruited through fliers and word of mouth	SCID‐II for DSM‐IV	Fair
Dingle et al. [52] [2018]	Australia	SUD	35 (13F/22M)	25 (4.48)	Age, gender, ethnicity, education level	Total score, all six subscales	108.43 (25.03)	Residents in a drug and alcohol therapeutic community	SCID for DSM‐IV administered to SUD group to confirm diagnosis	Fair
Dingle et al. [52] [2018]	Australia	Control	35 (13F/22M)	25 (4.34)	Age, gender, ethnicity, education level	Total score, all six subscales	66.69 (16.46)	Recruited from social media/word of mouth	Controls screened for alcohol misuse using AUDIT‐C	Fair
Fox et al. [53] [2007]	United States	Cocaine dependence	60 (28F/32M)	37.2 (5.9)	ND	Total score, four subscale scores reported and 2 estimated from graph (non‐acceptance and goals) ^b	85.8 (22.5)	Cocaine‐dependent patients and healthy volunteers recruited through local advertisement	SUD group met DSM‐IV criteria for cocaine dependence (but no diagnostic tool mentioned) and tested positive for cocaine in urine toxicology screen	Good
Fox et al. [53] [2007]	United States	Control	50 (24F/26M)	33.6 (10)			60.9 (15.0)			Good
Fox et al. [54] [2008]	United States	Alcohol dependence	50 (9F/41M)	37.5 (8.2)	ND	Total scores and all six subscale scores estimated from graphs ^b	ND	Treatment‐seeking alcohol‐dependent participants and social drinkers recruited through local advertisements	Met DSM‐IV criteria (but no diagnostic tool mentioned) and tested positive on a urine toxicology screen	Good
Fox et al. [54] [2008]	United States	Control	62 (32F/30M)	33.7 (9.4)	ND		ND			Good
Garke et al. [27] [2021]	United States	SUD	415 (180F/235M)	35.13 (10.21)	ND	Total score, all six subscales	88.67 (24.96)	The SUD groups were recruited from two state‐run residential treatment centers in Mississippi. For one of the two residential SUD groups, an alcohol and/or cocaine use disorder diagnosis was required The community sample was recruited through social media advertisements, flyers, and newspaper advertisements in the Greater New Haven, CT area	SCID‐I for DSM‐IV	Good
Garke et al. [27] [2021]	United States	Control ²	225 (134F/91M)	32.26 (11.37)			65.99 (19.44)		SCID‐I for DSM‐IV	Good
Ghorbani et al. [55] [2017] ³	Iran	Alcohol dependence	205 (51F/154M)	32.36 (10.26)	ND	Total score	115.17 (29.81)	Outpatient addiction treatment clinic in Karaj, Iran	AD group: DSM‐IV TR diagnosis of alcohol dependence	Good
Ghorbani et al. [55] [2017] ³	Iran	Control	100 (40F/60M)	30.32 (9.98)			72.35 (20.68)	Control group included undergraduate students from Tehran University	Controls: SCID‐I for DSM‐IV‐TR	Good
Hardy et al. [56] [2018]	United States	SUD	70 (70F/0M)	42.5 (11.8)	Sex	Total score and all subscale scores ^a	ND	Waiting rooms in diabetes, gynecology and primary care clinics in a public hospital in Atlanta, Georgia	MINI for DSM‐IV	Good
Hardy et al. [56] [2018]	United States	Control ⁴	117 (117F/0M)	37.9 (11.8)	Sex	Total score and all subscale scores ^a	ND		MINI for DSM‐IV	Good
Jakubczyk et al. [57] [2020]	Poland	Alcohol use disorder	165 (20F/145M)	44.0 (11.2)	ND	Total score, all six subscales	90.3 (19.3)	Recruited from inpatient alcohol treatment program	AUD group: MINI for ICD‐10	Good
Jakubczyk et al. [57] [2020]	Poland	Control	110 (28F/82M)	40.6 (8.1)			70.0 (14.5)	Recruited from individuals presenting to a general practitioner for medical advice, prophylactic examination, or infection treatment	AUD group: MINI for ICD‐10	Good
Krause‐Utz et al. [58] [2019]	Germany	SUD	28 (28F/0M)	31.14 (8.35)	Sex	Total score, all six subscales (data taken from paper’s supplement)	79.86 (20.50)	Both groups were recruited from the Central Institute of Mental Health in Mannheim and Leiden University	SCID‐I for DSM‐IV, International Personality disorder examination (IPDE), MINI, and Wender–Reimherr interview	Good
Krause‐Utz et al. [58] [2019]	Germany	Control ⁵	60 (60F/0M)	27.60 (6.84)	Sex		69.47 (13.32)			Good
London et al. [59] [2020]	United States	Methamphetamine dependence	11 (2F/9M)	38.3 (9.31)	ND	Total score	74.91 (19.83)	ND	MINI or SCID for DSM‐IV	Good
London et al. [59] [2020]	United States	Control	12 (5F/7M)	33.8 (7.48)			54.25 (15.96)	ND	MINI or SCID for DSM‐IV	Good
Okita et al. [60] [2016]	United States	Methamphetamine dependence	94 (38F/56M)	34.20 (9.7)	ND	Total score	76.2 (20.10)	Both groups recruited using internet and newspaper advertisements	SCID for DSM‐IV	Good
Okita et al. [60] [2016]	United States	Control	102 (54F/48M)	33.70 (8.94)			55.2 (12.53)		SCID for DSM‐IV	Good
Pelot et al. [61] [2020]	Canada	SUD	20 (13F/7M)	30.85 (ND)	Age, gender	Total score	90.30 (26.5)	Male SUD participants were recruited from an outpatient addiction program and female SUD participants were recruited from an inpatient addiction program	ND	Fair
Pelot et al. [61] [2020]	Canada	Control	20 (13F/7M)	30.85 (ND)			67.90 (11.4)	HCs recruited from the community	ND	Fair
Rettie et al. [62] [2018]	United Kingdom	Alcohol use disorder	45 (13F/32M)	44.83 (9.92)	ND	Total score	116.58 (25.63)	Clients undergoing alcohol detoxification who had completed pharmacological treatment	ND	Poor
Rettie et al. [62] [2018]	United Kingdom	Control	36 (28F/8M)	44.11 (13.38)			62.22 (13.85)	Control group of staff members at the same detoxification unit	ND	Poor
Russell et al. [63] [2017]	United States	SUD ⁶	86 (42F/44M)	17.16 (0.27)	ND	Total score	102.78 (28.17)	Students attending RHS in Massachusetts	ND	Fair
Russell et al. [63] [2017]	United States	Control	28 (20F/8M)	18.71 (0.6)			76.2 (26.1)	Recruited through convenience sampling from Connecticut high schools
Russell et al. [64] [2019]	United States	SUD ⁷	27 (9F/18M)	17.26 (1.48)	ND	Total score	101.72 (27.10)	Students attending RHS	ND	Fair
		Control	29 (9F/20M)	17.55 (2.03)			81.07 (21.58)	Recruited through local high schools, public libraries, and similar community settings
Seo et al. [48] [2016]	United States	Alcohol dependence	37 (8F/29M)	37.2 (7.9)	Age, intelligence, handedness and gender	Impulse subscale score	ND	Patients at inpatient treatment facility	SCID for DSM‐IV	Fair
		Alcohol dependence	37 (8F/29M)	37.2 (7.9)		Total scores obtained by e‐mailing authors ^a	ND	Recruited via community newspaper, flyers and web advertising
		Control	37 (14F/23M)	34.3 (8.6)		Total scores obtained by e‐mailing authors ^a	ND
Stover et al. [47] [2013]	United States	Substance abuse (alcohol, cocaine, or marijuana)	43 (0F/43M)	34.58 (9.93)	Sex	Total score	98.21 (14.38)	Recruited from substance abuse treatment programs and other community settings	Met DSM‐IV criteria, but no diagnostic tool mentioned	Fair
Stover et al. [47] [2013]	United States	Control	43 (0F/43M)	34.79 (8.5)			88.88 (10.88)	Recruited from same community	Met DSM‐IV criteria, but no diagnostic tool mentioned	Fair
Uhlmann et al. [17] [2016]	South Africa	Methamphetamine dependence	21 (4F/17M)	25 (ND)	Age, gender, handedness	Total score, all six subscales	82.71 (18.04)	Participants recruited from drug rehabilitation centers, hospitals, and communities in Cape Town	SCID‐I for DSM‐IV‐TR	Good
Uhlmann et al. [17] [2016]	South Africa	Control ⁸	19 (4F/15M)	23 (ND)			67.43 (17.84)		SCID‐I for DSM‐IV‐TR	Good
Zareban et al. [65] [2018]	Iran	SUD	166	32.65 (8.22)	ND	All six subscales	105.23	Recruited from addiction rehabilitation clinics	ND	Poor
Zareban et al. [65] [2018]	Iran	Control	120 (total sample 109F/177M but within group sex composition not reported)	22 (5.79)			88.44	Recruited using convenience sampling technique	ND	Poor

