Positive emotional intensity and substance use: the underlying role of positive emotional avoidance in a community sample of military veterans

Abstract

Background:

Military veterans are at greater risk for substance misuse. Positive emotional intensity is one well-established antecedent of substance misuse in this population. Positive emotional avoidance, or attempts to alter the form, frequency, or context of positive emotions, may help to explain this association. While clinical practice typically aims to increase positive emotions, such approaches may have iatrogenic effects, as high-intensity positive emotions may be experienced as distressing and prompt avoidance for some populations. This suggests a need to better understand responses to positive emotions to inform clinical practice.

Objectives:

The goal of the current study was to advance theory, research, and clinical practice by exploring the role of positive emotional avoidance in the associations between positive emotional intensity and both alcohol and drug misuse. We hypothesized that positive emotional intensity would indirectly influence alcohol and drug misuse through positive emotional avoidance.

Methods:

Participants were a community sample of United States military veterans recruited through Amazon’s Mechanical Turk (n = 535, M_age = 37.45, 71.8% male, 69.5% White).

Results:

Correlations among positive emotional intensity, positive emotional avoidance, and alcohol and drug misuse were significant and positive (rs range from.13 to.41). Further, positive emotional avoidance was found to account for the relations of positive emotional intensity to alcohol (indirect effect: b =.04, 95%CI [.01,.08]) and drug misuse (indirect effect: b =.01, 95%CI [.01,.02]).

Conclusions:

Results provide preliminary support for the potential clinical utility of targeting avoidance responses to positive emotions in interventions targeting alcohol and drug misuse among military veterans.

Introduction

Substance misuse is a significant problem among United States (US) military veterans (1). US military veterans use substances at high rates, with 36.0% reporting past-month alcohol use (2) and 10.5% reporting past-year drug use (3). Further, elevated rates of substance use disorders (SUDs) have been documented in the literature, such that 42.2% and 20.0% of US military veterans meet criteria for alcohol use disorders (AUDs) and drug use disorders (DUDs), respectively, in their lifetime (4,5). This is of concern, as military veterans’ substance misuse is associated with a number of negative outcomes, including those related to worse physical health (e.g., illness and injury; 6, 7), more risky behaviors (e.g., violent/aggressive behavior and risky sexual behavior; 8, 9), greater legal consequences (e.g., arrest and incarceration; 10), and increased risk for suicide (11). Indeed, nearly 30% of completed suicides among US military personnel are preceded by alcohol or drug use (12). These high prevalence rates and deleterious outcomes highlight the critical need for research that improves our understanding of factors that may underlie the misuse of alcohol and other drugs among military veterans.

One such potential mechanism that has been underexplored and therefore requires more investigation is the role of positive emotions in substance misuse within this population. Positive emotions refer to mental experiences that are both intense and pleasurable (13). Frequent experiences of positive emotion have generally been shown to be related to positive outcomes, including widening the scope of attention and encouraging novel and varied thoughts and actions (14,15), increased creativity (16) and intuition (17), positive future self-rated health (18), and even increased lifespan (19). Positive emotions have also been suggested to influence health outcomes via the “undoing hypothesis,” which states that positive emotions are able to undo the physiological effects of stress (such as those which may be experienced in the context of military combat), allowing individuals to recover more quickly (20). On the other hand, extant literature has identified times when positive emotions may precipitate substance use. For instance, related studies have found that higher positive affect is related to alcohol consumption later in the day (21–24). Some theorists have conceptualized this temporal association as evidence for positive reinforcement models of substance misuse, with alcohol and other drugs functioning to elicit, enhance, or prolong positive emotional experiences (25,26). However, there may be other factors or pathways that have not been sufficiently explored; one example being emotional avoidance (i.e., attempts to alter the form, frequency, or context of emotional experiences; 27).

