Alcohol Use and Aggression among Men in Residential Treatment for Substance Use Disorders: The Moderating Role of Mindfulness Facets

Abstract

Objective:

Aggressive behaviors are consistently linked to alcohol use and are over-represented among individuals in substance misuse treatment compared to the general population. Trait mindfulness (i.e., a multifaceted construct including attending to and accepting experiences) had an inverse relationship with aggression such that greater trait mindfulness may attenuate the risk alcohol use poses for aggression.

Method:

We examined the moderating effect of the five facets of trait mindfulness on the relationship between alcohol use/problems and aggressive attitudes, physical aggression, and verbal aggression among 516 men in residential treatment for substance use disorders.

Results:

Explication of the significant interactions utilizing the Johnson-Neyman technique revealed alcohol use/problems positively related to aggressive attitudes for men with below average and average Nonreactivity and Acting mindfulness facets. The positive relationship between alcohol use/problems and verbal aggression was present among men below average in Nonreactivity and Acting. Alcohol use/problems positively related to physical aggression among men below average and average in the mindfulness facets of Describe and Acting. Furthermore, alcohol use/problems positively related to aggressive attitudes and physical aggression among men average or above average in the mindfulness facet of Observe.

Conclusions:

Findings suggest that among men who misuse substances, the relation between alcohol use and aggression may be conditional and influenced by deficits in trait mindfulness.

Aggressive behaviors, such as verbal and physical aggression, are consistently linked to alcohol consumption (Bushman, 2002; Bushman & Cooper, 1990; Foran & O’Leary, 2008). Furthermore, emerging research suggests aggressive attitudes (e.g., aggression is acceptable when defending one’s ego) are also related to alcohol use (Archer, 2004; Nunes et al., 2014; Shorey et al., 2015; Stefanile et al., 2017; Wells et al., 2013). Alcohol myopia theory (AMT; Steele and Josephs, 1990) suggests that it is the pharmacological effects of alcohol which results in attentional resources focusing on the most salient cues within the environment. When in aggressive situations, alcohol intoxication results in attentional resources being allocated to aggression-provoking cues (e.g., perceived insult) and away from aggression-inhibiting cues (e.g., risk of injury). AMT is well supported in explaining the relationship between alcohol use and aggressive behaviors (e.g., Giancola et al., 2010). However, not all who consume alcohol become aggressive indicating a need to identify individual risk factors that moderate the alcohol-aggression relationship.

One potential risk factor for the alcohol-aggression link is low trait mindfulness. Mindfulness is defined as bringing one’s attention to, and the acceptance of, present moment experiences (Baer et al., 2006; Brown & Ryan, 2003). Mindfulness can be conceptualized as a state, brought about through practices such as mindful meditation, and as a dispositional trait (i.e., mindful tendencies and characteristics of an individual at any given time; Baer et al., 2006; Brown et al., 2007). Mindfulness is a multifaceted construct including the observation of internal and external experiences (i.e., Observe), describing internal experiences with words (i.e., Describe), purposefully acting upon present moment experiences (i.e., Acting), taking a nonjudging stance towards internal experiences (i.e., Nonjudging), and experiencing internal experiences without reacting to them (i.e., Nonreactivity; Baer et al., 2006; Baer et al., 2008). Facets of mindfulness were found to differentially relate to various outcomes. For example, a review of the literature showed that mindfulness facets of Acting, Nonjudging, and Nonreactivity negatively associated with alcohol use and this relationship was strongest among inpatient treatment-seeking samples (Karyadi et al., 2014). In the context of AMT and aggression, those with low trait mindfulness may be especially susceptible to the alcohol-related attentional shift towards aggression-provoking cues due to deficits in purposefully attending to, acting on, and accepting present experiences (Giancola et al., 2010). It is possible that the relationship between alcohol use and aggression may be conditional on deficits in specific facets of mindfulness.