Open in a new tab

SUD = substance use disorder; n = sample size; ND = no data; SD = standard deviation; M = male; F = female; NIH = National Institutes of Health; DERS = Difficulties in Emotion Regulation Scale.

RHS = recovery high school; DSM = Diagnostic and Statistical Manual of Mental Disorders; SCID = structured clinical interview for DSM; MINI = Mini‐International Neuropsychiatric Interview; BPD = borderline personality disorder; ADHD = attention deficit hyperactivity disorder.

^{^a}

Scores obtained by e‐mailing authors;

^{^b}

scores estimated from graphs;

^{^c}

discrepancies within the study were present for extracted values which are explained in Supporting information, Appendix S3.

^¹

Azizi et al. 2017 had an opium user group and a methadone maintenance treatment group that were combined.

^²

Garke et al. 2021 also had a community group with substance misuse that included participants with and without a SUD diagnosis. This group was excluded from our main analysis, but a sensitivity analysis was conducted that combined this group with the SUD group (see Supporting information, Figure S6).

^³

Ghorbani et al. 2017 administered both the DERS and ERQ. This study was included in both analyses.

^⁴

Hardy et al. 2018 also had a food addiction group (n = 42) that was not included in our analysis.

^⁵

Krause‐Utz et al. 2019 had three groups: a borderline personality disorder group, a clinical control group comprising individuals with SUD or ADHD and a control group. Data from this study were further subdivided into four groups (BPD, SUD, ADHD, HC) in the supplement, which is where we obtained the SUD and control data we extracted from this study.

^⁶

Russell et al. 2017 had three different SUD groups that were combined into one SUD group in our analysis.

^⁷

Russell et al. 2019 assessed DERS scores for the SUD group pre‐ and post‐treatment. Only the pre‐treatment data were included in our analysis.

^⁸

Uhlmann et al. 2016 also had a methamphetamine‐dependent group with a history of psychosis (n = 22) that was excluded from analysis, as all participants in this group had a comorbid diagnosis.

Risk of bias assessment

The selected papers were assessed for risk of bias using the National Heart, Lung, and Blood Institute’s quality assessment tools for observational cohort and cross‐sectional studies [23]. For details of our modified version of this tool, please see Supporting information, Appendix S2.

Effect measures

The primary measure used for the primary meta‐analysis was the mean difference in total DERS score and 95% confidence interval (CI). Additional measures included mean difference and 95% CI for DERS subscale scores, ERQ expressive suppression and cognitive reappraisal factors and scores from the shortened versions of the DERS. The standardized mean difference (SMD) was also provided to estimate the effect size for total scores and subscales. Standardization of the mean was used to estimate Hedges’ g. When moderators were included in the model we used the omnibus test QM, which is a joint test that analyzes if all the moderators’ coefficients are zero [24].

Synthesis methods

A meta‐analysis was conducted by using a random‐effect model that incorporates heterogeneity. The model specified studies as a random sample of all the studies that could have been conducted, so the result is interpreted as representing more than the studies being analyzed [25]. We assumed that heterogeneity would be present due to definition of study groups, study methodology and data collection procedures. I ² was used to estimate the proportion of the variance due to heterogeneity and the Q‐statistic was used to test for heterogeneity [26]. Prediction intervals, which represent a prediction of the possible range of effect sizes that could be found were a new study to be conducted, were reported in forest plots. Subgroup analysis (country where the study was conducted and primary drug used by the SUD group) and meta‐regression (year of publication, age of sample, sex) were used to explore heterogeneity. We used the metafor package in R version 4.0.2 [24]. Summary effect sizes were estimated using a two‐step approach, where first the heterogeneity between studies was estimated, then a weighted summary of the effect sizes was calculated, where the weight of each study was the inverse of the total variance; that is, the variance of the study plus the estimated heterogeneity. For country, European countries were grouped to achieve a reasonable sample size for subgroup analysis. Plots and descriptive statistics (tables with counts, means, SDs, maximum, minimum, number of missing values) were used for familiarization with the data, checking for outliers, associations and distribution of the data. Forest plots and funnel plots were used for displaying results and investigating evidence of publication bias and small study effect, respectively.

Six sensitivity analyses were conducted to assess the robustness of the results: (1) three studies were removed for which missing scale scores were imputed for the primary analysis; (2) three studies with effect sizes much larger than the others that fell outside the confidence region in the funnel plot (i.e. outlier studies) were removed; (3) two studies were removed where values were estimated from graphs in the paper; (4) an analysis was conducted with only 10 of 22 studies which were considered the lowest risk of bias and rated ‘good’ on the NIH quality assessment tool; (5) an analysis was conducted with only seven of 22 studies which contained a control group that was assessed with a diagnostic measure to assure ‘pure’ controls; and (6) an analysis was conducted in which an additional community group with substance misuse that included participants with and without a SUD diagnosis [27] was included (this group did not meet our inclusion criteria, but represented a large sample with substance misuse that was otherwise excluded from our analyses). Additional data synthesis and calculation methods [28] are included in Supporting information, Appendix S3.

Secondary analyses were conducted to investigate differences in emotion regulation between individuals with and without SUDs using studies that employed (1) the ERQ, (2) either shortened versions of the DERS or the full version of the DERS and (3) shortened versions of the DERS alone. Please see Supporting information, Appendix S5 for a full description of these secondary analyses [29, 30, 31].

Two studies [32, 33] had discrepancies in the values reported within their paper; these are described in Supporting information, Appendix S3.

Reporting bias assessments

Funnel plots and Egger’s test for small study effects were used to assess publication bias for the total sample and again after removing three outlier studies to assess their influence on the results.

RESULTS

Study selection

We identified 1123 papers from our original literature search and four papers from a preliminary search, resulting in a total of 1127 being screened by title and abstract. We found 36 papers that were eligible for full‐text review. Of these, 14 failed to meet inclusion criteria [16, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46]; reasons for study exclusion can be found in Supporting information, Table S1.1. We were left with 22 studies to be included for the primary analysis [17, 27, 33, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65]. A summary of the systematic review search process is shown in Figure 1 (PRISMA flow‐chart).

