Effect of Trauma-Related Stress After Alcohol Consumption on Perceived Likelihood of Negative Consequences and Willingness to Drive

Abstract

Background:

Alcohol-related motor vehicle crashes are a major preventable cause of death in the United States. One potential factor that may modulate the influence of alcohol on driving-related cognitions and decision-making is trauma-related stress. Indeed, in addition to compelling research indicating that both acute trauma-related stress and acute alcohol consumption may independently affect driving-related risky decision-making, there is reason to believe that the combination of these antecedents may have an exacerbating effect.

Methods:

The current study evaluated the influence of induction of acute trauma-related stress (via script-driven imagery) after alcohol consumption (.06% Breath Alcohol Concentration [BrAC]) on driving-related cognitions – perceived likelihood of negative consequences and willingness to drive – among 25 trauma-exposed (currently symptomatic) adult drinkers from the community (M = 24.08; 36.0% female).

Results:

Participants who were acutely exposed to trauma-related stress after alcohol consumption evidenced lower perceived likelihood of being pulled over by a police officer (η_p2 = 0.38, large effect size) and lower perceived likelihood of getting in an accident (η_p2 = 0.17, medium-to-large effect size) relative to participants exposed to a neutral cue; conversely, participants exposed to trauma-related stress after alcohol consumption evidenced greater willingness to drive (d = 1.16, large effect size) than participants exposed to a neutral cue.

Conclusions:

Generally, findings suggest that individuals with a trauma history that are acutely exposed to trauma-related stressors (e.g., reminders of their traumatic experience) may be particularly vulnerable to poorer driving-related decision-making after alcohol consumption. Results provide a meaningful target for the development of intoxicated driving prevention and intervention efforts geared specifically for individuals with trauma history.

1. Introduction

Alcohol-related motor vehicle crash (MVC) fatalities are a major preventable cause of death in the United States (US). Recent reports indicate that over 300,000 individuals drive under the influence of alcohol each day [1], resulting in more than 10,000 deaths annually and costing upwards of $52 billion in damages annually [2]–[4]. As such, an extensive literature has sought to delineate risk for driving under the influence of alcohol [5], as well as evaluate the efficacy of awareness campaigns and prevention efforts [6], [7]. However, despite laboratory-based research supporting longitudinal and cross-sectional associations between moderate to high levels of alcohol consumption (e.g., blood alcohol concentration [BAC] > .05%), risky decision-making, and impaired driving performance [8]–[11], little work has examined factors that may modulate the effects of alcohol use on driving-related risky decision-making.

One factor consistently linked to problematic alcohol use and driving under the influence of alcohol is trauma and trauma-related stress (e.g., posttraumatic stress disorder [PTSD] symptomology) [12]. Indeed, PTSD is one of the most commonly reported psychiatric disorders among those arrested for ‘driving under the influence’ (DUI) [13], with research indicating that individuals evidencing posttraumatic stress symptoms are more likely to drink and drive [14], and are at increased risk for DUI recidivism [15]. Further, recent work has found that individuals who have experienced trauma and report PTSD symptoms evidence an elevated propensity to engage in risky decision-making, including both general (e.g., thrill seeking) and alcohol-related risky behaviors (e.g., driving under the influence of alcohol) [16], [17]. Notably, national sample estimates indicate that between 31–41% of those experiencing trauma endorse increases in some risky/reckless behaviors after trauma exposure [18], [19], culminating in the addition of “reckless or self-destructive behavior” as a PTSD symptom in the updated Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [20].