Trait mindfulness and aggression have repeatedly shown to have a negative association with one another. Among samples of men and women undergraduates, trait mindfulness, as a single construct, was negatively correlated with anger, hostility, and verbal aggression (Heppner et al., 2008), as well as physical aggression (Borders et al., 2010). Prior work examined facets of mindfulness and found similar negative relations with aggression (Eisenlohr-Moul et al., 2016). For example, Peters et al. (2015) examined the five facets of mindfulness utilizing the Five Facet Mindfulness Questionnaire (FFMQ; Baer et al., 2006) among an undergraduate sample as they related to anger, hostility, verbal and physical aggression. They found that anger, verbal, and physical aggression were negatively related to the facets of Acting and Nonjudging and hostility negatively related to all facets except Observe. Of note, the Observe facet positively related to anger, hostility, verbal, and physical aggression, leading the researchers to echo calls questioning the validity of this facet within the FFMQ (Baer et al., 2008, Carpenter et al., 2019; Gillespie et al., 2018; Rudkin et al., 2018; Williams et al., 2014).

Furthermore, prior research supported the association between men’s trait mindfulness, specifically, and aggression. For example, trait mindfulness negatively associated with aggressive attitudes, verbal and physical aggression among male residential substance use disorder (SUD) patients when controlling for correlates of aggression including substance use (Shorey et al., 2015). Among a sample of men in prison for violent offenses, the mindfulness facets of Describe, Acting, and Nonjudging significantly and negatively correlated with physical aggression and aggressive traits such as anger and hostility (Velotti et al., 2016).

Emerging research supported low levels of mindfulness as a risk factor for alcohol-facilitated aggression. Trait mindfulness moderated the relationship between alcohol use and the perpetration of intimate partner sexual aggression among men, such that alcohol use positively associated with sexual aggression perpetration among men with low, but not high, levels of trait mindfulness (Gallagher et al., 2010). However, in Gallagher et al.’s (2010) study, trait mindfulness was examined as a singular construct rather than a multi-dimensional construct. Research is still needed to examine the individual facets of mindfulness as moderators of the alcohol-aggression link.

Prior research suggested low levels of the Acting facets of mindfulness may pose as a risk factor for alcohol-facilitated aggression. In a community sample of men and women presenting to the emergency department, a significant alcohol use by mindfulness facet of Acting interaction emerged, such that those low in Acting were more likely to perpetrate physical dating violence (Ngo et al., 2018). However, in this study there was no interaction between alcohol and mindfulness facet of Nonjudging relating to perpetration of physical dating violence. It should be noted, this study only examined interaction effects of the facets of mindfulness that exhibited significant main effects (i.e., Acting and Nonjudging).

There is a paucity of research in this area as no other study examined facets of mindfulness as a moderator of alcohol use and aggression. However, the alcohol-facilitated aggression literature provides guidance on deficits in specific facets of mindfulness that may increase the risk of aggression when under the influence of alcohol. For example, impulsivity has been consistently linked to aggressive behaviors (Bousardt et al., 2016; Roozen et al., 2011; Shorey et al., 2011; Stuart & Holtzworth-Munroe, 2005). Impulsivity, defined as quick, unplanned reactions to internal and/or external experiences without consideration of consequences (Moeller et al., 2001), could be considered the opposite of Acting with awareness or Nonreactivity to internal experiences given that these facets involve purposeful action and not acting on internal reactions. Impulsivity was negatively associated with facets of mindfulness (Peters et al., 2011). Furthermore, impulsive traits were found to moderate the association between alcohol use and aggression such that impulsivity increased the likelihood of aggression when under the influence of alcohol (Birkley et al., 2013; Kirwan et al., 2019; Watkins et al., 2014). Given these findings, it is possible that the association between alcohol use and aggression is conditional and influenced by low levels of the mindfulness facets of Acting and Nonreactivity.

In sum, alcohol’s disinhibiting effect as it relates to increased likelihood of aggression is well supported in prior research; however, researchers have called on the need to identify individual risk-factors that increase the likelihood of alcohol-facilitated aggression (Giancola et al., 2010). Dispositional mindfulness, both as a singular and multi-faceted construct, was negatively associated with aggression. Emerging research suggested that deficits in certain facets of mindfulness (e.g., Acting with awareness) may be particularly relevant moderators of the alcohol-aggression relationship.