Our secondary literature search yielded 1897 papers. After review, no additional studies were found using the full DERS that met our criteria. Ten additional studies were found using different measures, but only eight met all other inclusion criteria: five studies using the ERQ [30, 31, 55, 66, 67] and three using shortened versions of the DERS [29, 32, 68]. We identified one study using the emotion regulation interview (ERI) [69] and another study using the emotion regulation skills questionnaire (ERSQ) [70]; these studies were not included in our analysis. One study [55] administered both the DERS and ERQ to their sample and this study was included in both of our analyses. A summary of the secondary review process is shown in Supporting information, Figure S1. We excluded 34 studies that failed to meet the inclusion criteria [16, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85]; reasons for study exclusion can be found in Supporting information, Table S1.2.

Study characteristics

Twenty‐two studies were included in our primary analysis of total DERS scores. Thirteen studies reported all six subscale scores, and one additional study reported only the Impulse subscale scores. Study characteristics are summarized in Table 1. A summary of the data by SUD type is shown in Table 2. Study characteristics of papers using the ERQ and shortened versions of the DERS are summarized in Supporting information, Table S2.

TABLE 2.

Descriptive statistics for included studies by SUD type.

	All studies (N = 22)	Unspecified substance use disorder (n = 12)	Alcohol use disorder (n = 5)	Methamphetamine use disorder(n = 3)	Opioid/cocaine use disorder (n = 2)
Sample size
Mean (SD)	159 (137)	180 (165)	169 (112)	86.3 (95.4)	119 (12.7)
Median [min, max]	113 [23.0, 640]	123 [40.0, 640]	112 [74.0, 305]	40.0 [23.0, 196]	119 [110, 128]
Age
Mean (SD)	32.0 (7.1)	29.6 (7.3)	38.0 (5.4)	31.3 (6.4)	32.7 (4.0)
Median [min, max]	32.8 [17.4, 44.5]	29.8 [17.4, 39.9]	35.8 [31.7, 44.5]	33.9 [24.1, 36.0]	32.7 [29.9, 35.6]
% Female
Mean (SD)	42.9 (25.3)	53.0 (27.9)	33.0 (12.4)	32.3 (13.7)	23.5 (33.2)
Median [min, max]	40.5 [0, 100]	51.0 [0, 100]	30.0 [17.0, 51.0]	30.0 [20.0, 47.0]	23.5 [0, 47.0]
Average DERS score, SUD group
Mean (SD)	92.1 (14.6)	93.0 (10.2)	94.0 (21.1)	77.9 (4.2)	103 (25.0)
Median [min, max]	89.5 [73.0, 121]	91.5 [78.2, 108]	90.3 [73.0, 117]	76.2 [74.9, 82.7]	103 [85.8, 121]
Average DERS score, control group
Mean (SD)	70.6 (11.9)	73.6 (9.70)	65.8 (5.0)	59.0 (7.4)	81.8 (29.5)
Median [min, max]	67.7 [54.3, 103]	72.8 [59.0, 88.9]	62.5 [61.7, 72.4]	55.2 [54.3, 67.4]	81.8 [60.9, 103]
Risk of bias, NIH scale
Fair	9 (40.9%)	7 (58.3%)	1 (20.0%)	0 (0%)	1 (50.0%)
Good	10 (45.5%)	3 (25.0%)	3 (60.0%)	3 (100%)	1 (50.0%)
Poor	3 (13.6%)	2 (16.7%)	1 (20.0%)	0 (0%)	0 (0%)

Open in a new tab

N = sample size; SUD = substance use disorder; SD = standard deviation; min = minimum; max = maximum; NIH = National Institutes of Health. Unspecified substance use disorder refers to studies that did not indicate the specific substance use disorder or included individuals with a mix of diagnoses.

Risk of bias and publication bias

The quality of DERS studies varied, with three studies rated ‘poor’, nine studies rated ‘fair’ and 10 studies rated ‘good’. For studies using the ERQ, one study was rated ‘good’, three studies were rated ‘fair’ and one was rated ‘poor’. For studies using shortened versions of the DERS, two studies were rated ‘good’ and one was rated ‘fair’. The funnel plot for our primary analysis is shown in Figure 2. Egger’s test was not significant [t = 0.83, degrees of freedom (d.f.) = 20, P = 0.42], indicating a lack of asymmetry in the funnel plot. Three studies [52, 55, 62] with effect sizes over 40 were considered outliers and were removed for a sensitivity analysis. A separate funnel plot excluding these studies is shown in Figure 3.

Funnel plot of all studies using the full Difficulties in Emotion Regulation Scale (DERS) in the meta‐analysis to assess for publication bias. Each dot represents an individual study, with the y‐axis representing the standard error of each study (lower studies have higher standard error) and the x‐axis representing the effect size of each study. The vertical line in the middle of the funnel represents the summary effect size. The diagonal funnel lines represent the area where we would expect effect sizes of each study to be. Studies outside this area can be interpreted as outliers or studies with high heterogeneity.

Funnel plot of the studies included in the meta‐analysis, with three outlier studies containing effect sizes over 40 removed. Each dot represents an individual study, with the y‐axis representing the standard error of each study (lower studies have higher standard error) and the x‐axis representing the effect size of each study. The vertical line in the middle of the funnel represents the summary effect size. The diagonal funnel lines represent the area where we would expect effect sizes of each study to be. Studies outside this area can be interpreted as outliers or studies with high heterogeneity.

DERS total score

Individuals with SUDs had a significantly higher total DERS score than controls (mean difference = 21.44, 95% CI = 16.49–26.40, P < 0.001, Figure 4). When standardizing the mean difference, the Hedges’ g‐value was 1.05 (95% CI = 0.86–1.24, P < 0.001, Supporting information, Figure S2), indicating a large effect.

Forest plot of total Difficulties in Emotion Regulation Scale (DERS) scores, depicting the mean differences between the SUD group and control group for each study as well as the summary effect. Error bars for the individual studies represent the 95% CI. The 95% CI for the summary effect is represented by the diamond width. Dashed error bars for the summary effect represent the prediction interval. The prediction interval represents a prediction of the range of possible effect sizes that could be found were a new study to be conducted. N = sample size; MD = mean difference; CI = confidence interval. *Studies with total DERS score calculated from DERS subscales. **Studies with multiple SUD arms and one control arm: multiple SUD means and SDs were pooled into one.

Heterogeneity

A high degree of heterogeneity (I ² = 92.9%) was found, indicating that the differences between the studies were unlikely due to sampling error, but from methodological differences and external factors such as drug type (Figure 5), study country (Supporting information, Figure S3.1), year of publication (Supporting information, Figure S3.2), sex distribution of the sample (Supporting information, Figure S3.3) and age of the sample (Supporting information, Figure S3.4). However, these variables did not account for the heterogeneity, as the test of moderators for each was not statistically significant (Supporting information, Appendix S4). A single regression model including all of these variables showed no effect of moderators on effect sizes [QM (d.f. = 9) = 5.18, P = 0.82), indicating that other unmeasured factors explain the heterogeneity.

Forest plot of total Difficulties in Emotion Regulation Scale (DERS) scores by substance use disorder (SUD), depicting the mean differences between the SUD group and control group for each study as well as the summary effect broken down by the specific substance use disorder investigated within the study. Error bars for the individual studies represent the 95% CI. The 95% CI for the summary effect is represented by the diamond width. Dashed error bars for the summary effect represent the prediction interval. The prediction interval represents a prediction of the range of possible effect sizes that could be found were a new study to be conducted. N = sample size; MD = mean difference; CI = confidence interval. *Studies with total DERS score calculated from DERS subscales. **Studies with multiple SUD arms and one control arm: multiple SUD means and SDs were pooled into one.