Although extant work indicates that both alcohol consumption [8] and trauma-related stress [21] independently affect driving-related cognitions and decision-making, there is reason to believe that the combination of these antecedents may have a synergistic effect. First, recent reviews of the literature suggests that acute alcohol consumption – resulting in blood alcohol concentrations both below (i.e., .05%) and above the per se legal BAC threshold for DUI in the US of .08% – reliably contribute to marked impairment to motor and cognitive skills necessary for safely operating a vehicle [11], leading to numerous calls reduction of the .08% BAC threshold in the US [10], [22], [23]. Further, extant laboratory research indicates that acute alcohol consumption may influence driving-related risk behaviors by reducing perceptions and expectations of likelihood of negative consequences [8]. More specifically, individuals tend to rate negative consequences (e.g., getting in an accident) as less likely as more alcohol is consumed (e.g., ascending towards .08% BAC) [8]; therefore, some may be more willing to engage in risk-taking behavior, such as driving under the influence of alcohol, because they believe less risk is involved [24]. Second, early laboratory-based work suggests that acute stress may reduce perception of intoxication, purporting that stress attenuates some psychopharmacological impacts of alcohol at moderate to high levels of alcohol consumption (e.g., .05% BAC and above) [25]. However, recent research indicates that reduced perceptions of intoxication – which may be modulated by stress – are not related to actual improvements in associated functional performance, such as driving performance [26]. Thus, individuals experiencing acute stress may engage in riskier behaviors because they believe that the psychopharmacological effects of alcohol (e.g., reduced reaction time) are diminished – an assertion supported by research indicating that willingness to engage in driving under the influence of is impacted by lower subjective ratings of intoxication [24], [27]. This work suggests that both alcohol consumption and stress experienced after alcohol consumption may jointly affect driving-related cognitions and risky decision-making.

Finally, an extensive literature has linked trauma-related stress [28], [29] to avoidance coping and behaviors, which are often characterized by disengagement from the source of distress (e.g., leaving that location) [30]. Experimental work also indicates that acute induction of stress can lead to a bias towards short-term gains (e.g., immediate avoidance of stressor) [31] and a tendency to hastily make decisions before alternative options have been considered (e.g., alternative transportation) [32]. Laboratory-based work also demonstrates that this reactivity to both general and trauma-related stress is more pronounced among trauma-exposed individuals with PTSD symptoms [21], [33] as compared to both asymptomatic and non-trauma exposed individuals. Altogether, these findings suggest that individuals evidencing PTSD symptoms who are acutely exposed to trauma-related stimuli (e.g., reminders of a traumatic event) after drinking may be more likely to engage in immediately-reinforcing avoidance behavior, such as leaving the situation and driving home, despite their level of alcohol consumption.

Despite this compelling body of research, no work has experimentally assessed the influence of both acute alcohol consumption and trauma-related stress on subsequent driving-related cognitions, such as perceptions of driving-related negative consequences and willingness to drive. The current study expands the literature by utilizing a controlled between-subjects, experimental design to evaluate the following hypotheses. Drawing from extant work [8], [24], it was hypothesized that (1) participants who were exposed to a trauma cue (via script-driven imagery) following alcohol consumption would endorse lesser perceived likelihood of negative consequences (i.e., getting pulled over by a police officer, getting in an accident) than participants exposed to a neutral cue after alcohol consumption; and (2) that participants who were exposed to a trauma cue after alcohol consumption would endorse greater willingness to drive than participants exposed to a neutral cue after alcohol consumption. Notable, it was hypothesized that these differences would be robust to the empirically-derived inclusion of important covariates that have been identified in the extant literature to be associated with alcohol-related problematic driving behaviors (i.e., alcohol tolerance [26], [34], impulsivity [35], [36], and perceived dangerousness of driving after consuming alcohol [37]).

2. Materials and methods

2.1. Participants

The final sample comprised 25 community-recruited adults who were at or above the legal drinking age in the US; reported experiencing at least one DSM-5 [20] Criterion A traumatic event; endorsed at least minimal severity on at least one DSM-5 Criteria B-E PTSD symptom in the past 30 days (i.e., not asymptomatic); and did not have a current PTSD diagnosis and/or were not currently seeking PTSD treatment. Alcohol-related inclusion criteria were: consumption of at least four (for women) or five (for men) standard alcoholic beverages within a two-hour period at least once in the past year; willingness to consume alcohol in the laboratory; no history of allergic/adverse reactions to alcohol; no medical history/condition or use of medications for which alcohol use is contraindicated; ability to refrain from medication use for 24 hours prior to study and food for three hours prior to study; not currently completing/seeking or in need of treatment for substance use problems. Of 62 participants recruited for a larger parent study on risk-taking behavior, four participants were excluded due to alcohol-related criteria; two were excluded for trauma-related criteria; and 30 were excluded due to randomized assignment to placebo conditions meant for evaluation of variables outside the scope of the current study.¹ See Table 1 for full descriptives of the final sample and for each condition.

Table 1.

Descriptive Statistics for Total Sample, Descriptives for Condition Subgroups, and Conditional Comparisons

Variable	Total Sample (N = 25)	Alcohol/Neutral (n = 12)	Alcohol/Trauma (n = 13)	Comparative Analyses
				Test	p
Age	24.08 ± 3.23	23.58 ± 2.50	24.54 ± 3.82	t = 0.73	.471
Range	21–32	21–27	21–32
Biological Sex (Female)	9 (36.0%)	4 (33.3%)	5 (38.5%)	χ2 = 1.89	.170
Racea				χ2 = 0.53	.467
African-American	3 (12.0%)	3 (25.0%)	-
White	15 (60.0%)	7 (58.3%)	8 (61.5%)
Hispanic/Latinx	5 (20.0%)	2 (16.7%)	3 (23.1%)
Other	2 (8.0%)	-	2 (15.4%)
Alcohol Tolerance	6.88 ± 4.46	5.50 ± 3.26	8.15 ± 5.13	t = 1.53	.140
Index Trauma Typeb				χ2 = 1.47	.225
Transportation	1 (4.0%)	-	1 (7.7%)
Serious Accident	1 (4.0%)	-	1 (7.7%)
Physical Assault	6 (24.0%)	2 (16.7%)	4 (30.8%)
Assault with Weapon	1 (4.0%)	1 (8.3%)	-
Sexual Assault	7 (28.0%)	4 (33.3%)	3 (23.1%)
Other Sexual Experience	3 (12.0%)	2 (16.7%)	1 (7.7%)
Combat Exposure	1 (4.0%)	1 (8.3%)	-
Life-Threatening Injury	1 (4.0%)	1 (8.3%)	-
Sudden Violent Death	1 (4.0%)	-	1 (7.7%)
Sudden Accidental Death	3 (12.0%)	1 (8.3%)	2 (15.4%)
PTSD Symptom Severity	24.32 ± 18.52	22.92 ± 16.05	25.62 ± 21.12	t = 0.36	.724
Range	0–68	4–62	0–68
Impulsivityb
Negative Urgency	10.43 ± 2.93	11.24 ± 2.24	10.00 ± 3.38	t = 0.08	.452
Lack of Perseverance	11.85 ± 1.23	11.13 ± 1.36	12.33 ± 0.89	t = 2.42	.026
Lack of Premeditation	11.00 ± 1.65	11.00 ± 1.87	11.00 ± 1.54	-	-
Sensation Seeking	8.30 ± 2.94	8.44 ± 2.96	8.18 ± 3.06	t = 0.19	.849
Positive Urgency	11.75 ± 3.45	11.11 ± 3.14	12.27 ± 3.74	t = 0.74	.468
Perception of Dangerousness of Driving After Alcohol	86.96 ± 20.88	90.17 ± 14.62	84.00 ±25.62	t = 0.73	.472

Note: Data presented as M ± SD and n (%).

^aDue to small sample size, participants identifying as Asian (n = 1; 4.0%), Native American (n = 1; 4.0%) were combined in the Other category for analysis.

^bDue to small sample size, data were dichotomized by non-interpersonal (i.e., transportation, serious accident, life-threatening injury, sudden violent death, sudden accidental death) vs. interpersonal (i.e., physical assault, assault with a weapon, sexual assault, other sexual experience, combat exposure) index trauma for analysis

^cData not available for four participants on Negative Urgency (16.0%) subscale, five on Lack of Perseverance (20.0%), five on Lack of Premeditation (20.0%), five on Sensation Seeking (20.0%), and five on Positive Urgency (20.0%).

2.2. Measures

2.2.1. History of Alcohol Consumption and Alcohol Tolerance

Frequency of past-month alcohol use was evaluated concurrently with history of alcohol consumption via the timeline followback procedure (TLFB) [38]. Participants identified specific days they drank alcohol and recalled the number of drinks consumed on each of those days in standard alcohol beverages. The TFLB has high reliability [38] and has been used extensively in community-recruited trauma samples [39]. Given no validated self-report assessment of alcohol tolerance in the literature, the current study also used past-month alcohol use frequency as a proxy for alcohol tolerance.

2.2.2. Trauma Exposure and Index Trauma Type

The Life Events Checklist for DSM-5 (LEC-5 [40]) was administered to establish the presence of a DSM-5 Criteria A traumatic event [20]. The original LEC was developed concurrently with the Clinician-Administered PTSD Scale (CAPS). The LEC-5 consists of 16 specified potentially traumatic events, as well as one unspecified traumatic event. Sample items include: “Life-threatening illness or injury” and “Transportation accident (for example, car accident, boat accident, train wreck, plane crash).” Participants were additionally asked to index their most “impacting” trauma from their responses in the LEC-5 (i.e., index trauma), prompted by the question, “Which event have you identified as your worst event, which currently bothers you most?” This measure has been used extensively in research focused on trauma and substance use related outcomes [41]–[43].

2.2.3. Impulsivity

The short form of the UPPS-P Impulsive Behavior Scale (SUPPS-P) [44] is a 20-item, self-report questionnaire developed to measure five personality facets (four-items each) related to impulsive behavior: negative urgency, lack of perseverance, lack of premeditation, sensation seeking, and positive urgency. Samples items include: “When I am upset I often act without thinking” for negative urgency; “I generally like to see things through to the end” for lack of perseverance; “My thinking is usually careful and purposeful” for lack of premeditation; “I quite enjoy taking risks” for sensation seeking; and “I tend to act without thinking when I am really excited” for positive urgency. Participants responded to each question on a four-point, Likert-type scale from 1 (“strongly agree”) to 4 (“strongly disagree”). After accounting for reverse scoring, higher scores indicated greater impulsivity. The SUPPS-P evidences good psychometric properties [44] and has been used in trauma-exposed populations [45], [46] and in substance use research [47], [48]. Each subscale demonstrated good internal consistency in the present study (negative urgency α = .69, lack of perseverance α = .81, lack of premeditation α = .86, sensation seeking α = .73, positive urgency α = .82).

2.2.4. Perception of Dangerousness of Driving After Consuming Alcohol

Perception of dangerousness of driving after consuming alcohol were assessed via a single-item, face-valid question [37]. Participants were asked, “How dangerous do you feel it is for you to drive after consuming five standard alcoholic drinks in a two-hour period?” Responses range on a scale from 0 (“not at all”) to 100 (“extremely”).

2.2.5. Likelihood of Negative Consequences

Participants responded to two single-item, face-valid questions regarding likelihood of (1) “being stopped by a police officer” and (2) “being in an accident (e.g., hit a person/ object/another vehicle)” if they attempted to drive, given their current level of intoxication. Responses ranged on an analogue scale from 0 (“extremely unlikely”) to 100 (“extremely likely” [24].

2.2.6. Willingness to Drive

Following alcohol administration, participants responded to a single, face-valid question, “If given the option to leave the experiment, how willing would you be to drive yourself home right now at your current level of intoxication?” Responses ranged on an analogue scale from 0 (“extremely unwilling”) to 100 (“extremely willing”) [37].

2.3. Procedure

2.3.1. Trauma-Related Stress Induction

A script-driven imagery procedure was used to induce trauma-related or neutral cognitions/ affect. Following script development procedures [49], [50], participants completed a script preparation form where they were asked to provide a detailed written description of their most traumatic autobiographical event or a neutral autobiographical event (e.g., brushing teeth, doing laundry), including selecting up to five subjective visceral/muscular reactions (e.g., “heart racing,” “sweating,” “watery eyes/tears”) that accompanied their experience. Following script development, a research assistant recorded a 30-second script portraying that autobiographical experience in second person, present tense. Scripts were recorded in a neutral, gender-matched voice. For the script-driven imagery session, participants listened to a two-minute recording comprised of: (1) a brief introduction to the procedure (e.g., close eyes); (2) a 30-second relaxation baseline period; (3) the 30-second script; and (4) instructions to continue to imagine the experience for 30 seconds after the script had finished.

2.3.2. Alcohol Administration

Participants consumed an alcohol mixture that was three parts cranberry juice, one part 100-proof vodka (divided into three equal portions) containing .520 or .612g/kg of vodka, based on participants’ biological sex and weight [51]. Participants were given up to three minutes to finish each drink [51]; absorption periods were idiographically-determined by assessing BrAC every four minutes until participants reached the criterion BrAC (i.e., .05%) [52]. Generally, this criterion BrAC was chosen to ensure that participants were as close to the target peak BrAC (i.e., .06%) as possible while completing the driving-related outcome measures, given the addition absorption time that would occur in transitioning back to the experimenter room and completing the script-driven imagery session (see 2.3.3. Procedural Overview). In an effort to reduce expectancy effects, participants were not provided any information on their current or anticipated BrAC until after the study had concluded.

2.3.3. Procedural Overview

Approval from the institutional review board was obtained for the study. Participants were recruited via advertisements in the local community, screened for trauma- and alcohol-related eligibility criteria over the telephone, and provided relevant medical and dietary restrictions (i.e., no medications or alcohol for 24 hours prior to study, no food for three hours prior to study) and procedures for the study. Interested individuals meeting initial eligibility criteria were brought into the laboratory, consented, provided an overview and signed a Behavioral Contract (i.e., indicating they would not operate machinery or a vehicle for three hours after leaving the study; providing contact information for their pre-arranged designated ride home from the study), re-screened for eligibility criteria, and randomly assigned to either a trauma or neutral script condition. Participant then completed a series of baseline measures (e.g., sociodemographics) and a script developmental procedure. Following baseline assessment, participants completed an alcohol administration procedure that involved participants consuming three pre-mixed drinks over nine minutes, followed by an idiographically-determined absorption period in a designated ‘bar room’ – an experimenter room that cloth-covered bar and two bar stools that played bar-themed background noise. Once participants had met or surpassed the criterion BrAC (.05%) and were ascending towards the target peak BrAC (.06%), participants transitioned back to the original private participant room to start the two-minute (neutral or trauma) script-driven imagery procedure. At the end of the script-driven imagery procedure, participants completed the drinking-related outcome measures (i.e., perceived likelihood of negative consequences and willingness to drive) – given anticipated BrAC curve estimates, it was expected that participants were completing the driving-related outcome measures at their peak target BrAC (.06%). Following completion of the outcome measures, participants had their BrAC assessed every 15 minutes until their BrAC reached a predetermined safe threshold for release (i.e., ≤ .03%). Participants were then debriefed and compensated $30 for participation and taken to their pre-arranged transportation escort.

2.4. Data Analytic Plan

2.4.1. Preliminary Analyses

A manipulation check for the trauma-related stress induction procedure indicated that individuals in the trauma condition evidence significantly greater level of subjective destress (i.e., via scores on a Subjective Units of Distress Scale [SUDS][53]) than individuals in the neutral script condition (t = 2.58, p = .018). Sociodemographic (i.e., biological sex, age, race/ethnicity) and theoretically-relevant psychological/behavioral characteristics (i.e., alcohol tolerance, impulsivity, perceived dangerousness of driving after consuming alcohol) were examined in association with the primary outcome measures. All such variables that were significantly associated with a given primary outcome measure (i.e., p < .05) were included in the primary analyses as covariates.

2.4.2. Primary Analyses

Univariate analyses of covariance (ANCOVA) examined if participants in the trauma script condition endorsed (1) lower perceived likelihood of being stopped by a police officer, (2) lower perceived likelihood of getting in an accident, and (3) greater willingness to drive than participants in the neutral script condition after consuming alcohol, while controlling for statistically significant sociodemographic and theoretically-relevant psychological/behavioral characteristics. Covariates and condition variables were entered into the model simultaneously. If no covariates were identified, independent-samples t-tests were used. Given limited missing data, listwise deletion was used.²

3. Results

3.1. Preliminary Analyses

Group randomization appeared to be effective, as conditions did not significantly differ across age, biological sex, race/ethnicity, indexed trauma type, or alcohol use history. Results indicated that greater negative urgency, sensation seeking, and perceived dangerousness of driving after consuming alcohol were significantly associated with higher perceived likelihood of being stopped by a police officer. Further, results indicated that greater perceived dangerousness of driving after consuming alcohol was associated with higher perceived likelihood of getting into an accident. No covariates were associated with willingness to drive. Full correlational associations between variables can be found in Table 2.

Table 2.

Correlation Matrix for Sociodemographic, Covariate, and Driving-Related Outcome Measures

	Sociodemographics/Covariates										Driving-Related Cognitions
Variables	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.
Sociodemographics/Covariates
1. Age	-
2. Biological Sex	.05	-
3. Race	.03	−.58**	-
4. Alcohol Tolerance	−.21	.46*	−.50*	-
5. Negative Urgencya	.60**	.21	−.28	−.21	-
6. Lack of Perseverancea	−.16	−.09	.27	.04	−.44	-
7. Lack of Premeditationa	.11	−.07	.20	−.13	.34	−.17	-
8. Sensation Seekinga	.38	−.44	.47*	−.36	.13	.18	.10	-
9. Positive Urgencya	.34	−.02	.06	−.29	.56**	−.09	.57**	.20	-
10. Perception of Dangerousness of Driving After Alcohol	.22	−.43*	.12	−.24	.40	.21	.03	.35	.27	-
Driving-Related Outcomes
11. Likelihood of Being Pulled Over by Police	.10	.01	−.04	−.06	.38	−.13	.21	.23	.24	.42*	-
12. Likelihood of Accident	.21	−.16	.14	−.34	.45*	−.14	.13	.49*	.16	.44*	.83**	-
13. Willingness to Drive	.29	−.25	.38	−.15	−.34	.10	−.03	.18	.08	−.34	−.46*	.53*	-

Note. Perceived likelihood of being pulled over by police, perceived likelihood of accident, and willingness to drive was evaluated on 0 to 100 scale with higher scores indicated greater likelihood or willingness.

^aParticipants were categorized as “White” vs. “Non-White” for correlational analyses.

^bData not available for four participants on Negative Urgency (16.0%) subscale, five on Lack of Perseverance (20.0%), five on Lack of Premeditation (20.0%), five on Sensation Seeking (20.0%), and five on Positive Urgency (20.0%).

^*p < .05

^**p < .01

3.2. Primary Analyses

3.2.1. Likelihood of Being Stopped by a Police Officer.

Results of the one-way ANCOVA indicated that participants in the trauma condition evidenced significantly lower perceived likelihood of being stopped by a police officer if they were to drive than participants in the neutral condition (F = 9.27, p = .008, η_P² = .38, large effect size) after controlling for negative urgency, sensation seeking, and perceived dangerousness of driving after consuming alcohol.²

3.2.2. Likelihood of Accident.

Results of the one-way ANCOVA indicated that participants in the trauma condition (M = 25.08) evidenced significantly lower perceived likelihood of getting in an accident if they were to drive than participants in the neutral condition (M = 51.92; F = 4.34, p = .049, η_P² = .17, medium-to-large effect size) after controlling for perceived dangerousness of driving after consuming alcohol.

3.2.3. Willingness to Drive.

Given identification of no empirically-derived covariates, an independent-samples t-test was used for the willingness to drive analysis. Results indicated that participants in the trauma condition (M = 51.36) evidenced significantly greater willingness to drive than participants in the neutral condition (M = 14.88, SD = 23.13), t(17) = 2.43, p = .026, d = 1.16, large effect size]. Full descriptive and analytic information regarding likelihood of negative consequences and willingness to drive outcomes can be found in Table 3.

Table 3.

Results for Primary Analyses on Driving-Related Outcomes by Condition

	Condition		Statistics
Analysis	Alcohol/Neutral (n = 12)	Alcohol/Trauma (n = 13)	Test Statistics	p	Effect size
Likelihood of Being Pulled Over by Police a	54.78 ± 26.79	23.45 ± 23.08	F = 9.27	.008	ηp2 = 0.38
Likelihood of Accidentb	51.92 ± 34.58	25.08 ± 24.01	F = 4.35	.049	ηp2 = 0.17
Willingness to Drive	14.88 ± 25.21	51.36 ± 36.43	t = 2.43	.026	d = 1.16

Note. Data presented as M ± SD. Likelihood of Being Pulled Over by Police, Likelihood of Accident, and Willingness to drive were evaluated on 0 to 100 scale. Thresholds for effect size interpretation (Cohen, 1988; Vacha-Haase & Thompson, 2004): Cohen’s d = .20 (small), .40 (medium), .80 (large); η² = .01 (small), .06 (medium), .14 (large).

^aCovariates for Likelihood of Being Pulled Over by Police were negative urgency, sensation seeking, perceived dangerousness of driving after consuming alcohol. Five participants were excluded from primary analyses that included covariates due to missing data on negative urgency.

^bCovariates for Likelihood of Accident were perceived dangerousness of driving after consuming alcohol.