The purpose of the present study was to examine the moderating effects of the five facets of mindfulness (i.e., Observe, Describe, Acting, Nonjudging, and Nonreactivity), as a dispositional trait, on the relationship between alcohol use and aggression. Given prior research which suggested that alcohol has a greater effect of increasing aggression among men than women (Bushman, 2002; Giancola et al., 2009; Gussler-Burkhardt & Giancola, 2005) and that aggressive behaviors are over-represented among substance use treatment samples compared to non-treatment seeking samples (Foran & O’Leary, 2008; Murray et al., 2008; Stuart et al., 2009), we examined the moderating role of mindfulness facets within the alcohol-aggression relation in a sample of men in residential treatment for SUD. We examined the interactive effects of alcohol use and the five facets of mindfulness on aggressive attitudes, verbal and physical aggression, while statistically controlling for known correlates of aggression including age, drug use, positive and negative impression management, and antisocial features (Caputo, 2019; Gardner et al., 2015; Hennessy & Wiesenthal, 2004; Stuart et al., 2008). We hypothesized that the relationship between alcohol use and aggressive attitudes, verbal and physical aggression would be present among patients with low, but not high, levels of the five mindfulness facets.

Methods

Participants

We reviewed medical records from 516 men in a Southeastern United States residential treatment program for SUDs. The average age of the sample was 41.15 years (SD = 11.71). Eighty-eight percent of the sample reported their race/ethnicity as White, 5.6% Black, 5.4% Hispanic/Latino, .6% Native American, and less than one percent identified as Bi/multiracial and Other, respectively. Regarding substance use disorder diagnoses, 71% were diagnosed with alcohol use disorder and 29% were given a drug use disorder including: 20% opioid use disorder, 17% cannabis use disorder, 16% Sedative, Hypnotic, or Anxiolytic Use Disorder, 14% Amphetamine-type Substance Use Disorder, 11% Cocaine Use Disorder, 2% Other Hallucinogen Use Disorder, 2% Other or Unknown Substance Use Disorder, and .4% Inhalant use disorder. Thirty-eight percent of the sample were diagnosed with two or more SUDs. Table 1 provides t-tests, with Bonferroni correction, means and standard deviations of study variables comparing patients with an alcohol use disorder versus patients with a drug use disorder.

Table 1.

T-tests, Means and Standard Deviations of Study Variables by Substance Use Disorder Diagnosis

Variable	M	SD	t
Age
AUD Dx	41.97	11.31	−2.52
DUD Dx	39.11	12.45
Anti-social Features
AUD Dx	55.34	11.44	−8.97***
DUD Dx	59.78	13.24
DUDIT
AUD Dx	6.21	11.80	5.75***
DUD Dx	14.40	14.03
Positive Impression Management
AUD Dx	48.84	12.11	−1.25
DUD Dx	47.28	12.28
Negative Impression Management
AUD Dx	52.73	10.55	.36
DUD Dx	53.11	9.93
AUDIT
AUD Dx	20.03	10.49	−8.97***
DUD Dx	8.85	12.05
Acting
AUD Dx	28.78	7.68	.238
DUD Dx	29.00	8.11
Describe
AUD Dx	27.26	6.91	−1.60
DUD Dx	25.94	7.10
Nonjudging
AUD Dx	22.09	5.96	−2.13
DUD Dx	20.59	5.75
Nonreactivity
AUD Dx	25.48	7.49	−1.73
DUD Dx	23.89	8.34
Observe
AUD Dx	30.49	8.02	−1.35
DUD Dx	29.20	8.20
Aggressive Attitudes
AUD Dx	49.48	11.74	.23
DUD Dx	49.76	12.34
Verbal Aggression
AUD Dx	49.40	10.12	.83
DUD Dx	50.29	11.15
Physical Aggression
AUD Dx	52.05	11.54	2.03
DUD Dx	54.62	14.05

Note. AUD Dx = Alcohol use disorder diagnosis; DUD Dx = Drug use disorder diagnosis; AUDIT = Alcohol Use Disorders Identification Test; DUDIT = Drug Use Disorders Identification Test.

^*p < .05.

^**p < .01.

^***p < .001.

Procedures

We reviewed de-identified medical records (i.e., demographic information, self-report assessments, diagnoses) of men entering residential treatment for SUDs from May 11, 2018 to March 12, 2019. Medical records examined contained only scored assessments and not individual responses to items; thus, Cronbach alphas could not be calculated for the present study. Patients provided consent for their de-identified medical information to be used for research purposes. Self-report measures were administered at intake or following completion of medical detoxification, when necessary. SUD diagnoses, based on criteria in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5; American Psychiatric Association, 2013), were given by the patients’ treatment team including a psychiatrist, psychologists, and SUD counselors. The residential treatment program follows the 12-step model and patients are typically referred to the 28- to 30-day residential program. The Institutional Review Board of the last author approved the procedures in the present study.

Measures

Alcohol Use Disorder Identification Test (AUDIT).

The 10-item AUDIT (Saunders et al., 1993) is a self-report measure of the frequency and quantity of alcohol use, as well as negative consequences due to alcohol use including symptoms of alcohol dependence, in the year prior to entrance into treatment. Responses on AUDIT items ranged from 0 to 4. AUDIT scores ranged from 0 to 40 and higher scores suggested greater alcohol use/problems. The AUDIT is a reliable and valid assessment of alcohol use/problems in clinical samples (Babor et al., 2001).

Drug Use Disorders Identification Test (DUDIT).

The 14-item DUDIT (Stuart et al., 2004) is a self-report assessment of drug use frequency and drug-related problematic consequences. The DUDIT assessed drug use/problems in the year prior to treatment, including seven categories of drugs: cannabis, cocaine, stimulants, sedatives/hypnotics/anxiolytics, opiates, hallucinogens, and other drugs. Scores on the DUDIT range from 0 to 70, with higher scores indicating greater drug use/problems. The DUDIT demonstrated good psychometric properties (Stuart et al., 2004) and was used as a measure of drug/use problems among multiple populations (Brem et al., 2017; Rothman et al., 2018; Wolford-Clevenger et al., 2017).

Five Facet Mindfulness Questionnaire (FFMQ).

The FFMQ (Baer et al., 2006) is a 39-item self-report measure of five facets of trait mindfulness including: Observe (i.e., observation of experiences), Describe (i.e., describing with words), Acting (i.e., acting with awareness), Nonjudging (i.e., nonjudging of inner experiences), and Nonreactivity (i.e., nonreactivity to inner experiences). Participants responded to items (e.g., “I am good at finding words to describe my feelings”) on a 1 (never or very rarely true) to 5 (very often or always true) scale, with total scores ranging from 8 to 40 on the Observe, Describe, Acting, and Nonjudging facets and 7 to 35 on the Nonreactivity facet. Higher scores indicate greater trait mindfulness within the respective facet. The FFMQ is a reliable and valid measure of the five facets of mindfulness among multiple populations (Baer et al., 2008; Brown et al., 2009).

Personality Assessment Inventory (PAI).

The PAI’s (Morey, 1991) Aggression subscales measured aggressive traits and beliefs and physically and verbally aggressive behaviors. The Aggressive Attitudes subscale assessed aggressive traits, such as difficulty controlling one’s anger, and aggression-related beliefs (e.g., the utility of aggression). The Physical Aggression subscale measured a propensity for physically aggressive behaviors such as physical fights or destroying property. The Verbal Aggression subscale examined a tendency towards verbal displays of aggression including yelling and abusive language. Additionally, The PAI’s Antisocial Features scale assessed antisocial personality disorder’s behavioral and characterological traits. The Negative Impression Management and Positive Impression Management subscales of the PAI assessed for attempts to present oneself in an extremely negative or positive manner. Scores for the scales were obtained by summing the items, rated on a 4-point scale (1=False to 4=Very True), and converting summations into T scores (M = 50, SD = 10). T scores of 70 or greater suggest clinically significance tendencies within the respective scale (Morey, 1991). The Aggression subscales, Impression Management subscales, and Antisocial Features scale demonstrated adequate or greater psychometric properties and predictive validity (Gardner et al., 2015, Karlin et al., 2005; Morey, 1991).

Data Analyses

Descriptive statistics were obtained using SPSS Version 25.0. Skewness and kurtosis of the study variables were examined, and no highly skewed variables were found. In regard to missing data, Little’s (1988) test of Missing Complete at Random (MCAR) was not significant, χ²(89) = 89.99, p = .54. Given prior concerns regarding the Observe facet within the FFMQ (Baer et al., 2008, Carpenter et al., 2019; Gillespie et al., 2018; Rudkin et al., 2018; Williams et al., 2014), we examined the bivariate associations between the Observe facet and other study variables as prior research had removed this facet when it was found to have nonsignificant or positive associations with aggression outcome variables (Peters et al., 2015). We found the Observe facet to have a significant negative relationship with alcohol use, aggressive attitudes, verbal and physical aggression, as well as a significant positive relationship with all of the other facets except nonreactivity. Furthermore, Carpenter et al. (2019) in their meta-analysis examining the mindfulness facets as they relate to psychological outcomes found that counterintuitive findings of positive associations between the Observe facet and psychological outcomes were not present among samples of older adults and non-student samples. Given that the current study examined the five facets of mindfulness within a clinical sample with a mean age of 41.15 and the significant negative correlations between the Observe facet and the aggression dependent variables, the Observe facet was retained in subsequent analyses.

Moderation analyses in Mplus Version 8.0 tested the relationships between alcohol use/problems, the five facets of mindfulness (i.e., Observe, Describe, Acting, Nonjudging, and Nonreactivity), and aggression (i.e., aggressive attitudes, verbal aggression, and physical aggression), while controlling for the effects of drug use/problems, age, positive and negative impression management, and antisocial features. Variables were mean centered prior to calculating the interaction terms. Missing data were handled using full information maximum likelihood (FIML) estimation, which is robust to issues of nonnormality and provides less biased estimates compared to alternative strategies such as listwise deletion (Kline, 2016).

Figure 1 illustrates the paths examined in the aggression model. Paths were examined simultaneously including: (a) regressive paths of aggressive attitudes, verbal and physical aggression onto alcohol use/problems and the five facets of mindfulness (b) regressive paths of aggressive attitudes, verbal and physical aggression onto the interaction terms between alcohol use/problems and the five facets of mindfulness (e.g., alcohol use/problems × Observe), (c) regressive paths of aggressive attitudes, verbal and physical aggression onto the covariates of drug use/problems, age, positive and negative impression management, and antisocial features, and (d) the bivariate relations between the alcohol use/problems, the five facets of mindfulness, and the covariates (drug use/problems, age, antisocial features), as well as the bivariate relations between aggressive attitudes, verbal and physical aggression. Significant interaction effects were decomposed in Mplus using the Johnson-Neyman technique as outlined by Hayes and Matthes (2009). Each significant interaction was explicated separately. The Johnson-Neyman procedure allows researchers to identify the exact level the moderator (i.e., the five facets of mindfulness) influences the relationship between the independent (i.e., alcohol use/problems) and dependent variables (i.e., aggressive attitudes, physical and verbal aggression).

Figure 1.

Paths tested in the aggression model. Covariate paths among predictor variables, with the exception of interaction terms, were included in the model but not presented for clarity. DUDIT = Drug Use Disorders Identification Test; PIM = Positive Impression Management; NIM = Negative Impression Management; AUDIT = Alcohol Use Disorders Identification Test.

Results

Table 2 includes bivariate correlations, means, and standard deviations among the study’s variables. AUDIT scores negatively and significantly associated with mindfulness facets of Observe and Acting. All five mindfulness had a significant negative relationship with aggressive attitudes, physical and verbal aggression, with the exception of Describe with verbal aggression and Nonreactivity with verbal and physical aggression. The three forms of aggression were all positively and significantly related to one another.

Table 2.

Summary of Bivariate Correlations, Means, and Standard Deviations of Study Variables

Measure	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1. Age	--
2. DUDIT	−.35†	--
3. Antisocial Features	−.38†	.55†	--
4. NIM	−.10*	.21†	.44†	--
5. PIM	.15**	−.28†	.54†	−.49†	--
6. AUDIT	.04	−.22†	.09	.15**	−.29†	--
7. Observe	.17**	−.26†	−.42†	−.44†	.52†	−.27†	--
8. Describe	.03	−.12*	−.07	−.22†	.25†	−.06	.42†	--
9. Acting	.07	−.12*	−.21†	−.46†	.46†	−.30†	.50†	.25†	--
10. Nonjudging	.05	−.14*	−.21†	−.24†	.31†	−.001	.30†	.38†	.24†	--
11. Nonreactivity	.02	−.06	.002	−.03	.01	.10	.09	.25†	−.05	.42†	--
12. Aggressive Attitudes	−.17†	.20†	.46†	.39†	−.51†	.23†	−.34†	−.21†	−.32†	−.38†	−.12*	--
13. Verbal Aggression	−.18†	.22†	.44†	.20†	−.33†	.11*	−.16**	−.08	−.14*	−.24†	−.11	.68†	--
14. Physical Aggression	−.25†	.35†	.57†	.46†	−.47†	.16**	−.34†	−.13*	−.29†	−.34†	−.09	.76†	.64†	--
M	41.15	8.54	56.60	52.84	48.40	16.86	30.15	26.91	28.84	21.70	25.05	49.56	49.65	52.78
SD	11.71	12.99	12.13	10.37	12.17	12.05	8.08	6.98	7.78	5.94	7.75	11.90	10.42	12.35

Note. AUDIT = Alcohol Use Disorders Identification Test; DUDIT = Drug Use Disorders Identification Test; PIM = Positive Impression Management; NIM = Negative Impression Management.

^*p < .05.

^**p < .01.

^†p < .001.

Results of the moderation model are included in Table 3. The model fit the data well: χ² = 64.26, p = .18; RMSEA = .02; CFI = .99; TLI = .99. AUDIT scores and anti-social features positively associated with aggressive attitudes, while positive impression management and the mindfulness facet of Nonjudging showed a negative relationship. Similar relationships were found with physical aggression, with the addition of positive associations between DUDIT scores and negative impression management with physical aggression. Anti-social features positively related to, and positive impression management negatively related to, verbal aggression.

Table 3.

Standardized Path Estimates for Moderation Model

	Outcome Variables
Predictors	Aggressive Attitudes	Verbal Aggression	Physical Aggression
Age	.07 (.05)	.08 (.05)	.02 (.04)
DUDIT	.04 (.06)	.04 (.07)	.14 (.06)*
Anti-social Features	.18 (.07)**	.30 (.07)***	.31 (.06)***
Negative Impression Management	.07 (.06)	−.04 (.06)	.20 (.05)***
Positive Impression Management	−.26 (.06)***	−.23 (.07)**	−.13 (.06)**
AUDIT	.15 (.05)**	.06 (.06)	.12 (.05)*
Observe	.03 (.06)	.12 (.07)	.07 (.06)
Describe	−.07 (.05)	−.01 (.06)	.01 (.05)
Acting	−.06 (.05)	−.04 (.07)	−.02 (.05)
Nonjudging	−.04 (.06)***	−.10 (.06)	−.19 (.05)***
Nonreactivity	−.11 (.05)	−.10 (.06)	−.02 (.05)
AUDIT x Observe	.12 (.06)*	.09 (.06)	.15 (.05)**
AUDIT x Describe	−.02 (.05)	.07 (.06)	−.06 (.05)*
AUDIT x Acting	−.17 (.05)**	−.12 (.06)*	−.11 (.05)*
AUDIT x Nonjudging	−.04 (.06)	.03 (.06)	−.06 (.05)
AUDIT x Nonreactivity	−.11 (05)*	−.13 (.06)*	−.04 (.05)

Note. AUDIT = Alcohol Use Disorders Identification Test; DUDIT = Drug Use Disorders Identification Test; Standard errors are in parentheses. Variables included in interaction terms were mean centered.

^*p < .05.

^**p < .01.

^***p < .001.

Regarding aggressive attitudes, the interaction terms of AUDIT x Acting and AUDIT x Nonreactivity had significant, negative associations, while the interaction term between AUDIT x Observe had a significant, positive association. Results of the Johnson-Neyman technique showed that AUDIT scores positively related to aggressive attitudes when levels of non-reactivity were below .16 SDs from the mean, B = .18, SE = .05, p = .001. In other words, AUDIT scores significantly and positively related to aggressive attitudes among those with non-reactivity scores below .16 SDs from the mean, but not above. In regard to the AUDIT x Acting interaction, AUDIT scores positively related to aggressive attitudes when levels of Acting were below .12 SDs from the mean, B = .19, SE = .05, p = .001, but not above. These findings suggest that AUDIT scores positively related to aggressive attitudes among male SUD patients who reported approximately average, or below average, levels of the mindfulness facets of Non-reactivity and Acting. For the AUDIT x Observe interaction, AUDIT scores positively related to aggressive attitudes when levels of Observe were above −.16 SDs from the mean, B = .18, SE = .06, p = .001, but not below. That is, AUDIT scores positively related to aggressive attitudes among male SUD patients with approximately average or greater levels of the mindfulness facet of Observe.

For verbal aggression, there was a significant positive association with AUDIT scores when Nonreactivity scores were below, but not above, −.26 SDs from the mean, B = .27, SE = .11, p = .012. The significant positive relationship between AUDIT scores and verbal aggression was found when levels of Acting were below, but not above, −.32 SDs from the mean, B = .24, SE = .10, p = .021. These findings suggest that AUDIT scores positively related with verbal aggression among male SUD patients who reported below average levels of the mindfulness facets of Acting and Nonreactivity.

For physical aggression, there was a significant and positive relationship between AUDIT scores and physical aggression when levels of Acting were below, but not above, .11 SDs from the mean, B = .15, SE = .05, p = .003. Similarly, the positive association between AUDIT scores and physical aggression was found for Describe scores below, but not above, .12 SDs from the mean, B = .15, SE = .05, p = .003. That is, AUDIT scores positively associated with physical aggression among male SUD patients who reported average, or below average, levels of Acting and Describe. Lastly, the mindfulness facet of Observe moderated the positively relationship of AUDIT and physical aggression among levels of Observe above, but not below, −.09 SDs from the mean, B = .16, SE = .05, p = .003. In other words, AUDIT scores had a significant positive relationship with physical aggression among male SUD who reported average, or above average, levels of Observe.

Discussion

We examined the five facets of mindfulness as moderators of the relationship between alcohol use/problems and aggressive attitudes, verbal and physical aggression among men in residential SUD treatment. Our findings showed that alcohol use/problems positively related to aggressive attitudes and physical aggression among men with average and below average levels of the mindfulness facet of Acting. A positive relationship was found between alcohol use/problems and verbal aggression but only among those with below average levels of Acting. Alcohol use/problems also had a positive association with aggressive attitudes among male SUD patients with average and below average levels of the mindfulness facet of Nonreactivity. Alcohol use/problems positively related to verbal aggression among those with below average levels of Nonreactivity. Alcohol use/problems positively associated with physical aggression among men with average, or below average, levels of the mindfulness facet of Describe. Lastly, the mindfulness facet of Observe had the opposite expected effect in that it strengthened the relationship between alcohol use and aggressive attitudes and physical aggression among those average, and above average, in Observe scores. The only mindfulness facet that did not moderate any of the three aggression models was Nonjudging.

These findings are consistent with, and extend, past research that men low in trait mindfulness have an increased risk of alcohol-related aggression (Gallagher et al., 2010), and that the association between alcohol use and aggression was found among those with deficits in the mindfulness facet of Acting (Ngo et al., 2018). Deficits in Acting with awareness result in engaging in tasks while in “autopilot” or with little awareness or attention to present moment behaviors (Baer et al., 2006). That is, individuals low in trait Acting with awareness tend to engage in behaviors based on automatic urges and impulses rather than conscious, thoughtful decisions. AMT suggested alcohol’s disinhibiting effect results in attentional resources being diverted towards aggression-instigating cues leading to an increased likelihood of aggressive behaviors (Giancola et al., 2010). It may be that men with greater alcohol use and/or problems who also exhibit low trait Acting with awareness, with their attentional resources focused on aggression-instigating cues, act on urges such as anger and hostility without awareness resulting in an increased likelihood of aggressive behaviors. This may also explain why the Nonjudging facet did not moderate the relationship between alcohol use and the three types of aggression. It may be that when facing aggression-instigating cues and acting on impulse, the judgement of one’s internal experiences is not a salient mitigating factor.

Our findings indicated that the Nonreactivity facet moderated the association between alcohol use and aggressive attitudes and verbal aggression. Men with average and low levels of Nonreactivity had a positive relationship between alcohol use and aggressive attitudes and beliefs while men with low levels of Nonreactivity had a positive relationship between alcohol use and verbal aggression. While no prior study examined the mindfulness facet of Nonreactivity as a moderator of the alcohol-aggression link, prior work supported that a greater tendency to impulsively react to experiences increased the risk of alcohol use resulting in aggression (Birkley et al., 2013; Kirwan et al., 2019; Watkins et al., 2014). Our findings suggested that Nonreactivity moderates the alcohol-aggressive attitudes and verbal aggression link, but not physical aggression. Prior research supported a negative association between mindfulness and aggressive attitudes and mindfulness and verbal aggression, as well as constructs associated with aggressive attitudes such as anger and hostility (Borders et al., 2010; Heppner et al., 2008; Shorey et al., 2015). Men in treatment for SUDs with deficits in Nonreactivity may have a tendency to react to distressing thoughts and feelings with emotional responses such as anger and hostility or verbal aggression, rather than physical displays of aggression. Given the positive association between aggressive attitudes and physical and verbal displays of aggression (Archer, 2004; Nunes et al., 2014; Stefanile et al., 2017), it is possible that aggressive attitudes may be a mechanism through which deficits in Nonreactivity lead to alcohol-related verbal and physical aggression. Further research is needed to replicate our findings and examine this potential underlying mechanism.

The mindfulness facet of Describe moderated the relationship between alcohol use and physical aggression, such that men with average or low levels of Describe had a positive association between alcohol use and physical aggression. The ability to verbally describe and differentiate thoughts and emotions is an important aspect of emotion regulation (Hill & Updegraff, 2012; Iani et al., 2019; Thompson & Waltz, 2010). Difficulties in emotional regulation is a well-supported correlate of aggressive behaviors (Garofalo et al., 2020; Roberton et al., 2012). It is possible that men with SUD with deficits in the mindfulness facets of Describe lack emotional regulation skills needed to suppress physically aggressive responses when experiencing aggression-instigating cues.

While at the bivariate level the Observe facet was found to have negative relationships with the aggression variables, results of the moderation analyses showed a positive effect of the mindfulness facet of Observe on the association between alcohol use and aggressive attitudes and physical aggression. Prior research showed the Observe facet to have positive correlations with physical aggression and anger and hostility, constructs associated with aggressive attitudes (Garofalo et al., 2020; Peters et al., 2015). It is possible that this finding of a positive effect of the Observe facet is due to problematic construct validity with this facet with its focus on observing external and bodily sensations rather than internal experiences, resulting in counterintuitive relationships between the facet and psychological phenomena (Bergomi et al., 2013; Goldberg et al., 2016; Rudkin et al., 2017). It is also possible that the myopic effect of alcohol, combined with high levels of the ability to observe bodily sensations may increase risk of aggression given prior research which suggested that physiological responses to anger positively related to reactive aggression (Zwets et al., 2014). However, researchers have urged caution when interpreting the Observe facet (Baer et al., 2008, Carpenter et al., 2019; Gillespie et al., 2018; Rudkin et al., 2018; Williams et al., 2014), and thus caution should be taken when interpreting the Observe results in the present study. Further research is needed to see if the positive effect of the Observe facet stands when other measures of observing are utilized.

Limitations and Future Research

There were limitations of the current study. The present study utilized cross-sectional methodology and thus temporal associations cannot be determined among the study variables. In order to assess for risk factors that increase the likelihood for alcohol-facilitated aggression, future research should consider longitudinal and experimental designs to elucidate directionality. This study utilized self-report measures which can result in common method bias, or an artificial inflation of the correlations between constructs when an individual completes measures in a single survey as a result of social desirability in responding, priming effects, and/or response styles (Mackenzie & Podsakoff, 2012). Thus, the methods utilized in the present study may result in correlations between variables that are inflated. Future research should consider utilizing collateral reports such as clinician assessments or participants completing measures at different points in time to address concerns regarding common method bias. Our sample consisted of primarily White men which limits the generalizability of the findings. Future research would benefit from examining the study variables within ethnically and gender diverse samples to examine whether these findings can be generalized to different populations. General aggression, as measured by the PAI, was examined in the present study which limits the conclusions regarding specific types of aggression that can be drawn from our findings. While this study aimed to examine the relationship between alcohol and aggression, the PAI aggression scales do not examine aggressive attitudes and behaviors as a result of alcohol use as they are measures of general aggression. Thus, no conclusions can be made regarding whether the aggression reported in the present sample was alcohol- facilitated. Furthermore, prior research supported the validity of the total Aggression scale of the PAI rather than the subscales (Gardner et al., 2015; Morey, 2007). Thus, caution should be taken when interpreting the results. Future work would benefit from examining specific forms of aggression, using validated behavioral measures of aggression, such as intimate partner violence, familial aggression, violence toward strangers, etc., to examine if our findings extend to these specific types of aggressive behaviors. In order to examine alcohol facilitated aggression specifically, future research should utilize measures that examine aggressive attitudes and behaviors while drinking alcohol or methodologies such as ecological momentary assessment to examine concurrent alcohol use and aggressive behaviors.

The present findings have important research implications despite these limitations. Our results provide preliminary support for the possibility that deficits in mindfulness facets of Acting and Nonreactivity may influence the strength of the relationship between alcohol use and aggression. Future researchers might consider investigating whether addressing trait mindfulness among men with SUDs may reduce aggressive attitudes and behaviors.