Sensitivity analyses

We conducted six sensitivity analyses (Supporting information, Figure S4.1–S4.6). When removing the three studies with imputed total DERS scores [49, 50, 65], the value of the effect size was similar (21.49, 95% CI = 15.76–27.22, P < 0.001) and heterogeneity remained high (I ² = 93.2%). After removing the three outlier studies [52, 55, 62], the effect size decreased to 17.43 (95% CI = 14.39–20.47, P < 0.001) and heterogeneity decreased (I ² = 77.72%) but remained high. When removing the two studies where values were estimated from graphs [53, 54], the effect size was 21.74 (95% CI = 16.35–27.12, P < 0.001). When only ‘good’ studies (n = 10) based on the quality assessment were included in the analysis, there was a minimal change in the effect size (20.81, 95% CI = 14.94–26.69, P < 0.001) and heterogeneity remained high (I ² = 89.5%). When only the seven studies that assessed the control group with a diagnostic measure were included, the effect size was slightly higher at 22.13 (95% CI = 14.07–30.19, P < 0.001), with heterogeneity remaining high (I ² = 91.1%). Lastly, when an additional community sample with substance misuse was included in our analysis [27], the effect size was 20.97 (95% CI = 15.96–25.99, P < 0.001) and heterogeneity remained high (I ² = 93.3%).

Secondary analyses

DERS scores by SUD type

Specific SUDs that were studied included alcohol use disorder, methamphetamine use disorder, opioid use disorder and cocaine use disorder. In some studies, the type of SUD was not specified or there was a mix of different SUD diagnoses (designated as the unspecified SUD group). In our analysis, opioid use disorder and cocaine use disorder were combined, as there was only one study for each. Among substance types, SUD groups had a higher total DERS score than control groups. The unspecified SUD group and AUD group had similar sample sizes that were more than twice the size of the methamphetamine and opioid/cocaine use disorder groups. The unspecified SUD group had a significant effect size of 18.93 (95% CI = 13.35–24.52, P < 0.001). The AUD group had the highest significant effect size of 28.24 (95% CI = 11.46–45.02, P = 0.001). The methamphetamine use disorder group had a significant effect size of 20.17 (95% CI = 15.99–24.35, P < 0.001) and the cocaine/opioid use disorder group had a significant effect size of 21.44 (95% CI = 15.24–27.63, P < 0.001). Heterogeneity for each drug type is as follows: unspecified SUD = 90.3%, AUD = 97.1%, methamphetamine use disorder = 0.0%, cocaine/opioid use disorder = 44.9%. A forest plot of the total DERS scores by drug type is shown in Figure 5.

DERS subscale scores

The SUD group displayed significantly higher scores on all subscales when compared to the control group (Supporting information, Appendix S6, Figure S5.1–S5.6), with the Strategies (Hedges’ g = 1.01, 95% CI = 0.77–1.26, P < 0.001) and Impulse (Hedges’ g = 0.81, 95% CI = 0.58–1.04, P < 0.001) subscales having the largest standardized mean differences. The results of the remaining subscales are as follows: Non‐Acceptance’ g = 0.57 (95% CI = 0.38–0.75, P < 0.001), Goals Hedges’ g = 0.55 (95% CI = 0.34–0.76, P < 0.001), Awareness Hedges’ g = 0.55 (95% CI = 0.34–0.75, P < 0.001) and Clarity Hedges’ g = 0.67 (95% CI = 0.50–0.85, P < 0.001). Heterogeneity was high for each subscale (Non‐Acceptance = 77.7%, Goals = 83.5%, Impulse = 85.7%, Awareness = 82.0%, Strategies = 86.7%, Clarity = 74.1%). Additional information on the subscale analyses can be found in Supporting information, Appendix S6.

Shortened versions of the DERS

The addition of studies that used three shortened versions of the DERS into the primary analysis had minimal effect on the standardized mean difference between individuals with and without SUDs, with Hedges’ g = 1.13 (95% CI = 0.91–1.35, P < 0.001, Supporting information, Figure S6).

ERQ

Individuals with SUDs displayed higher levels of expressive suppression (mean difference = 3.38, 95% CI = 1.29–5.46, P = 0.002) and lower levels of cognitive reappraisal (mean difference = −3.65, 95% CI = −8.14 to 0.85, P = 0.11; Figure 6a,b) than individuals without SUDs, but the difference in cognitive reappraisal scores between cases and controls did not reach statistical significance. Heterogeneity was 83.0% for expressive suppression and 93.5% for cognitive reappraisal. When standardizing the mean difference, Hedges’ g value was 0.76 (95% CI = 0.25–1.28, P = 0.004) for expressive suppression and −0.62 (95% CI = −1.38 to 0.14, p = 0.11) for cognitive reappraisal.

Forest plot of Emotion Regulation Questionnaire (ERQ) expressive suppression factor (a) and cognitive reappraisal factor (b) scores, depicting the mean differences between the substance use disorder (SUD) group and control group for each study as well as the summary effect. Error bars for the individual studies represent the 95% CI. The 95% CI for the summary effect is represented by the diamond width. Dashed error bars for the summary effect represent the prediction interval. The prediction interval represents a prediction of the range of possible effect sizes that could be found were a new study to be conducted. N = sample size; MD = mean difference; CI = confidence interval.

DISCUSSION

This study confirmed our hypothesis that adults with SUDs display greater difficulties in emotion regulation compared to adults without SUDs. Individuals with SUDs had significantly higher total DERS scores compared to those without a SUD and scored significantly higher on all subscales of the DERS, with the Strategies and Impulse subscales having the largest effect sizes. Studies using the ERQ found that individuals with SUDs were also more likely to use suppression than controls, but did not find significant differences in the use of cognitive reappraisal. When looking at specific drug types, AUD had the largest effect size; however, this could be due at least in part to small sample sizes in the other groups.

Our analysis found a large amount of heterogeneity which could not be attributed to age, sex, country of publication or year of publication. This is not surprising, given the wide variability in study protocols, including population and setting. We could not address certain potential contributors to heterogeneity as they were not measured in the included studies. For example, length of abstinence prior to assessment was not reported in most studies, even though periods of abstinence may lead to improved emotion regulation in individuals with SUDs [53]. Furthermore, psychiatric comorbidities such as post‐traumatic stress disorder (PTSD) may also impact emotion regulation [86] and many studies did not assess comorbidities. Finally, severity of substance use was not measured in many of the studies included in our review. It has been found that higher DERS scores are associated with greater alcohol dependence severity [13, 86, 87] and experiencing more motives to drink alcohol [88]. In other studies, heavy cannabis users had higher DERS scores than less frequent cannabis users [89], and polydrug users had higher DERS scores than those only diagnosed with AUD [90, 91]. Future studies should attempt to report data on co‐occurring illnesses, length of abstinence prior to assessment and substance use severity.

Our findings raise the possibility that promoting effective emotion regulation skills in this population could improve treatment outcomes. There is some evidence that dialectical behavior therapy (DBT) interventions can be used to improve emotion regulation in individuals with SUDs. After 3 months of DBT skills training, individuals with AUD had significantly lower DERS scores [13]. Further, those with lower overall DERS scores had more consecutive days of abstinence. Women with borderline personality disorder and substance dependence also reported lower DERS scores after a 20‐week DBT program [92]. Emotion regulation skills may also be useful for common comorbidities in individuals with SUDs, as interventions that target emotion regulation have proved effective for mood and anxiety disorders [93]. Additional clinical trials are necessary to more clearly delineate the role of DBT‐related interventions in SUD treatment. Our analysis also found that the largest emotion regulation deficits were in the Strategies and Impulse subscales of the DERS. The Strategies subscale assesses whether a person feels that they have healthy approaches to manage feeling upset. The Impulse subscale assesses if a person can control their behavior when they are upset. Targeting these specific deficits in therapy may be especially helpful, including providing skills that allow individuals with SUDs to prevent themselves from acting impulsively on emotions as well as distress tolerance and mindfulness skills to prevent emotions from becoming overwhelming.