There are a number of possible explanations as to why military veterans may attempt to avoid positive emotions. It may be that, for some, intense positive emotions are experienced as aversive or threatening. Military veterans may experience secondary negative emotions to intense positive emotional stimuli (28,29), subsequently resulting in fear responses to positive emotions (30–32). For instance, physiological arousal originally associated with negative emotions may generalize to intense positive emotions (33), which may lead the individual to avoid positive emotions over time. Alternatively, military veteran’s appraisals of intense positive emotional experiences – including the extent to which they are viewed as undeserving (e.g., “I do not deserve to be happy”) or unpredictable (e.g., “Positive emotions are always short lived”) – may elicit distress (28). In turn, military veterans may misuse alcohol or other drugs in an attempt to escape or avoid aversive positive emotions (34,35), consistent with negative reinforcement models of substance misuse (36). However, no work to date has examined the role of positive emotional avoidance underlying the association between positive emotional intensity and substance misuse in general, nor among military veterans in particular.

Addressing these gaps, the current study explored the relations among positive emotional intensity, positive emotional avoidance, and substance (i.e., alcohol and drug) misuse among military veterans in the community. Investigations in this area with military populations have the potential to advance clinical practice and present a valuable contribution to the literature. Positive emotions are frequently overlooked in clinical interventions. When they are attended to, it is generally to attempt to increase positive emotional experiences (37), as the assumption is that doing so will result in positive outcomes. Existing evidence suggests that military veterans may be at increased risk for substance misuse in the context of intense positive emotions (35). This may be because the military socializes individuals to value qualities that promote avoidance of intense emotions, whether positive or negative (e.g., emotional control, stoicism; 38, 39). In turn, military individuals may be more likely to misuse alcohol or drugs to escape or avoid intense positive emotion states. Thus, clinical strategies that take alternative approaches to target intense positive emotions may be warranted for this population. Based on the past research, we hypothesized that positive emotional intensity, positive emotional avoidance, and alcohol and drug misuse would be significantly positively correlated. Further, we expected that positive emotional intensity would indirectly influence alcohol and drug misuse through positive emotional avoidance.

Methods

Participants and procedure

Military veterans were recruited from Amazon’s Mechanical Turk (MTurk), an internet-based crowdsourcing platform. MTurk is capable of generating reliable data (40,41), and represents the general population in terms of demographics (42) and prevalence of mental health problems (41). Participants were screened for the study based upon four inclusionary criteria (1): at least 18 years old (2), living in North America (3), working knowledge of the English language, and (4) a veteran of the US military. Participants who met eligibility criteria provided informed consent and completed the survey on Qualtrics (data collection platform). To improve data quality, we embedded validity checks in the MTurk survey assessing attentive responding and comprehension (n = 4; e.g., “I have never brushed my teeth;” 43–46) and military-specific knowledge (n = 2; i.e., “What is the acronym for the locations where final physicals are taken prior to shipping off for basic training,” and “What is the acronym for the generic term that the military uses for various job fields;” 47). To ensure the quality of the data, participants who failed to correctly respond to any of the six validity checks were excluded. Participants were compensated 2.50 USD for study participation. All procedures were approved by the Institutional Review Board at [redacted].

Exclusions and missing data

Of the obtained 2,644 responses, 997 participants were excluded for not meeting one or more inclusionary criteria (see Participants and Procedures; remaining n = 1,647). We then excluded 899 participants who failed to pass any of the four attention/comprehension validity questions (remaining n = 748), 134 participants who failed to pass either of the two military-specific validity questions (remaining n = 614), and 79 participants who attempted to complete the survey more than once (remaining n = 535). Thus, the final sample for the present study included 535 participants who were mostly male (71.8%) and White (69.5%), with an average age of 37.45 (SD = 11.24). See Table 1 for additional demographic information on the sample.

Table 1.

Sample characteristics.