4. Discussion

Building upon prior findings [8], [24], the current study indicated that trauma-related stress may lessen perceived likelihood of negative consequences (i.e., being stopped by a police officer, accident), above and beyond the influence of acute alcohol consumption. These findings may be partially explained by potential alterations in arousal and reactivity (AAR; e.g., hypervigilance) linked to re-exposure to trauma-related stimuli [20], [54]. More specifically, research suggests that trauma-related stress may trigger specific AAR symptoms, manifesting in increased visual attention towards threatening stimuli (e.g., impending red light) [55] and increased vigilance (cf. avoidance) when visually processing information [56]. Consequently, individuals may be misinterpreting this heightened vigilance as ‘sobriety’ (i.e., perceived diminishing of pharmacological effects) which may, in turn, decrease their perceived likelihood that negative driving-related consequences may occur, despite actual levels of intoxication [57], [58]. This assertion is generally supported by extant laboratory-based research indicating that stress may attenuate perceptions of the psychopharmacological impact of alcohol [25], [59].

Consistent with laboratory-based research on driving-related decision-making [60], findings from the current study also indicated that trauma-related stress may exacerbate willingness to drive under the influence of alcohol. Similar to the finding related to driving-related negative consequences, this may be due to perceived attenuation – and subsequent misinterpretation – of the pharmacological effects of alcohol following acute re-exposure to trauma-related stimuli [25], [55]. Further, decision-making research on both general and trauma-related stress may partially explain this finding. For example, research across populations indicates that stress negatively impacts general decision-making [61], such as eliciting bias towards short-term gains (e.g., immediately disengaging from the source of distress) [31], and offering solutions and making decisions before all options have been considered (e.g., alternative transportation) [32] – all of which may lead to increased willingness to engage in driving under the influence of alcohol.

A confluence of theoretical research highlighting cognitive changes that occur in response to trauma-related stimuli may also serve to partially explain these findings [62]. For example, research focused on models of tension-reduction [63] and mood maintenance [64] indicate that individuals may be more willing to engage in risky behaviors, such as driving under the influence of alcohol, to reduce threat-related emotions (via disengagement) caused by trauma-related stimuli [65], [66]. Other cognitive research posits that trauma exposed individuals may be willing to engage in riskier behaviors in an effort to enhance self-related cognitions, wherein individuals may weigh short-term rewards (e.g., disengaging from the source of trauma-related stress) more strongly when that risky behavior involves some form of potential self-enhancement (e.g., perceived control over the situation) [67]. As such, individuals acutely exposed to trauma-related stimuli may be more willing to drive under the influence of alcohol as a maladaptive means of avoidance coping - attempting to both alleviate distress, as well as enhance their self-perception of control via behaviorally disengaging from the source of the distress [62], [67].

Lastly, theoretical models of risk behavior may help to partially explain the influence of trauma-related stress on willingness to drive. For example, the universal theory of planned behavior (TPB) [68] posits that reasoned decision-making and behavioral intentions are the main determinants of behavioral outcomes. However, this broad modeling of behavior struggles in explaining ‘intention-behavior’ disparities, which often occur in instances where reasoned decision-making is hindered [9], [69]. As such, dual process models examining risky behaviors have been developed [70], [71], purporting that there may be alternative decision-making pathways that are triggered by various cognitive and affective states, leading to spontaneous risk decision-making (e.g., subconscious). Results from the current study support work indicating that negative affective states [72], [73] – resulting from exposure to trauma-related stimuli – may serve as an exacerbating mechanism for more spontaneous risky decision-making pathways, above and beyond the influence of acute alcohol consumption [70], [74]. Importantly, the current study did not directly examine or evaluate the rationale underlying driving-related cognitions. Future work aiming to better understand the influence of trauma-related stress after alcohol consumption should consider more detailed evaluation to identify the specific underlying mechanisms that may be impacting proximal driving-related cognitions.

4.1. Implications

Current study findings have notable theoretical and clinical implications. Findings underscore the complexity of understanding intoxicated driving, including highlighting important discrepancies between various driving-related cognitions. Understanding these unique driving-related cognitions may be imperative to education and intervention efforts, as individuals may extend or extrapolate potential improvements in one area of driving-related risk (e.g., increased vigilance/reaction time) to other areas that are impaired. Future work incorporating more direct and ecologically-valid assessments of driving-related cognitions are needed to advance scientific understanding of which aspects of driving may be exacerbated or mitigated by trauma-related stress, ultimately aiding in the continued development of education and intervention efforts targeting the reduction of incidences and negative outcomes (e.g., recidivism, death) related to intoxicated driving.

Further, findings may provide some meaningful contributions to education efforts. Specifically, there have been numerous calls highlighting the importance and efficacy of broad education on prevention and intervention of intoxicated driving [75]. Current study results provide additional information for practitioners working with symptomatic trauma-exposed individuals, indicating that trauma-related stimuli (e.g., thoughts related to the trauma) may negatively impact driving-related decision-making. This ‘awareness’ of (potential) behavior has been shown to decrease engagement in negative alcohol-related behaviors [76], [77], which may serve to increase the number of individuals that abstain from driving after alcohol consumption. Further, findings provide some preliminary support for extending or combining ecological momentary intervention (EMI) techniques that have been demonstrated as efficacious for reducing both acute trauma-related stress [78] and problematic alcohol behaviors [79]. However, given the novelty of the current study, caution should be taken in advancing education and intervention efforts until findings have been replicated and thoroughly extended in future research.

4.2. Limitations and Future Directions

Findings from the current study should be considered in the context of limitations. First, results may be limited by inherent biases in self-report assessments (e.g., social desirability), as well as the lack of additional context in the questions (e.g., anticipated driving distance). Although questions were purposefully chosen due to their simplicity and potential ecological validity, future work may consider other direct and/or indirect assessments of driving-related cognitions that may be less prone to social desirability responding, as well as consider other factors that may modulate willingness to drive [80]. Second, findings may have been limited by threats to internal validity incurred through the experimental manipulations. For example, generally, although the alcohol administration procedure was successful in reaching the target BrAC, there were incidents in which individuals either did not meet or exceeded the criterion and/or target threshold. Although the target BrAC was specifically selected to reliably impact driving-performance within a reasonable range [10], future work may consider alternative forms of alcohol administration that are more precise [81]. Further, given the sensitive nature of the trauma manipulation (i.e., guided reimagining) and the potential for threats to internal validity (e.g., experimenter, expectancy effects, reactivity), researchers did not obtain objective measures (i.e., BrAC estimates) or subjective measures of intoxication immediately before completion of the driving-related outcomes. Future research should consider passive assessment techniques (e.g., alcohol biosensors) [82] that would allow for continuous assessment of BrAC without requiring more invasive breathalyzer equipment or, at minimum, include a subjective assessment of intoxication (e.g., single-item, face-valid question) [24] prior to or concurrent with completion of the driving-related outcomes measures. Regarding the trauma-related stress induction paradigm, script-driven imagery was an effective means of inducing distress. However, between-subjects experimental designs may be limited by potential within-subjects effects; for example, it may be that, despite randomization, participants in the trauma conditions simply reacted more strongly to the trauma reimagining condition than participants assigned to the neutral reimagining condition. To ameliorate this concern, future work may consider developing a within-subjects design, evaluating driving-related decision-making to both neutral and trauma-related stimuli under the influence of alcohol. Notably, this design would also help in addressing any statistical power concerns, allowing for more nuanced assessments of other proximal and distal driving-related factors, including potentially mediating and moderating factors (e.g., perceived intoxication). Such information may advance our understanding of mechanisms underlying relations between trauma-related stress after alcohol consumption and driving-related cognitions. Indeed, future research aimed at replicating and extending the current findings would benefit not only from within-subject designs, but also by incorporating additional manipulations (e.g., nocebo alcohol condition) that can directly assess potentially mediating and moderating mechanisms (e.g., perceived intoxication) through tightly-controlled experimental paradigms.

4.3. Conclusions

Despite these limitations, the current study meaningfully adds to the literature as the first study to experimentally examine the influence of trauma-related stress after alcohol consumption on driving-related cognitions. Results indicate that ‘real-time’ evocation of trauma-related stimuli may lessen perceived likelihood of negative consequences, while increasing willingness to drive under the influence of alcohol. Generally, these findings highlight the complexity of understanding driving-related cognitions, highlighting trauma-related stress as potential mechanistic trigger for more spontaneous decision-making, and provide meaningful targets for the development and improvement of education and intervention efforts aimed at reducing incidents of driving under the influence of alcohol.

Highlights.

• Script-driven imagery was used to elicit trauma-related stress after alcohol consumption

• Trauma-related stress increased willingness to drive after alcohol consumption

• Trauma-related stress lessened perceived likelihood of getting in an accident

• Trauma-related stress lessened perceived likelihood of getting arrested