Our study was limited by a lack of ability to establish temporality. It is unclear whether emotional dysregulation or substance use presented first in the individuals examined in these studies. Emotion dysregulation at the ages of 12 and 16 years has been found to predict the risk of developing a SUD in early adulthood [94]. Conversely, it has been shown that abstinence results in improved emotion regulation in adults with SUDs [53], suggesting that recovery from a SUD could potentially lead to recovery in emotion regulation capacity. Other studies have found that individuals with SUDs continued to have greater emotion regulation difficulties compared to controls after 3–6 months of abstinence [63, 64], suggesting that emotion regulation deficits may persist after sobriety. Additionally, individuals with SUDs may have negative biases in the way they process emotional facial expressions [95], which may further exacerbate emotion dysregulation. It is likely that emotion dysregulation is both a risk factor for and sequela of addiction, but more longitudinal studies are needed to confirm this.

A shortcoming of the included studies is that some did not use validated tools to assess the control group for SUDs or other psychiatric disorders. However, we included a sensitivity analysis containing only studies which assessed controls with a validated diagnostic assessment and this analysis showed a similar effect size to our original analysis. Some studies did not assess psychiatric comorbidities in the SUD group, and this may be a limitation as it has been found that individuals with alcohol use disorders and co‐occurring mood or anxiety disorders have higher overall DERS scores than those who were only diagnosed with AUD [96]. Another limitation of many studies was poor matching between groups. For example, two studies used undergraduate students or patients’ family members as the comparison group for a treatment‐engaged group. Furthermore, not all studies indicated at which point in treatment the questionnaires were administered or the state of the participants during administration. For example, only seven of the studies in our analysis stated that participants were abstinent when completing the DERS. Future studies should aim to assess all participants with standardized diagnostic and dimensional measures of depression and anxiety, should ask about the length of abstinence prior to questionnaire administration and should include measures of intoxication and withdrawal at the time of questionnaire administration. Future studies should also carefully consider how comparison groups are matched to the SUD group. Finally, our review was limited to studies that utilized self‐report questionnaires. Task‐based assessments may provide less subjective measures of emotion regulation [97], but were beyond the scope of the current review.

Our findings demonstrate that individuals with SUDs have more difficulty regulating their emotions than individuals without SUDs. We observed a large effect, suggesting that this may be an important process leading to continued substance misuse. Compared to those without a SUD, individuals with a SUD had the most difficulty controlling impulsive behaviors and accessing strategies to regulate their emotions effectively. Assessing emotion regulation in patients with addiction may eventually allow clinicians to gain a clearer understanding of treatment trajectories and could potentially serve as an important target for therapeutic intervention.

DECLARATION OF INTERESTS

None.

AUTHOR CONTRIBUTIONS

Jordan Stellern: Data curation; investigation. Ke Bin Xiao: Data curation; investigation. Erin Grennell: Data curation; investigation. Marcos Sanches: Formal analysis; visualization. Joshua Gowin: Conceptualization; supervision. Matthew Sloan: Conceptualization; data curation; investigation; supervision.

Supporting information

Appendix S1. Search Strategy and Details on Included Studies

Appendix S2. Quality Assessment Tool (modified National Heart, Lung, and Blood Institute‘s Quality Assessment Tools for Observational Cohort and Cross‐Sectional Studies)

Appendix S3. Additional Synthesis and Calculation Methods and Discrepancies

Appendix S4. Testing for Heterogeneity

Appendix S5. Details on Analyses of Studies Using the Emotion Regulation Questionnaire and Shortened Versions of the Difficulties in Emotion Regulation Scale

Table S1.1 Excluded studies from DERS analysis (n = 14)

Table S1.2 Excluded studies in secondary analysis (n = 34)

Table S2. Studies Using the ERQ and Shortened Versions of the DERS

Figure S1. Systematic Review Search Process including Embase (Secondary Search)

Figure S2. Forest Plot for Standardized Mean Differences (SMD)

Figure S3.1. Forest Plot Organized by Geographical Region of Each Study

Figure S3.2. Forest Plot Organized by Year of Publication of Each Study

Figure S3.3. Forest Plot Organized by Sex Distribution of Each Study

Figure S3.4. Forest Plot Organized by Average Age of Each Study

Figure S4.1. Sensitivity Analysis with Exclusion of Three Studies that were Missing Total DERS Score

Figure S4.2. Sensitivity Analysis with Studies Categorized as ‘Poor’, ‘Fair’, or ‘Good’ Based on the Modified NIH Quality Assessment Tool

Figure S4.3. Sensitivity Analysis with Exclusion of Three Outlier Studies with Mean Difference Between Cases and Controls > 40

Figure S4.4. Sensitivity Analysis with Exclusion of Studies with DERS Score Values Estimated from Graphs

Figure S4.5. Sensitivity Analysis with Exclusion of Studies that did not Assess Controls Using a Validated Assessment Tool for SUD Diagnosis

Figure S4.6. Sensitivity Analysis with Addition of a Community Substance Misuse Group

Figure S5.1. Forest Plot for Nonacceptance Subscale Scores

Figure S5.2. Forest Plot for Goals Subscale Scores

Figure S5.3. Forest Plot for Impulse Subscale Scores

Figure S5.4. Forest Plot for Awareness Subscale Scores

Figure S5.5. Forest Plot for Strategies Subscale Scores

Figure S5.6. Forest Plot for Clarity Subscale Scores

Figure S6. Forest Plot with Shortened Versions of the DERS Included in the Primary Analysis

Click here for additional data file.^{(1.1MB, docx)}

ACKNOWLEDGEMENTS

K.B.X.’s funding was supported by the Ontario Ministry of Health and Long‐Term Care and the Canadian Institutes of Health Research. E.G.’s funding was supported by the Comprehensive Research Experience for Medical Students Summer Research Program through the University of Toronto. J.S.’s funding was supported by the Research Track at the University of Colorado School of Medicine. We would like to thank all the authors we contacted for taking the time to provide us with any missing information that was necessary for our review.

Stellern J, Xiao KB, Grennell E, Sanches M, Gowin JL, Sloan ME. Emotion regulation in substance use disorders: a systematic review and meta‐analysis. Addiction. 2023;118(1):30–47. 10.1111/add.16001

Contributor Information

Joshua L. Gowin, Email: joshua.gowin@cuanschutz.edu.

Matthew E. Sloan, Email: matthew.sloan@camh.ca.