	M (SD)	n (%)	Range
Age	37.45 (11.24)		18–76
Gender
Female		148 (27.9%)
Male		381 (71.8%)
Transgender		2 (0.4%)
Race
White		372 (69.5%)
Black		118 (22.1%)
Asian		28 (5.2%)
American Indian/Alaska		22 (4.1%)
Native
Native Hawaiian/Other		6 (1.1%)
Pacific Islander
Ethnicity
Hispanic or Latino/a		118 (22.1%)
Not Hispanic or Latino/a		417 (77.9%)
Employment Status
Employed full-time		451 (85.7%)
Employed part-time		48 (9.1%)
Not in labor force (student, homemaker)		15 (2.9%)
Unemployed		12 (2.3%)
Branch of service
Army		342 (63.9%)
Navy		51 (9.6%)
Air Force		102 (19.1%)
Marines		33 (6.1%)
Coast Guard		6 (1.1%)
Positive emotional intensity	29.93 (9.10)		6–48
Positive emotional avoidance	12.03 (5.93)		3–25
Alcohol misuse	9.23 (9.68)		0–35
Drug misuse	1.60 (2.61)		0–10
Negative emotional intensity	21.46 (7.06)		4–36

Note. Reported percentages represent valid percentages to account for missing data.

Measures

Short Affect Intensity Scale – Positive (SAIS-P; 48).

The SAIS-P is an 8-item scale used to measure positive emotional intensity. Participants indicated the extent to which they typically experience positive emotions using a 7-point Likert-type scale (1 = strongly disagree, 7 = strongly agree). Items were summed to create a total scale score, with higher scores representing greater positive emotional intensity. This scale has previously demonstrated excellent psychometric properties (48), and reliability in the current sample was excellent (Cronbach’s α = .93).

Emotional Avoidance Questionnaire – Positive (EAQ-Positive; 49).

The EAQ-Positive is a 5-item self-report measure assessing avoidance of positive emotions. Participants rated each item on a 5-point Likert-type scale (1 = not true of me, 5 = very true of me). Items were summed to create a total scale score, with higher scores indicating greater avoidance of positive emotions. The EAQ-Positive has been found to have adequate psychometric properties (49), and reliability in the current sample was excellent (Cronbach’s α = .91).

Alcohol Use Disorder Identification Test (AUDIT; 50).

The AUDIT is a 10-item self-report measure assessing alcohol consumption, drinking behaviors, adverse reactions to drinking, and alcohol-related problems. Participants rate each item using a 5-point Likert-type scale (0 = never, 4 = daily or almost daily). Higher scores indicate greater likelihood of alcohol misuse, with a recommended cutoff score of 8 or higher to probable AUD (50). The AUDIT has adequate psychometric properties (51), and reliability in the current sample was excellent (Cronbach’s α = .92).

Drug Abuse Screening Test (DAST; 52).

The DAST is a 10-item self-report measure that assesses the presence of problems related to drug use, such as occupational or relational problems, illegal activities, or regret. Responses to each item have 1 (yes) and 0 (no) options. Higher scores indicate greater drug misuse, with scores greater than or equal to 3 on the DAST indicating probable DUD (52). The DAST demonstrates good reliability and validity (52), and reliability in the current sample was excellent (Cronbach’s α = .91).

Covariates

Demographics.

Participants were asked to self-identify their age and gender identity. Research indicates that individuals who are younger and male report more alcohol and drug misuse (53,54).

Short Affect Intensity Scale – Negative (SAIS-N; 48).

The SAIS-N is a 6-item scale used to measure negative emotional intensity. Participants indicated the extent to which they typically experience negative emotions using a 7-point Likert-type scale (1 = strongly disagree, 7 = strongly agree). Items were summed to create a total scale score, with higher scores representing greater negative emotional intensity. This scale has previously demonstrated excellent psychometric properties (48), and reliability in the current sample was excellent (Cronbach’s α = .86). Research suggests an association between negative emotional intensity and alcohol and drug misuse (22,55,56).

Data analysis

As recommended by Tabachnick and Fidell (57), all study variables were assessed for assumptions of normality. Then, Pearson product-moment correlations were calculated among potential covariates (i.e., age, gender, negative emotional intensity) and primary study variables to explore their bivariate associations. In line with recommendations (58,59), those variables that were significantly correlated with alcohol and drug misuse were retained as covariates in subsequent analyses. Next, to address the question of whether positive emotional avoidance explains the relation between positive emotional intensity and alcohol and drug misuse, separately, we conducted indirect effect analyses (60) with the PROCESS SPSS macro (Model 4; 61). The PROCESS procedures use ordinary least squares regression and bootstrapping methodology, which confers more statistical power than do standard approaches to statistical inference and does not rely on distributional assumptions. Bootstrapping was done with 5,000 random samples generated from the observed covariance matrix to estimate bias-corrected 95% confidence intervals (CIs) and significance values (60,62). The effect is significant if the 95% confidence interval does not contain zero (60). Separate models were estimated to examine continuous scores representing severity of alcohol and drug misuse and to examine likelihood of belonging to the probable AUD and DUD groups (based on cutoff scores on the AUDIT and DAST, respectively; see Measures).

Results

All study variables of interest were approximately normally distributed based on benchmarks of skewness >2 and kurtosis >4 reflecting non-normality (63,64). Additionally, residuals were found to be normally distributed (skewness = .45 for alcohol misuse and 1.32 for drug misuse). Of the current sample, 43.4% (n = 232) reported probable AUD based on a cutoff score of 8 on the AUDIT (50) and 23.9% (n = 128) reported probable DUD based on a cutoff score of 3 on the DAST (52). Pearson product moment correlations among positive emotional intensity, positive emotional avoidance, alcohol misuse, and drug misuse are presented in Table 2. Positive emotional intensity was significantly positively associated with positive emotional avoidance as well as both alcohol and drug misuse. Positive emotional avoidance was also significantly positively associated with both alcohol and drug misuse.

Table 2.

Intercorrelations among the primary study variables of interest.

	1	2	3	4	5	6	7
1. Positive emotional intensity	-	.24**	.13*	.19**	.30**	.01	−.01
2. Positive emotional avoidance		-	.37**	.41**	.40**	−.15*	−.06
3. Alcohol misuse			-	.61**	.29**	−.05	−.06
4. Drug misuse				-	.25**	−.08	−.01
5. Negative emotional intensity					-	−.10	.14*
6. Age						-	−.02
7. Female gender							-

^*p <.01.

^**p <.001.

Models examining alcohol and drug misuse severity

The model explicating the role of positive emotional avoidance, controlling for the effect of negative emotional intensity, in the relation between positive emotional intensity and both alcohol and drug misuse, separately, is shown in Figure 1. In the model with alcohol misuse, the indirect effect of positive emotional intensity on alcohol misuse through the pathway of positive emotional avoidance was significant, b = .04, SE = .02, 95%CI [.01, .08], and the direct effect linking positive emotional intensity and alcohol misuse was nonsignificant after controlling for positive emotional avoidance, b = .004, SE = .05, t = 0.09, p = .93, 95%CI [−.09, .09]. The effect of negative emotional intensity on alcohol misuse was also significant, b = .23, SE = .06, t = 3.76, p < .001, 95%CI [.11, .35]. In the model with drug misuse, the indirect effect of positive emotional intensity on drug misuse through the pathway of positive emotional avoidance was significant, b = .01, SE = .01, 95%CI [.01, .02], and the direct effect linking positive emotional intensity and drug misuse was not significant after controlling for positive emotional avoidance, b = .02, SE = .01, t = 1.83, p = .07, 95%CI [−.002, .05]. The effect of negative emotional intensity on drug misuse was not significant, b = .03, SE = .02, t = 1.84, p = .07, 95%CI [−.002, .06].

Figure 1.

Summary of analyses explicating the role of positive emotional avoidance in the relations between positive emotional intensity and (separately) alcohol and drug misuse severity.

In analyses restricted only to those reporting alcohol and/or drug use (remaining n = 416, reflecting removing 111 abstainers), our pattern of findings remained the same. Regarding alcohol misuse, the indirect effect of positive emotional intensity through the pathway of positive emotional avoidance was significant, b = .08, 95%CI [.03, .12], while the direct effect was not significant, b = −.02, p = .66, 95%CI [−.12, .07]. Regarding drug misuse, the indirect effect of positive emotional intensity through the pathway of positive emotional avoidance was significant, b = .02, 95%CI [.01, .04], while the direct effect was not significant, b = .02, p = .18, 95% CI [−.01, .05].

Models examining likelihood of membership in probable AUD and DUD groups

The model explicating the role of positive emotional avoidance, controlling for the effect of negative emotional intensity, in the relation between positive emotional intensity and likelihood of membership in the probable AUD and DUD groups, separately, is shown in Figure 2. In the model examining likelihood of membership in the probable AUD group, the indirect effect of positive emotional intensity through the pathway of positive emotional avoidance was significant, b = .01, SE = .003, 95%CI [.003, .01], and the direct effect linking positive emotional intensity and likelihood of membership in the probable AUD group was nonsignificant after controlling for positive emotional avoidance, b = −.02, SE = .01, z = −1.36, p = .17, 95%CI [−.04, .01]. The effect of negative emotional avoidance on likelihood of membership in the probable AUD group was also significant, b = .06, SE = .02, z = 3.84, p < .001, 95%CI [.03, .09]. In the model examining likelihood of membership in the probable DUD group, the indirect effect of positive emotional intensity through the pathway of positive emotional avoidance was significant, b = .02, SE = .01, 95%CI [.01, .03], and the direct effect linking positive emotional intensity and likelihood of membership in the probable DUD group was nonsignificant after controlling for positive emotional avoidance, b = −.02, SE = .02, z = 1.51, p = .13, 95%CI [−.01, .05]. The effect of negative emotional avoidance on likelihood of membership in the probable DUD group was also significant, b = .05, SE = .02, z = 2.31, p = .02, 95%CI [.01, .08].

Figure 2.

Summary of analyses explicating the role of positive emotional avoidance in the relations between positive emotional intensity and (separately) likelihood of membership in probable AUD and DUD groups.

In analyses restricted only to those reporting alcohol and/or drug use (remaining n = 416, reflecting removing 111 abstainers), our pattern of findings remained the same. Regarding likelihood of membership in the probable AUD group, the indirect effect of positive emotional intensity through the pathway of positive emotional avoidance was significant, b = .02, 95%CI [.01, .03], while the direct effect was not significant, b = −.02, p = .08, 95%CI [−.05, .002]. Regarding the likelihood of membership in the probable DUD group, the indirect effect of positive emotional intensity through the pathway of positive emotional avoidance was significant, b = .02, 95%CI [.01, .04], while the direct effect was not significant, b = .02, p = .20, 95% CI [−.01, .05].

Discussion

A body of research has supported the notion that military veterans are at greater risk for misusing alcohol and other drugs in response to positive emotions (35). Extending this work, the goal of the present study was to better understand why intense positive emotions are associated with substance misuse for military veterans. Specifically, we sought to examine the role of positive emotional avoidance in the association between positive emotional intensity and both alcohol and drug misuse. As expected, positive emotional intensity, positive emotional avoidance, and alcohol and drug misuse were significantly and positively correlated at the bivariate level. Further, positive emotional avoidance was found to account for the relations of positive emotional intensity to alcohol and drug misuse severity and to likelihood of membership in the probable AUD and DUD groups. Specifically, greater positive emotional intensity was significantly associated with greater positive emotional avoidance, which, in turn, was significantly associated with greater alcohol and drug misuse severity and greater likelihood of belonging to the probable AUD and DUD groups. These results held even controlling for the effect of negative affect intensity, suggesting that findings are not simply tapping into intensity of emotional experience, but are specifically related to positively valanced emotional states. These findings have important implications for advancing clinical practice for substance misuse with military veterans.

Findings of the present study provide support for the relevance of positive emotional avoidance in the relation between positive emotional intensity and alcohol and drug misuse. These findings are well aligned with previous literature suggesting that individuals who experience high-intensity emotions may be motivated to avoid distress that is associated with these arousal states; in an effort to cope, alcohol and other drug use may be one strategy (65). For instance, there are strong permissive norms associated with substance misuse, and alcohol misuse in particular, among military personnel (66), such that substance use may be seen as a viable means by which to avoid undesirable emotional states. Such positive emotional states may be perceived by military veterans as portraying vulnerability (67). This, combined with military expectations around being stoic and emotionally controlled (38,39), may mean that military veterans experiencing high-intensity positive emotions misuse alcohol or other drugs as a way to dampen these emotional experiences. It also may be that military veterans are more likely to experience negative affect interference, or negative emotions in situations that typically elicit positive emotions (28). Thus, they may be more likely to engage in activities to alter positive emotions as a means to avoid the resulting negative states.

While preliminary in nature, findings of the present study have important implications for clinical practice with military veterans. Specifically, our results suggest that interventions aiming to indiscriminately increase positive emotions may have unintended adverse consequences for some individuals. Specifically, our findings suggest that more intense positive emotions may be associated with increased positive emotional avoidance, which in turn may lead to alcohol and drug misuse. Thus, clinicians working with military veterans may benefit from first assessing their client’s responses to positive emotions. Then, if the client reports adverse responses to positive emotions, these concerns can be addressed. For instance, individuals who report adverse responses to positive emotions may benefit from interoceptive exposures focused on positive emotions to allow for habituation to the elicited distress, which may indirectly decrease the client’s reliance on avoidance behaviors, such as alcohol and drug use (68). Alternatively, treatment approaches which seek to encourage the use of acceptance and mindfulness strategies to manage distress rather than avoiding or attempting to suppress stressors, such as Acceptance and Commitment Therapy (ACT) and Mindfulness-Based Cognitive Therapy (MBCT), may be useful. Indeed, studies examining the effectiveness of ACT and MBCT have found support for their use to address concerns related to both trauma exposure (69,70) and substance misuse (71,72), including among military veterans (73,74). Future work would benefit from assessing the utility of incorporating adverse responses to positive emotions in treatments targeting substance misuse among military veterans.

Results of the present investigation should be considered within the context of the study’s limitations. First, the cross-sectional and correlational nature of the data precludes the ability to make determinations about the causal and temporal patterns among the examined associations. Future research is needed to investigate the nature and direction of these associations through prospective, longitudinal designs. Second, this study relied exclusively on self-report measures, which may be influenced by one’s willingness and/or ability to report accurately. It may be that individuals have poor insight into their emotional experiences, or that their retrospective reports of substance use are over- or under-inflated. Indeed, previous work has found that individuals who misuse alcohol and other drugs may face difficulty reporting accurately on their emotion states (75,76). Future investigations should include objective (e.g., behavioral, physiological) measures of emotional responding (77,78,79). Third, collecting data via the Internet has disadvantages that may limit the generalizability of results, such as sample biases (e.g., self-selection) and lack of control over the research environment (e.g., unable to limit distractions, no opportunity to ask clarifying questions; 79). Thus, future research which integrates other data collection methods (e.g., interviewing, focus groups) is warranted. Finally, although our focus on military veterans in the community is a strength of this study, our findings cannot be assumed to generalize to other military (e.g., clinical) or nonmilitary populations. Further, the lack of a nonmilitary control group precludes the ability to draw definitive conclusions about the specificity of these associations among military veterans. Thus, investigations are needed to replicate our findings in larger, more diverse samples.

Despite these limitations, results of the current study extend our understanding of the role of emotional avoidance in the association between positive emotional intensity and substance misuse. Our findings suggest that existing models of emotional avoidance should incorporate avoidance of positive emotions, as it is related to clinically relevant outcomes such as alcohol and drug misuse. Further, they highlight the potential utility of targeting avoidance responses to positive emotions as a means to prevent and treat alcohol and drug misuse among military veterans in clinical practice. Extensions and replications of this work will enable refinement of clinical targets to address substance use among military veterans.