REFERENCES

1. Carvalho AF, Heilig M, Perez A, Probst C, Rehm J. Alcohol use disorders. Lancet. 2019;394:781–92. [DOI] [PubMed] [Google Scholar]
2. United Nations Office on Drugs and Crime (UNODC) . World Drug Report Executive Summary. Vienna, Austria: UNODC; 2019. [Google Scholar]
3. Wellman RJ, Chaiton M, Morgenstern M, O’Loughlin J. Untangling influences in the longitudinal relationship between depressive symptoms and drinking frequency in high school. J Adolesc Health. 2020;66:308–14. [DOI] [PubMed] [Google Scholar]
4. Wellman RJ, Contreras GA, Dugas EN, O’Loughlin EK, O’Loughlin JL. Determinants of sustained binge drinking in young adults. Alcohol Clin Exp Res. 2014;38:1409–15. [DOI] [PubMed] [Google Scholar]
5. Kenneson A, Funderburk JS, Maisto SA. Substance use disorders increase the odds of subsequent mood disorders. Drug Alcohol Depend. 2013;133:338–43. [DOI] [PubMed] [Google Scholar]
6. Murphy A, Taylor E, Elliott R. The detrimental effects of emotional process dysregulation on decision‐making in substance dependence. Front Integr Neurosci. 2012;6:101. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Thorberg FA, Lyvers M. Negative mood regulation (NMR) expectancies, mood, and affect intensity among clients in substance disorder treatment facilities. Addict Behav. 2006;31:811–20. [DOI] [PubMed] [Google Scholar]
8. McRae K, Gross JJ. Emotion regulation. Emotion. 2020;20:1–9. [DOI] [PubMed] [Google Scholar]
9. D’Agostino A, Covanti S, Rossi Monti M, Starcevic V. Reconsidering emotion dysregulation. Psychiatry Q. 2017;88:807–25. [DOI] [PubMed] [Google Scholar]
10. Vøllestad J, Nielsen MB, Nielsen GH. Mindfulness‐ and acceptance‐based interventions for anxiety disorders: a systematic review and meta‐analysis. Br J Clin Psychol. 2012;51:239–60. [DOI] [PubMed] [Google Scholar]
11. Gross JJ, John OP. Individual differences in two emotion regulation processes: implications for affect, relationships, and well‐being. J Pers Soc Psychol. 2003;85:348–62. [DOI] [PubMed] [Google Scholar]
12. Linehan MM. Cognitive–Behavioral Treatment of Borderline Personality Disorder New York, NY: Guilford Press; 2018. [Google Scholar]
13. Cavicchioli M, Movalli M, Vassena G, Ramella P, Prudenziati F, Maffei C. The therapeutic role of emotion regulation and coping strategies during a stand‐alone DBT skills training program for alcohol use disorder and concurrent substance use disorders. Addict Behav. 2019;98:106035. [DOI] [PubMed] [Google Scholar]
14. Gratz KL, Roemer L. Multidimensional assessment of emotion regulation and dysregulation: development, factor structure, and initial validation of the difficulties in emotion regulation scale. J Psychopathol Behav Assess. 2004;26:41–54. [Google Scholar]
15. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Magd S, Rakhawy M, Mamdouh R, Shaheen S. Impulsivity, suicidality, and emotional dysregulation in women having borderline personality disorder with and without substance dependence. Egypt J Psychiatry. 2019;40:59–63. [Google Scholar]
17. Uhlmann A, Fouche J‐P, Koen N, Meintjes EM, Wilson D, Stein DJ. Fronto‐temporal alterations and affect regulation in methamphetamine dependence with and without a history of psychosis. Psych Res: Neuroimaging. 2016;248:30–8. [DOI] [PubMed] [Google Scholar]
18. Fowler JC, Charak R, Elhai JD, Allen JG, Frueh BC, Oldham JM. Construct validity and factor structure of the difficulties in emotion regulation scale among adults with severe mental illness. J Psychiatr Res. 2014;58:175–80. [DOI] [PubMed] [Google Scholar]
19. Giromini L, Velotti P, de Campora G, Bonalume L, Cesare Zavattini G. Cultural adaptation of the difficulties in emotion regulation scale: reliability and validity of an Italian version. J Clin Psychol. 2012;68:989–1007. [DOI] [PubMed] [Google Scholar]
20. Ritschel LA, Tone EB, Schoemann AM, Lim NE. Psychometric properties of the difficulties in emotion regulation scale across demographic groups. Psychol Assess. 2015;27:944–54. [DOI] [PubMed] [Google Scholar]
21. Giromini L, Ales F, de Campora G, Zennaro A, Pignolo C. Developing age and gender adjusted normative reference values for the difficulties in emotion regulation scale (DERS). J Psychopathol Behav Assess. 2017;39:705–14. [Google Scholar]
22. Veritas Health Innovation . Covidence Systematic Review Software Melbourne, Australia; 2021. Available at: www.covidence.org. Accessed 1 May 2021.
23. National Heart Lung and Blood Institute , National Institutes of Health , US Department of Health and Human Services . Quality Assessment Tool for Observational Cohort and Cross‐Sectional Studies; 2014. Available at: https://www.nhlbi.nih.gov/health-pro/guidelines/in-develop/cardiovascular-risk-reduction/tools/cohort. Accessed 1 July 2021.
24. Viechtbauer W. Conducting meta‐analyses in R with the metafor package. J Stat Softw. 2010;36:1–48. [Google Scholar]
25. Hedges LV, Vevea JL. Fixed‐ and random‐effects models in meta‐analysis. Psychol Methods. 1998;3:486–504. [Google Scholar]
26. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta‐analysis. Stat Med. 2002;21:1539–58. [DOI] [PubMed] [Google Scholar]
27. Garke MÅ, Isacsson NH, Sörman K, Bjureberg J, Hellner C, Gratz KL, et al. Emotion dysregulation across levels of substance use. Psychiatry Res. 2021;296:113662. [DOI] [PubMed] [Google Scholar]
28. Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ et al. Cochrane Handbook for Systematic Reviews of Interventions, version 6.3. Cochrane; 2022. Available at: www.training.cochrane.org/handbook. Accessed 1 July 2021.
29. Wolff S, Holl J, Stopsack M, Arens EA, Höcker A, Staben KA, et al. Does emotion dysregulation mediate the relationship between early maltreatment and later substance dependence? Findings of the CANSAS study. Eur Addict Res. 2016;22:292–300. [DOI] [PubMed] [Google Scholar]
30. Hiebler‐Ragger M, Perchtold‐Stefan CM, Unterrainer HF, Fuchshuber J, Koschutnig K, Nausner L, et al. Lower cognitive reappraisal capacity is related to impairments in attachment and personality structure in poly‐drug use: an fMRI study. Brain Imaging Behav. 2021;15:2187–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Xin Z, Lu X, Li F, Haitao H, Ling Y, Aibao Z. Emotion regulation in male abstinent heroin abusers. Psychol Rep. 2014;114:14–9. [DOI] [PubMed] [Google Scholar]
32. Balducci T, González‐Olvera JJ, Angeles‐Valdez D, Espinoza‐Luna I, Garza‐Villarreal EA. Borderline personality disorder with cocaine dependence: impulsivity, emotional dysregulation and amygdala functional connectivity. Front Psychiatry. 2018;9:328. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Buckholdt KE, Parra GR, Anestis MD, Lavender JM, Jobe‐Shields LE, Tull MT, et al. Emotion regulation difficulties and maladaptive behaviors: examination of deliberate self‐harm, disordered eating, and substance misuse in two samples. Cognit Ther Res. 2015;39:140–52. [Google Scholar]
34. Árpási‐Kmoskó E, Szemán‐Nagy A. Examination of emotion regulation difficulties, coping, impulsivity, external‐internal control and sensation seeking among occasional and regular marijuana users. Psychiatr Hung. 2021;36:53–66. [PubMed] [Google Scholar]
35. Charfi N, Mseddi N, Sallemi R, Zahaf A, Maâlej‐Bouali M, Omri S, et al. Tempéraments affectifs des toxicomanes: étude cas‐témoins [Temperaments of drug addicts: a case‐control study]. Encéphale. 2019;45:226–31. [DOI] [PubMed] [Google Scholar]
36. Lillaz C, Varescon I. Emotion regulation among psychostimulants drug users at techno parties. L’Encephale. 2012;38:390–6. [DOI] [PubMed] [Google Scholar]
37. Faulkner P, Dean AC, Ghahremani DG, London ED. Neural basis of smoking‐related difficulties in emotion regulation. Int J Neuropsychopharmacol. 2020;23:409–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Ivezaj V, Saules KK, Schuh LM. New‐onset substance use disorder after gastric bypass surgery: rates and associated characteristics. Obes Surg. 2014;24:1975–80. [DOI] [PubMed] [Google Scholar]
39. Prosek EA, Giordano AL, Woehler ES, Price E, McCullough R. Differences in emotion dysregulation and symptoms of depression and anxiety among illicit substance users and nonusers. Subst Use Misuse. 2018;53:1915–8. [DOI] [PubMed] [Google Scholar]
40. Migliore S, Paolucci M, Quintiliani L, Altamura C, Maffi S, D’Aurizio G, et al. Psychopathological comorbidities and clinical variables in patients with medication overuse headache. Front Hum Neurosci. 2020;14:571035. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Claudat K, Brown TA, Anderson L, Bongiorno G, Berner LA, Reilly E, et al. Correlates of co‐occurring eating disorders and substance use disorders: a case for dialectical behavior therapy. Eat Disord. 2020;28:142–56. [DOI] [PubMed] [Google Scholar]
42. Berlingeri M, Losasso D, Girolo A, Cozzolino E, Masullo T, Scotto M, et al. Resting state brain connectivity patterns before eventual relapse into cocaine abuse. Behav Brain Res. 2017;327:121–32. [DOI] [PubMed] [Google Scholar]
43. Fox HC, Bergquist KL, Casey J, Hong KA, Sinha R. Selective cocaine‐related difficulties in emotional intelligence: relationship to stress and impulse control. Am J Addict. 2011;20:151–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Kopera M, Trucco EM, Suszek H, Kobyliński P, Wiśniewski P, Wojnar M, et al. Pain sensitivity, negative affect, and alcohol use disorder status: a moderated mediation study of emotion dysregulation. J Clin Med. 2021;10:1321. [DOI] [PMC free article] [PubMed] [Google Scholar]
45. C‐sR L, Huang C, Yan P, Bhagwagar Z, Milivojevic V, Sinha R. Neural correlates of impulse control during stop signal inhibition in cocaine‐dependent men. Neuropsychopharmacology. 2008;33:1798–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
46. Ejei J, Gholamali Lavasani M, Erami H. Comparison of emotional regulation in substance abusers and normal subjects. J Psychol. 2015;19:137–46. [Google Scholar]
47. Stover CS, Easton CJ, McMahon TJ. Parenting of men with co‐occurring intimate partner violence and substance abuse. J Interpers Violence. 2013;28:2290–314. [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Seo D, Lacadie CM, Sinha R. Neural correlates and connectivity underlying stress‐related impulse control difficulties in alcoholism. Alcohol Clin Exp Res. 2016;40:1884–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
49. Azizi S, Maghsoudloo A, Baheshmat S. Comparison of coping strategies and emotion regulation tendencies among opium users, methadone maintenance treatment clients and normal individuals. J Res Health. 2019;9:533–43. [Google Scholar]
50. Bottesi G, Ghisi M, Caggiu I, Lauriola M. How is intolerance of uncertainty related to negative affect in individuals with substance use disorders? The role of the inability to control behaviors when experiencing emotional distress. Addict Behav. 2021;115:106785. [DOI] [PubMed] [Google Scholar]
51. Di Pierro R, Benzi IMA, Madeddu F. Difficulties in emotion regulation among inpatients with substance use disorders: the mediating effect of mature defenses mechanisms. Clin Neuropsychiatry J Treat Eval. 2015;12:83–9. [Google Scholar]
52. Dingle GA, Neves DC, Alhadad SSJ, Hides L. Individual and interpersonal emotion regulation among adults with substance use disorders and matched controls. Br J Clin Psychol. 2018;57:186–202. [DOI] [PubMed] [Google Scholar]
53. Fox HC, Axelrod SR, Paliwal P, Sleeper J, Sinha R. Difficulties in emotion regulation and impulse control during cocaine abstinence. Drug Alcohol Depend. 2007;89:298–301. [DOI] [PubMed] [Google Scholar]
54. Fox HC, Hong KA, Sinha R. Difficulties in emotion regulation and impulse control in recently abstinent alcoholics compared with social drinkers. Addict Behav. 2008;33:388–94. [DOI] [PubMed] [Google Scholar]
55. Ghorbani F, Khosravani V, Sharifi Bastan F, Jamaati Ardakani R. The alexithymia, emotion regulation, emotion regulation difficulties, positive and negative affects, and suicidal risk in alcohol‐dependent outpatients. Psychiatry Res. 2017;252:223–30. [DOI] [PubMed] [Google Scholar]
56. Hardy R, Fani N, Jovanovic T, Michopoulos V. Food addiction and substance addiction in women: common clinical characteristics. Appetite. 2018;120:367–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
57. Jakubczyk A, Trucco EM, Klimkiewicz A, Skrzeszewski J, Suszek H, Zaorska J, et al. Association between interoception and emotion regulation in individuals with alcohol use disorder. Front Psychiatry. 2020;10:1028. [DOI] [PMC free article] [PubMed] [Google Scholar]
58. Krause‐Utz A, Erol E, Brousianou AV, Cackowski S, Paret C, Ende G, et al. Self‐reported impulsivity in women with borderline personality disorder: the role of childhood maltreatment severity and emotion regulation difficulties. Bord Personal Disord Emot Dysregul. 2019;6:6. [DOI] [PMC free article] [PubMed] [Google Scholar]
59. London ED, Okita K, Kinney KR, Dean AC, McClintick MN, Rizor EJ, et al. No significant elevation of translocator protein binding in the brains of recently abstinent methamphetamine users. Drug Alcohol Depend. 2020;213:108104. [DOI] [PMC free article] [PubMed] [Google Scholar]
60. Okita K, Ghahremani DG, Payer DE, Robertson CL, Dean AC, Mandelkern MA, et al. Emotion dysregulation and amygdala dopamine D2‐type receptor availability in methamphetamine users. Drug Alcohol Depend. 2016;161:163–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
61. Pelot A, Perron M, Lewis S, Roy‐Charland A. Smile judgment in substance use disorders and its relationship to interpersonal and emotional functioning: an eye‐tracking investigation. Drug Alcohol Depend. 2020;208:107842. [DOI] [PubMed] [Google Scholar]
62. Rettie HC, Hogan LM, Cox WM. Negative attentional bias for positive recovery‐related words as a predictor of treatment success among individuals with an alcohol use disorder. Addict Behav. 2018;84:86–91. [DOI] [PubMed] [Google Scholar]
63. Russell BS, Heller AT, Hutchison M. Differences in adolescent emotion regulation and impulsivity: a group comparison study of school‐based recovery students. Subst Use Misuse. 2017;52:1085–97. [DOI] [PubMed] [Google Scholar]
64. Russell BS, Hutchison M, Fusco A. Emotion regulation outcomes and preliminary feasibility evidence from a mindfulness intervention for adolescent substance use. J Child Adolesc Subst Abuse. 2019;28:21–31. [Google Scholar]
65. Zareban I, Bakhshani N, Bor M, Bakhshani S. Emotion regulation difficulties in drug abusers. Ann Trop Med Public Health. 2017;10:S129. [Google Scholar]
66. Akbari M, Bahadori MH, Mohammadkhani S, Kolubinski DC, Nikčević AV, Spada MM. A discriminant analysis model of psychosocial predictors of problematic internet use and cannabis use disorder in university students. Addict Behav Rep. 2021;14:100354. [DOI] [PMC free article] [PubMed] [Google Scholar]
67. Ding X, Jiang H, Xu M, Li Y, Liang J, Xie R. The ineffective emotion regulation strategies of heroin use disorder patients: an event‐related potential study. Drug Alcohol Depend. 2021;228:109076. [DOI] [PubMed] [Google Scholar]
68. McDonnell E, Hevey D, McCauley M, Ducray KN. Exploration of associations between early maladaptive schemas, impaired emotional regulation, coping strategies and resilience in opioid dependent poly‐drug users. Subst Use Misuse. 2018;53:2320–9. [DOI] [PubMed] [Google Scholar]
69. Petit G, Luminet O, Maurage F, Tecco J, Lechantre S, Ferauge M, et al. Emotion regulation in alcohol dependence. Alcohol Clin Exp Res. 2015;39:2471–9. [DOI] [PubMed] [Google Scholar]
70. Berking M, Margraf M, Ebert D, Wupperman P, Hofmann SG, Junghanns K. Deficits in emotion‐regulation skills predict alcohol use during and after cognitive–behavioral therapy for alcohol dependence. J Consult Clin Psychol. 2011;79:307–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
71. Ogel K, Sarp N, Gurol DT, Ermagan E. Investigation of mindfulness and affecting factors of mindfulness among substance users and non users [Bagimli olan ve olmayan bireylerde farkindalik (mindfulness) ve farkindaligi etkileyen etkenlerin incelenmesi]. Anadolu Psikiyatri Dergisi. 2014;15:282–9. [Google Scholar]
72. Vierhaus M, Ewering J, Klein F, Ködding C, Petry J. Zur Validität des Modells zur psychischen Vulnerabilität der Glücksspielsucht. Sucht. 2012;58:183–93. [Google Scholar]
73. Eiden RD, Godleski SA, Colder CR, Livingston JA, Leising MC, Leonard KE. Early childhood risk and protective factors predicting resilience against adolescent substance use. Advers Resil Sci. 2020;1:107–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
74. Monteschi M, Ignacio de Padua A, Riva Perez C, Vera Castellano M, Cedano S, Luiz Rodrigues‐Júnior A, et al. A simple test to assess the emotional status of smokers. Clin Respir J. 2018;12:2606–12. [DOI] [PubMed] [Google Scholar]
75. Sheets ES, Bujarski S, Leventhal AM, Ray LA. Emotion differentiation and intensity during acute tobacco abstinence: a comparison of heavy and light smokers. Addict Behav. 2015;47:70–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
76. Tang Y‐Y, Tang R, Posner MI. Mindfulness meditation improves emotion regulation and reduces drug abuse. Drug Alcohol Depend. 2016;163:S13–8. [DOI] [PubMed] [Google Scholar]
77. Jakubczyk A, Trucco EM, Kopera M, Kobyliński P, Suszek H, Fudalej S, et al. The association between impulsivity, emotion regulation, and symptoms of alcohol use disorder. J Subst Abuse Treat. 2018;91:49–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
78. Moselhy HF, Fahmy E, Mikhael VS, El‐Sheikh H. Emotional control in patients with opioid dependence syndrome and reported history of negative life events. Addict Disord Treat. 2012;11:93–100. [Google Scholar]
79. Okasha T, Abd Elsamie A, Azzam H, Elserafi D, Morsy M, Shorub E. Emotional regulation as a mediating factor in substance use disorders. Addict Disord Treat. 2021;20:202–10. [Google Scholar]
80. Okasha T, Abdelsamie A, Azzam H, Elserafi D, Shorub EM, El Hawary Y, et al. Emotion regulation, impulsivity, and personality profile among Egyptian patients with substance use disorders. Addict Disord Treat. 2021;20:85–90. [Google Scholar]
81. Ramirez V, Wiers CE, Wang GJ, Volkow ND. Personality traits in substance use disorders and obesity when compared to healthy controls. Addiction. 2020;115:2130–9. [DOI] [PubMed] [Google Scholar]
82. Spada MM, Wells A. Metacognitions across the continuum of drinking behaviour. Personal Individ Differ. 2010;49:425–9. [Google Scholar]
83. Tsavou E, Petkari E. Associations of personality traits and emotional intelligence: comparing individuals in rehabilitation from drug misuse, occasional users and non‐users. Subst Use Misuse. 2020;55:252–60. [DOI] [PubMed] [Google Scholar]
84. Ramezani S, Afkhamzadeh AR, Qorbani H, Naimi S, Rahmani S. Effect of mindfulness‐based cognitive therapy on substance dependence intensity and cognitive emotion regulation in patients under methadone maintenance treatment. Pract Clin Psychol. 2019;7:225–34. [Google Scholar]
85. Ford KA, Wammes M, Neufeld RW, Mitchell D, Théberge J, Williamson P, Osuch EA. Unique functional abnormalities in youth with combined marijuana use and depression: an fMRI study. Front Psychiatry. 2014;5:130. [DOI] [PMC free article] [PubMed] [Google Scholar]
86. Watkins LE, Schumacher JA, Coffey SF. A preliminary investigation of the relationship between emotion dysregulation and partner violence perpetration among individuals with PTSD and alcohol dependence. J Aggress Maltreat Trauma. 2016;25:305–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
87. Khosravani V, Sharifi Bastan F, Ghorbani F, Kamali Z. Difficulties in emotion regulation mediate negative and positive affects and craving in alcoholic patients. Addict Behav. 2017;71:75–81. [DOI] [PubMed] [Google Scholar]
88. Sung‐Doo W. The influence of daily stress on impaired control through emotion dysregulation and drinking motives. Korean J Clin Psychol. 2017;36:565–73. [Google Scholar]
89. Houck JM, Bryan AD, Feldstein Ewing SW. Functional connectivity and cannabis use in high‐risk adolescents. Am J Drug Alcohol Abuse. 2013;39:414–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
90. Cavicchioli M, Movalli M, Maffei C. Difficulties with emotion regulation, mindfulness, and substance use disorder severity: the mediating role of self‐regulation of attention and acceptance attitudes. Am J Drug Alcohol Abuse. 2019;45:97–107. [DOI] [PubMed] [Google Scholar]
91. Lucke HR, Harbke CR, Mathes EW, Hammersley JJ. Higher emotion dysregulation and coping motives in alcohol and marijuana users. Subst Use Misuse. 2021;56:950–61. [DOI] [PubMed] [Google Scholar]
92. Axelrod SR, Perepletchikova F, Holtzman K, Sinha R. Emotion regulation and substance use frequency in women with substance dependence and borderline personality disorder receiving dialectical behavior therapy. Am J Drug Alcohol Abuse. 2011;37:37–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
93. Daros AR, Haefner SA, Asadi S, Kazi S, Rodak T, Quilty LC. A meta‐analysis of emotional regulation outcomes in psychological interventions for youth with depression and anxiety. Nat Hum Behav. 2021;5:1443–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
94. Kirisci L, Tarter R, Ridenour T, Reynolds M, Horner M, Vanyukov M. Externalizing behavior and emotion dysregulation are indicators of transmissible risk for substance use disorder. Addict Behav. 2015;42:57–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
95. Freeman CR, Wiers CE, Sloan ME, Zehra A, Ramirez V, Wang G‐J, et al. Emotion recognition biases in alcohol use disorder. Alcohol Clin Exp Res. 2018;42:1541–7. [DOI] [PubMed] [Google Scholar]
96. Bradizza CM, Brown WC, Ruszczyk MU, Dermen KH, Lucke JF, Stasiewicz PR. Difficulties in emotion regulation in treatment‐seeking alcoholics with and without co‐occurring mood and anxiety disorders. Addict Behav. 2018;80:6–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
97. Ochsner KN, Bunge SA, Gross JJ, Gabrieli JDE. Rethinking feelings: an fMRI study of the cognitive regulation of emotion. J Cogn Neurosci. 2002;14:1215–29. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials