Real-Time Assessment of Alcohol Reward, Stimulation, and Negative Affect in Individuals With and Without Alcohol Use Disorder and Depressive Disorders
Publication: American Journal of Psychiatry
Abstract
Objective:
The allostasis theory states that, as addiction develops, alcohol is consumed to relieve negative affect rather than to produce positive effects. This study aimed to investigate the real-time subjective effects of alcohol in individuals with alcohol use disorder (AUD) and those prone to negative affect by virtue of having comorbid depressive disorder (DEP).
Methods:
Participants (N=221) completed high-resolution ecological momentary assessments during 3-hour monitoring of one alcohol drinking episode and one non-alcohol drinking episode in their natural environment. Participants also completed daily mood surveys and next-day surveys. Linear mixed-effect models were used to compare drinking behavior and subjective responses (stimulation, sedation, liking, wanting, negative affect) among 120 participants with AUD (AUD+; with depression [DEP+]: N=64, without depression [DEP−]: N=56) and 101 participants without AUD (AUD−; DEP+: N=45, DEP−: N=56).
Results:
During the monitoring period, participants with AUD consumed an average of 8.5 standard alcohol drinks (estimated blood alcohol concentration [eBAC]=0.115 g/dl) versus 3.7 drinks (eBAC=0.040 g/dl) for non-AUD participants. The AUD group, regardless of comorbid DEP, reported increases in stimulation and rewarding effects that persisted throughout most of the alcohol episode relative to the non-alcohol episode. To a lesser extent, alcohol relieved negative affect but this was not specific to AUD or DEP groups.
Conclusions:
Contrary to the allostasis model of addiction’s emphasis on negative reinforcement drinking, findings demonstrated that people with AUD prone to negative affect displayed positive alcohol reinforcement with pronounced and prolonged sensitivity to alcohol’s pleasurable effects, akin to their noncomorbid counterparts. The findings provided critical testing of addiction theories in the natural environment to enhance external validity.
Alcohol is a causal factor in more than 200 disease and injury conditions and is attributed to the deaths of over 175,000 people annually in the United States (1) and 3 million people (5.3% of all deaths) worldwide (2). The excessive use of alcohol remains highly prevalent, with 10.6% of Americans aged 12 and older (29.5 million) meeting criteria for past-year alcohol use disorder (AUD) (3). Further, there is a high rate of psychiatric comorbidity in AUD, with depressive disorders being the most common comorbidity (33%–50% [4]). AUD and depression have a bidirectional relationship, as having one of these disorders increases the risk of the other (5), and there is some evidence for potential shared vulnerability via genetic predisposition (6) and brain reward and stress system dysfunction (7).
Advances in neuroscience research over the past few decades have fostered models of addiction involving a spiraling process associated with increased negative affect and withdrawal symptoms (8), with alcohol and/or drugs used compulsively to alleviate such aversive states. Related theories suggest alcohol is misused in those prone to negative affect as a form of tension reduction or “self-medication” (9). The allostasis model of addiction (10) hypothesizes that excessive alcohol use produces neuroadaptations from an initial positive reinforcement phase to a negative reinforcement phase. Thus, alcohol is consumed primarily to relieve negative affect and withdrawal (i.e., “the dark side of addiction”) rather than for pleasure or reward (10). Support for these heuristic models is derived from basic science animal models (7, 8) and human neuroimaging showing decreased reward circuit activity in regular drug users versus nonusers to sexual content (11) and in heavy versus light drinkers to intravenous alcohol (12).
Related to experiences of alcohol reward and punishment, alcohol response phenotype predicts the development and maintenance of AUD (13–15). These phenotypes vary, but two large longitudinal studies have shown that enhanced sensitivity to the stimulating and rewarding effects of alcohol (13, 14) and lower sensitivity to its sedating effects (15) predicts heavy drinking behavior and AUD. Debate remains whether excessive drinking reflects overall acute alcohol tolerance, desire for relief from negative mood states (e.g., depression), or heightened sensitivity to alcohol’s pleasurable effects (16). Most of this research has examined otherwise healthy young people who drink and are at risk for future alcohol problems, with those meeting criteria for moderate/severe AUD at baseline excluded (13), so little is known about subjective and motivational alcohol properties in individuals with AUD and comorbid depressive disorders. Examining subjective alcohol effects among these individuals would provide a critical test of theories of alcohol misuse and guide intervention development for these comorbidities.
Thus, to test whether excessive alcohol is associated primarily with relief of negative mood states or reward and enhancement of pleasurable effects, we examined real-time positive and negative subjective effects in individuals with and without AUD and depressive disorders. To this end, we used smartphone-delivered ambulatory monitoring to capture data in participants’ natural environments during an alcohol drinking episode and a comparison non-alcohol drinking episode. The non-alcohol drinking episode allowed us to examine whether participant reports of subjective responses during alcohol use are specific to that substance, versus a nonspecific response bias (17). This approach used high-resolution ecological momentary assessment (HR-EMA) (18), a reliable and valid method for measuring alcohol use and subjective effects during excessive drinking events (17, 19). In this study, we examined a typical drinking event to determine whether persons with AUD and comorbid depressive disorder report a reduction in negative affect, or enhancement of positive affect, as shown previously in nondepressed heavy social (13, 14) and AUD drinkers (17, 20). To our knowledge, this represents the first real-time assessment comparing hedonic alcohol effects (stimulation, liking), negative affect, motivational salience (wanting), and sedative effects in individuals with and without AUD and depressive disorders.
Methods
Design and Overview
Data were collected between October 2020 and October 2021. All procedures were approved by the University of Chicago Institutional Review Board. Among individuals who drink, this study compared those with and without AUD and depressive disorders (between subjects) on natural environment responses during alcohol drinking versus non-alcohol drinking episodes (within subjects). All participants owned a smartphone (iOS or Android) and received survey prompts via a dedicated ecological momentary assessment (EMA) app (Metricwire, Inc.). During the 1-week study period, participants completed 1) daily mood surveys, 2) surveys during one typical alcohol episode and one typical non-alcohol episode separated by at least 24 hours, and 3) next-day surveys after each episode. Data from a portion of this study were published elsewhere (17).
Screening and Eligibility
Study candidates were recruited from online advertisements and word-of-mouth referrals. To facilitate comparison across prior laboratory or EMA studies assessing alcohol responses, the general eligibility criteria were age 21–35 years, ≥10 years of education, English fluency, generally healthy, no current desire to stop drinking, cannabis use ≤3 times/week, recreational drugs ≤2 times/month, smoking ≤5 cigarettes per day, and no history of DSM-5 bipolar disorder, premenstrual dysphoric disorder, eating disorder, obsessive-compulsive disorder, or psychotic disorder. Candidates with current suicidal ideation or ratings ≥5 on the Columbia Suicide Severity Rating Scale (CSSRS) (21) were excluded.
A videoconference screening session assessed candidates’ self-reported demographics (i.e., sex assigned at birth, gender identity, race/ethnicity) and substance use patterns via the Online Timeline Followback (O-TLFB) interview (22), Alcohol Use Disorder Identification Test (AUDIT) (23), Beck Depression Inventory (BDI) (24), State-Trait Anxiety Inventory (STAI) (25), and Structured Clinical Interview for the DSM-5, non-patient version (SCID-5) (26).
Criteria for the AUD+ group were meeting two or more of the 11 criteria for AUD in the past year based on the SCID-5 (26) with current alcohol use of ≥10 standard alcohol drinks/weekly for men (eight for women) and heavy drinking (five or more drinks in an occasion for men, four for women) on a weekly basis for the past 2 years or longer. Participants in the AUD− group did not meet AUD criteria and consumed one to nine standard alcohol drinks/weekly for men (seven for women) with rare or infrequent heavy drinking episodes (five or fewer occasions/year). Moderate drinkers or those with highly variable drinking were excluded. Participants within each AUD group (AUD+: N=120, AUD−: N=101) were further classified into cohorts according to the presence or absence of a depressive disorder (DEP). Among AUD+ participants, there were 56 DEP− and 64 DEP+; among AUD− participants, there were 56 DEP− and 45 DEP+. Those classified as DEP+ met criteria for past-year major depressive disorder (MDD), persistent depressive disorder (PDD), or unspecified depressive disorder (UDD). Participants classified as DEP− did not meet past year criteria for these disorders. Candidates with premenstrual dysphoric disorder or substance- or medication-induced depressive disorder were not included.
Procedure
Participation and study orientation.
Of the 419 candidates screened, 232 (55%) were deemed eligible to participate, and most enrollees (N=221 of 226, 98%) completed the study and comprised the study sample (i.e., four did not follow instructions, one was fraudulent).
The study began with a 30-minute videoconference orientation to explain study procedures, describe surveys, and facilitate use and downloading of the EMA app (Metricwire, Inc.). The participant’s past month O-TLFB interview was used as a guide to inform potential typical alcohol and non-alcohol drinking days for their real-time HR-EMA episodes. Instructions for the alcohol episode were to consume alcohol as typical with no drink amount requirements (for safety and ethical reasons) and, for the non-alcohol episode, to consume non-alcohol beverages of their choice (e.g., water, juice, soda). The participant was also instructed to refrain from alcohol or psychoactive substance use on the day of each recorded episode and to refrain from using psychoactive substances during the episodes.
Daily noon assessments.
Each day at noon local time, the participant received a notification from the app to complete a brief ∼1-minute survey (see “episode subjective surveys” section). If the survey was not completed, reminder prompts were given at 12:15 p.m. and 12:30 p.m. local time. Surveys left uncompleted by 3:00 p.m. local time were counted as missed.
Alcohol and non-alcohol episodes.
The participant used the study app to complete HR-EMA of their behavior and subjective responses during the first 3 hours of one typical alcohol episode and one typical non-alcohol episode. This time frame was chosen to minimize participant burden, capture data during rising-to-peak blood alcohol concentration (BAC) associated with the greatest number of drinking-related consequences (27), and because participants in naturalistic observation studies often go to sleep when BAC is declining (18). Each episode was completed on separate days of the participant’s choosing, at approximately the same time of day, and in any order.
The HR-EMA episodes began with the participant self-initiating a brief baseline survey between 30 minutes to 3 hours before their first beverage to assess mood state and environmental context (28). If alcohol had been already consumed, the episode would not proceed, with a reminder message to avoid alcohol before the episode and to try again at least 24 hours later. Once the participant was ready to initiate their first drink, they were prompted to submit a photo of that drink as a validity check of episode type. Participants were not asked to submit photos of subsequent drinks consumed during the episodes to reduce burden and minimize interference with typical drinking patterns. After finishing the first beverage, the participant self-initiated a post-first-drink survey to assess subjective effects, drink type, quantity, and size. The app initiated all subsequent survey prompts at regular intervals at 15, 30, 60, 90, 120, 150, and 180 minutes after the first drink.
Estimated blood alcohol concentration (eBAC).
As in our prior work (17, 19), eBAC was calculated at each time point for the alcohol drinking episodes based upon participant reports of alcohol consumed (converted to standard drink units) as follows: eBAC = [(c/2) × (GC/w)] − (β60 × t), where c is the total standard drinks consumed to that point since drinking started, GC is a gender constant (7.5 for men and 9.0 for women), w is the participant’s weight in pounds, β60 is a constant representing the mean population alcohol metabolism rate (0.017 g/dl/h), and t is the time in hours since drinking started (29). Of note, as in prior studies (17, 19, 30) if eBAC was ≥0.30 g/dl for any time point, data were censored and treated as missing as that level of intoxication is associated with coma/death and assumed to reflect an error; 1.4% of the data (22/1,629 datapoints) were censored in this fashion.
Episode subjective surveys.
At every time point, the participant completed six items from the Brief Biphasic Alcohol Effects Scale (B-BAES) (31). The participant rated each item (energized, excited, and up [stimulation subscale] and sedated, sluggish, and slow thoughts [sedation subscale]) using a 10-point touchscreen sliding scale. The mean scores were transformed to a 100-point scale to facilitate comparison with the other measures. The same scoring metric was used for negative affect, with an average of three items: sad/down and stressed, based on prior EMA work (32), and another item, anxious/worried, to more fully capture the construct of negative affect. These three items had factor loadings ≥0.82, while a fourth item, lonely, was excluded due to its lower 0.69 factor loading. After the first drink and for all remaining survey prompts, the participant rated three Drug Effects Questionnaire (DEQ) (33) items using a 100-point sliding scale: “I feel the effects of the drink,” “I like the effects I am feeling,” and “I want more of what I consumed.”
Next day survey.
At 11:00 a.m. local time the next morning, the participant completed a next-day survey including the 12-item Alcohol Hangover Severity Scale (AHSS) (34), items assessing whether the prior day’s episode was typical, and whether any more alcohol was consumed after monitoring. Last, the participant completed items assessing their HR-EMA experience and were debriefed and compensated with an electronic payment or gift card worth $140.
Statistical Analyses
Background variables were compared across groups via ANOVA for continuous outcomes and logistic regression for categorical variables. Linear mixed-effects models were used for the primary outcomes to test the allostasis model and included stimulation, liking (hedonic reward), and negative affect. For significance testing, a Bonferroni correction was applied at p<0.017 (0.05/3). Other subjective effects included eBAC, and ratings of feeling the effects, wanting, and sedation, as these variables are not direct components of the allostasis model per se but important to discern in the scope of alcohol responses. The models used random-intercept and slope terms to control for intrasubject correlation. To simplify interpretation of the results and focus on within-group differences on DEP status, these models were conducted separately in the AUD groups and included three factors, including cohort (DEP+ and DEP−), episode type (alcohol and non-alcohol), and time, and their interactions. All models included baseline subjective ratings, when applicable (stimulation, negative affect, and sedation), and eBAC as a time-varying covariate. Post-estimation testing of significant main effects and interaction terms are included in the supplement that accompanies the online version of this article. As there were group differences in education and family history of alcohol use disorder, models were repeated including them as covariates, but they did not significantly change the results and were not included in the final models. For next-day survey data, chi-square and ANOVA were conducted separately in the AUD groups and included DEP cohort, episode type, and their interactions. All analyses were conducted using SAS 9.4 software (Cary, NC).
Results
Participants
The sample had a mean age of 27.0 years (range 21–35), 75% were White, and 54% were male. Demographic characteristics were similar across the subgroups with the exception that AUD− participants had more years of education than those who were AUD+ (Table 1). AUD+ participants endorsed a mean of 4.3 (±2.0) AUD symptoms (range 2–10 symptoms). Forty-five percent of AUD+ participants met criteria for mild AUD. Participants with moderate or severe AUD comprised 30% and 25% of the AUD+ sample, respectively, and were overrepresented relative to United States prevalence rates among people with AUD (20.2% moderate, 20.7% severe) (35). While participants reported a range of drinking behaviors, more than half reported consuming 24 or more drinks weekly (range 24–85 drinks). The AUD+ group was higher than the AUD− group on all alcohol drinking measures, including frequency of both heavy drinking (five or more drinks per occasion for men, four for women) (36) and high-intensity drinking episodes (10 or more drinks in an occasion for men, eight for women) (37) as well as cigarette, cannabis, and other substance use (Table 2). For the DEP+ cohort, most met criteria for past year MDD (84.4%), with fewer meeting criteria for PDD (8.3%) or UDD (7.3%). DEP+ (vs. DEP−) participants also had significantly higher BDI and STAI-Trait scores as well as daily sedation and negative affect scores but not stimulation ratings (Table 2).
| | AUD− (N=101) | AUD+ (N=120) | Significance | ||||||
|---|---|---|---|---|---|---|---|---|---|
| DEP− N=56 | DEP+ N=45 | DEP− N=56 | DEP+ N=64 | ||||||
| N | % | N | % | N | % | N | % | ||
| Sex assigned at birth (male) | 27 | 48.2 | 23 | 51.1 | 34 | 60.7 | 36 | 56.3 | ns |
| Race (White) | 38 | 67.9 | 34 | 75.6 | 46 | 82.1 | 47 | 73.4 | ns |
| Family history of AUD (positive) | 7 | 12.5a | 18 | 40.0b | 24 | 42.9b | 31 | 48.4b | Group × Cohort, p<0.05 |
| | Mean | SD | Mean | SD | Mean | SD | Mean | SD | |
| Age (years) | 27.7 | 3.7 | 26.7 | 3.8 | 26.7 | 3.8 | 26.8 | 4.1 | ns |
| Education (years) | 16.9a | 1.8 | 16.5a | 1.9 | 15.6b | 1.7 | 15.8b | 1.9 | Group, p<0.001 |
| DSM-5 AUD symptoms (past year) | 0.3a | 0.5 | 0.4a | 0.5 | 4.1b | 1.7 | 4.5b | 2.3 | Group, p<0.001 |
| AUDIT (total score) | 5.9a | 2.8 | 6.3a | 2.9 | 14.8b | 5.1 | 17.9c | 5.3 | Group × Cohort, p<0.05 |
| Age at first alcohol drink (years) | 17.5a | 2.3 | 16.6a | 2.2 | 16.0b | 2.1 | 16.0b | 2.4 | Group, p<0.001 |
| Drinks per drinking day (past month) | 2.5a,c | 1.3 | 2.3b,d | 0.9 | 6.6b,c | 2.9 | 5.4b,d | 2.4 | Group, Cohort, ps<0.01 |
| Drinks per week (past month) | 8.0a,c | 6.8 | 6.2b,d | 4.4 | 28.7b,c | 16.4 | 23.9b,d | 12.2 | Group, Cohort, ps<0.05 |
| % drinking days (past month) | 42.4a | 20.4 | 37.0a | 18.6 | 61.8b | 18.1 | 64.7b | 17.6 | Group, p<0.001 |
| % heavy drinking days (past month) | 7.9a | 12.1 | 4.8a | 8.6 | 41.4b | 21.3 | 37.4b | 20.6 | Group, p<0.001 |
| % high intensity days (past month) | 0.8a | 3.0 | 0.4a | 1.3 | 18.2b | 18.6 | 10.0c | 15.2 | Group × Cohort, p<0.05 |
a
Standard drink unit was 14 g ethanol per drink. Heavy drinking defined as consuming ≥4 standard drinks for a woman (≥5 standard drinks for a man). High intensity drinking defined as consuming ≥8 standard drinks for a woman (≥10 standard drinks for a man). Differing subscripts (a-d) denote significant differences for main effects of group (AUD), cohort (DEP), or their interaction.
| | AUD− (N=101) | AUD+ (N=120) | Significance | ||||||
|---|---|---|---|---|---|---|---|---|---|
| DEP− (N=56) | DEP+ (N=45) | DEP− (N=56) | DEP+ (N=64) | ||||||
| N | % | N | % | N | % | N | % | ||
| Any cigarette use (past year) | 7a | 12.5 | 11a | 24.4 | 18b | 32.1 | 29b | 45.3 | Group, p<0.05 |
| Any cannabis use (past year) | 25a | 44.6 | 24a | 53.3 | 39b | 69.6 | 37b | 57.8 | Group, p<0.05 |
| Any other substance use (past year) | 7a | 12.5 | 8a | 17.8 | 18b | 32.1 | 21b | 32.8 | Group, p<0.05 |
| | Mean | SD | Mean | SD | Mean | SD | Mean | SD | |
| Beck Depression Inventoryb | 3.8a | 3.6 | 13.5b | 6.0 | 4.9a | 4.7 | 14.0b | 7.4 | Cohort, p<0.001 |
| State-Trait Anxiety Inventory, traitc | 51.9a | 6.8 | 63.1b | 6.0 | 52.2a | 6.5 | 64.0b | 8.2 | Cohort, p<0.001 |
| Stimulation (daily report)d | 46.1 | 17.3 | 45.4 | 12.5 | 44.5 | 15.1 | 42.0 | 16.1 | n.s. |
| Sedation (daily report)d | 15.8a,c | 13.6 | 26.6a,d | 14.2 | 22.4b,c | 14.0 | 31.9b,d | 14.8 | Group, Cohort, p<0.01 |
| Negative affect (daily report)e | 18.8a | 15.5 | 32.8b | 17.6 | 22.5a | 15.5 | 34.5b | 17.1 | Cohort, p<0.001 |
a
Differing subscripts denote significant differences for main effects of group (AUD+ versus AUD−), cohort (DEP+ versus DEP−), or their interaction.
b
Scores range from 0 to 63.
c
Scores are based on age- and sex-adjusted t-scores, range from 20 to 80.
d
From the Brief Biphasic Alcohol Effects Scale, scores range from 0 to 100.
e
Factor score (mean of sad/depressed, anxious/worried, and stressed ratings), with range of 0 to 100.
Characteristics of real-time drinking episodes.
Most alcohol episodes occurred on weekends (Thursday through Sunday [71%]), and most non-alcohol episodes were on weekdays (Monday through Wednesday [62%]). The mean local start time for alcohol episodes was 6:29 p.m. (±2.7 hr) and non-alcohol episodes at 5:37 p.m. (±3.6 hr). Participants completed 89.2% of system-initiated prompts during the HR-EMA episodes. No participant reported using cannabis, nicotine, or other (non-alcohol) drugs during either HR-EMA episode. Notably, at pre-drink baseline, negative affect was higher in DEP+ versus DEP− participants (p<0.001) but did not differ by episode type.
Alcohol and non-alcohol episodes: AUD− group by DEP status.
As expected, AUD− participants engaged in moderate drinking, with a mean of 3.7 (±2.1) alcohol-containing drinks consumed in their alcohol episode (peak eBAC=0.04 g/dl; Figure 1A) and 2.6 (±1.3) non-alcohol drinks in the non-alcohol episode. Ratings of feeling the effects of the alcohol beverage increased in the first hour and was sustained thereafter (episode × time, p=0.03; Figure 1C). DEP+ participants reported higher feel ratings than did those who were DEP− (cohort × episode, p=0.004; Figure 1C) during the alcohol-drinking episode.
FIGURE 1. Mean estimated blood alcohol concentration (eBAC) and subjective ratings of feeling the effects for participants with and without alcohol use disorder (AUD) during real-world alcohol and non-alcohol drinking episodes, by presence or absence of comorbid depressive disorder (DEP)a
aData are shown as mean (±SEM) and depicted separately for the alcohol (orange) and the non-alcohol (green) drinking episodes. Data depict outcomes from baseline to the conclusion of a 3-hour high-resolution ecological momentary assessment (HR-EMA) of one typical alcohol drinking episode and one typical non-alcohol drinking episode in participants’ natural environments. AUD groups and DEP cohort were determined via DSM-5 symptom counts. Letters (A–D) indicate outcomes for the two DEP cohorts, separately by AUD status, with outcomes for AUD− participants shown in the left panels (A and C) and outcomes for AUD+ participants shown in the right panels (B and D). For each AUD group separately, linear mixed effects models tested the effects of DEP cohort, episode type, time, and their interactions on outcomes; significant effects of those variables are depicted within each lettered subgraph, where applicable. For details on the results of the statistical models and significance testing, see text and Table S1 in the online supplement .
Among AUD− participants, relative to the non-alcohol episode, alcohol increased stimulation during the early portion of the episode followed by a precipitous decline (episode × time, p<0.01; Figure 2A) and alcohol also increased liking (episode, p<0.0001; Figure 2C). For both stimulation and liking, there were no differences between DEP cohorts. Alcohol also decreased negative affect (episode × time, p<0.01; Figure 2E) particularly during the early portion of the episode with no differences between DEP cohorts. Also, for AUD− participants, alcohol increased wanting during the first hour with a precipitous decline over time (episode × time, p<0.0001; Figure 3A). Sedation was higher for the DEP+ cohort regardless of episode type (cohort × time, p<0.01; Figure 3C) and alcohol steadily increased sedation in the last 2 hours of the episode compared with the non-alcohol episode (episode × time, p<0.001). Detailed model outcomes and post-estimation results are in Table S1 in the online supplement .
FIGURE 2. Mean subjective ratings of stimulation, liking, and negative affect for participants with and without alcohol use disorder (AUD) during real-world alcohol and non-alcohol drinking episodes, by presence or absence of comorbid depressive disorder (DEP)a
aData are shown as mean (±SEM) and depicted separately for the alcohol (orange) and the non-alcohol (green) drinking episodes. Data depict outcomes from baseline to the conclusion of a 3-hour high-resolution ecological momentary assessment (HR-EMA) of one typical alcohol drinking episode and one typical non-alcohol drinking episode in participants’ natural environments. AUD groups and DEP cohorts were determined via DSM-5 criteria. Letters (A–F) indicate outcomes for the two DEP cohorts, separately by AUD group, with outcomes for AUD− participants shown in the left panels (A, C, and E) and outcomes for AUD+ participants shown in the right panels (B, D, and F). For each AUD group separately, linear mixed effects models tested the effects of DEP cohort, episode type, time, and their interactions on outcomes; significant effects of those variables are depicted within each lettered subgraph, where applicable. For details on the results of the statistical models and significance testing, see text and Table S1 in the online supplement .
FIGURE 3. Mean subjective ratings of wanting more and sedation for participants with and without alcohol use disorder (AUD) during real-world alcohol and non-alcohol drinking episodes, by presence or absence of comorbid depressive disorder (DEP)a
aData are shown as mean (±SEM) and depicted separately for the alcohol (orange) and the non-alcohol (green) drinking episodes. Data depict outcomes from baseline to the conclusion of a 3-hour high-resolution ecological momentary assessment (HR-EMA) of one typical alcohol drinking episode and one typical non-alcohol drinking episode in participants’ natural environments. AUD groups and DEP cohort were determined via DSM-5 criteria. Letters (A–D) indicate outcomes for the two DEP cohorts, separately by AUD group, with outcomes for AUD− participants shown in the left panels (A and C) and outcomes for AUD+ participants shown in the right panels (B and D). For each AUD group separately, linear mixed effects models tested the effects of DEP cohort, episode type, time, and their interactions on outcomes; significant effects of those variables are depicted within each lettered subgraph, where applicable. For details on the results of the statistical models and significance testing, see text and Table S1 in the online supplement .
Alcohol and non-alcohol episodes: AUD+ group by DEP status.
AUD+ participants engaged in heavier alcohol consumption, with higher magnitude subjective effects, compared to AUD− participants. AUD+ participants consumed 8.5 (±5.1) standard alcohol drinks and 3.1 (±1.5) non-alcohol drinks in each episode, respectively. The eBAC slopes increased steadily for AUD+ participants such that by 60–90 minutes, mean eBAC exceeded the 0.08 g/dl U.S. legal limit and averaged 0.115 g/dl (±0.08) at the end of monitoring (episode × time, p<0.001; Figure 1B). Alcohol drink totals and eBAC levels were similar between DEP cohorts (cohort × time, p=0.45) and feeling the effects of the beverage ratings increased accordingly (episode × time, p<0.0001; Figure 1D).
Among AUD+ participants, relative to the non-alcohol episode, alcohol increased stimulation, with pronounced elevated ratings throughout the episode (episode, p<0.0001; Figure 2B), and these ratings did not differ by DEP cohort. For liking, there was a three-way interaction (cohort × episode × time, p=0.01; Figure 2D) such that alcohol increased liking more in the DEP+ cohort early in the alcohol episode, but liking was also increased in those who were DEP+ during the middle of the non-alcohol episode. Alcohol produced an overall decrease in negative affect (episode, p<0.0001; Figure 2F), and there were no differences between DEP cohorts. Finally, in these AUD+ participants, alcohol produced heightened and sustained ratings of wanting more, with an eventual decline in the last hour (episode × time, p=0.015; Figure 3A) with no differences between DEP cohorts. Alcohol also increased sedation over time (episode × time, p<0.01) in both DEP cohorts. Detailed model outcomes and post-estimation results are in Table S1 in the online supplement .
As AUD is a heterogeneous disorder and our sample ranged on drinking behaviors and alcohol-related consequences, exploratory linear mixed-effects models compared alcohol and non-alcohol episode responses, controlling for eBAC, in mild (N=54) versus moderate-to-severe AUD (N=66) subgroups, regardless of DEP status. As expected, moderate and severe AUD reported more excess drinking behavior than did those with mild AUD (see Table S2 in the online supplement ). Relative to persons with mild AUD, those with moderate and severe AUD exhibited a sustained elevation in stimulation throughout the episode (group × episode × time, p<0.001), higher wanting (group × episode, p<0.05), and similar increases in liking and decreases in negative affect (see the online supplement ).
Next-Day Surveys
More than half of the AUD+ participants (53%) reported continued alcohol drinking after the alcohol episode compared with 20% of those who were AUD− (χ2=26.2, df=1,p<0.001). As expected, next-day hangover scores were higher after alcohol versus non-alcohol episodes (p<0.0001) and for the AUD+ group after the alcohol episode (AHSS mean score 2.5 [±1.5] versus 1.4 [±1.0] in the AUD− group [p<0.0001]). As in our prior work (17, 19), participants found the HR-EMA method acceptable and feasible. Next-day surveys revealed that 95% of participants endorsed that the alcohol episodes were typical.
Discussion
During alcohol intoxication, people with comorbid AUD and depressive disorder showed a positive alcohol response phenotype characterized by heightened stimulation, liking, and wanting, resembling the response profile of AUD drinkers without depression. Thus, persons with AUD, even those with depressive disorders, experience the pleasurable effects of alcohol soon after consuming alcohol, and these effects remain elevated for much of the course of a heavy drinking bout. In contrast, persons without AUD, regardless of depressive disorder status, engaged in moderate drinking with less pronounced positive-type subjective responses. Finally, alcohol reduced negative affect, but this reduction was small in magnitude and nonspecific to depressive disorder or AUD status. Subjective responses during the alcohol drinking episode differed significantly from those in the non-alcohol drinking episode, validating that participants’ subjective experiences reflected the effects of alcohol rather than a nonspecific response bias (17). These findings extend prior results of placebo-controlled laboratory studies demonstrating pronounced positive-type alcohol effects in non-comorbid heavy social alcohol users and individuals with AUD (14, 20).
It is notable that the current findings from natural environment alcohol and non-alcohol drinking episodes do not support the allostasis model’s contention that an initial positive reinforcement phase wanes in addiction, particularly for individuals with conditions characterized by negative affect who are theorized to drink to excess for negative reinforcement factors. As such, this real-time assessment of drinking in comorbid drinkers disputes conventional thinking that the motivation to consume large quantities of alcohol serve as self-medication (9) to lessen negative affective states (10) or within a relief drinking profile (38). Rather, these data indicate that pronounced and prolonged stimulating, hedonic, and motivational rewarding effects of alcohol may underlie alcohol misuse in persons with AUD and comorbid depressive disorder. Finally, the present study findings challenge the widely held concept of global tolerance in addiction (39). While behavioral tolerance may be present at lower (but not higher) intoxicating alcohol doses in persons with AUD (40), controlled studies of subjective responses in persons with AUD are lacking. In the present study’s real-time assessment of naturalistic drinking episodes, we found evidence of sensitivity to alcohol’s desirable subjective effects, rather than tolerance to these effects, with persistently high stimulation, hedonic reward, and motivational reward in the alcohol (versus non-alcohol) episodes in persons with AUD, regardless of depression status. While exploratory AUD subgroup analyses yielded small sample sizes, these results indicated that the most pronounced AUD drinkers, i.e., individuals with moderate and severe AUD, exhibited higher alcohol stimulation and hedonic reward compared with those with mild AUD.
Alcohol affects the brain in complex ways (41), and a greater understanding of the factors that affect vulnerability to AUD and comorbid depression is critical for early identification and more effective prevention and treatment. Despite advances in the measurement of acute alcohol responses in the laboratory and natural environment (18, 42), only a few studies have examined acute alcohol effects in individuals who drink with co-occurring mental health disorders, with most (43–45) but not all (46), showing higher positive-like alcohol responses in people with psychiatric conditions (bipolar type 1, schizophrenia, anxiety-like disorders). However, none of these studies examined individuals with depressive disorders. Thus, our understanding of alcohol response phenotypes in problem drinkers with affective disturbance is hampered by the paucity of research, lack of inclusion of people with AUD, and small sample sizes (subgroup Ns of 20 to 30). Related, the acute alcohol response findings may have implications for treatment such that for dual diagnosis patients, medications targeting opioidergic, dopaminergic, or impulsivity-related neural mechanisms (47) to combat hedonic and motivational salience effects of alcohol in combination with antidepressant medications may show more efficacy than medications focused on relief of negative affect or the stress response system (48). Treatment outcomes could be further optimized with behavioral approaches that address a positive alcohol response phenotype (49).
While controlled laboratory studies are the gold standard for measuring acute alcohol response phenotypes in persons at risk for AUD (13, 15) and in the context of the progression and maintenance of AUD (14, 20), such paradigms have limited external validity. The novel HR-EMA approach used in this study leveraged smartphone technology with brief assessments to bridge prior studies using fixed-dose oral or intravenous alcohol administration (42) to real-time assessment in the natural environment of persons who drink alcohol. The most notable difference between the laboratory-based controlled dosing paradigm and HR-EMA is that the former allows assessment of biphasic alcohol effects when breath alcohol levels are rising, at its peak, and declining whereas the latter reveals what occurs in the natural environment: steady paced drinking in light or moderate drinkers and progressive excessive drinking to intoxication in risky drinkers (17). There is also debate about people’s ability to accurately report their internal sensations. Our HR-EMA assessments with psychometrically valid adjective items, a comparison non-alcohol episode, and instructions not requiring participants to ascertain effects based on the stimulus (i.e., beverage consumed) may serve to lessen the potential confounds of complex cognitive appraisals (50).
The present study featured several strengths, such as enrolling participants with past-year depressive disorders, conducting real-time natural environment assessments to enhance external validity, controlling for eBAC levels in analyses, and demonstrating excellent survey completion rates during the real-world episodes. Also, the participants in the AUD+ group not only represent a clinically relevant sample but also a more excessive drinking group than in most prior studies of heavy social drinkers at risk for AUD. The heavy drinking frequency was 41% and 37% of days for AUD+ participants without and with depressive disorders, respectively, compared with 28% and 22% of days in prior laboratory (13) and HR-EMA samples (28). However, there were some limitations. First, to reduce burden, the episodes were limited to 3 hours and so may have missed longer drinking episodes, which may have been particularly important among participants with AUD. Second, while DEP+ participants met past-year criteria for a depressive disorder, those with severe symptoms (e.g., current suicidal ideation) were excluded for safety reasons. Third, alcohol and non-alcohol drinking events were based upon single episodes. However, prior studies have shown 97% agreement between self-report of drinking and alcohol biosensor data (40), as well as good correspondence of subjective effects across multiple episodes and laboratory sessions (19). Fourth, while participants were young adults between 21 and 35 years of age, corresponding to the highest rates of heavy drinking across the lifespan (51), generalizability to older adults may be limited. Of note, the allostasis model does not specify age ranges or duration intervals for the purported stage transitions. Last, as participants completed the mobile assessments outside of the laboratory and in their usual environments, there is the possibility that they misreported their alcohol or other substance use. Alcohol biosensors have the potential to augment future EMA work in this area by providing an objective measure of alcohol use in naturalistic settings (40, 52, 53).
Positive reinforcement in which alcohol produces hedonic effects may not be solely confined to the early stages of addiction but rather may persist throughout the addiction cycle (7) and in persons with high negative affect by virtue of depressive disorder. In this first study to examine comorbid drinkers’ real-time assessment of alcohol versus non-alcohol drinking, drinking to intoxication in individuals with AUD was more closely associated with enhancement of positive affect than with relief of negative affect and observed in both those with and without concomitant depressive disorder. This real-world drinking study adds external validity to prior controlled laboratory studies suggesting the continuation of the reward-sensitive stage of drinking in the addiction process, even in persons with high negative affect by virtue of depressive disorder and suggests a co-existence, rather than progression from positive to negative reinforcement, in young adults most prone to negative affect-driven drinking by virtue of AUD and DEP comorbidity.
Supplementary Material
File (appi.ajp.20240069.ds001.pdf)
- View/Download
- 311.54 KB
References
1.
Alcohol and Public Health. Atlanta, GA, CDC, 2024. https://www.cdc.gov/alcohol/index.htm
2.
Alcohol. Geneva, World Health Organization, 2022. https://www.who.int/news-room/fact-sheets/detail/alcohol
3.
National Survey on Drug Use and Health: Detailed Tables. Rockville, MD, Substance Abuse and Mental Health Service Administration, 2021
4.
McHugh RK, Weiss RD: Alcohol use disorder and depressive disorders. Alcohol Res 2019; 40:arcr.v40.1.01
5.
Castillo-Carniglia A, Keyes KM, Hasin DS, et al: Psychiatric comorbidities in alcohol use disorder. Lancet Psychiatry 2019; 6:1068–1080
6.
Andersen AM, Pietrzak RH, Kranzler HR, et al: Polygenic scores for major depressive disorder and risk of alcohol dependence. JAMA Psychiatry 2017; 74:1153–1160
7.
Vendruscolo LF, Koob GF, Zorrilla EP, et al: Alcohol use disorder: stress, negative reinforcement, and negative urgency; in Alcohol and Alcohol-Related Diseases. Edited by Mueller S, Heilig M. Cham, Springer International Publishing, 2023, pp 687–719. https://doi.org/10.1007/978-3-031-32483-3_36
8.
Koob GF, Volkow ND: Neurobiology of addiction: a neurocircuitry analysis. Lancet Psychiatry 2016; 3:760–773
9.
Khantzian EJ: The self-medication hypothesis of substance use disorders: a reconsideration and recent applications. Harv Rev Psychiatry 1997; 4:231–244
10.
Koob GF, Le Moal M: Plasticity of reward neurocircuitry and the “dark side” of drug addiction. Nat Neurosci 2005; 8:1442–1444
11.
Garavan H, Pankiewic J, Bloom A, et al: Cue-induced cocaine craving: neuroanatomical specificity for drug users and drug stimuli. Am J Psychiatry 2000; 157:1789–1798
12.
Gilman JM, Ramchandani VA, Crouss T, et al: Subjective and neural responses to intravenous alcohol in young adults with light and heavy drinking patterns. Neuropsychopharmacology 2012; 37:467–477
13.
King AC, de Wit H, McNamara PJ, et al: Rewarding, stimulant, and sedative alcohol responses and relationship to future binge drinking. Arch Gen Psychiatry 2011; 68:389–399
14.
King A, Vena A, Hasin DS, et al: Subjective responses to alcohol in the development and maintenance of alcohol use disorder. Am J Psychiatry 2021; 178:560–571
15.
Schuckit MA: Low level of response to alcohol as a predictor of future alcoholism. Am J Psychiatry 1994; 151:184–189
16.
King A: Alcohol responses as phenotypic markers of AUD risk: lessons from longitudinal studies; in Alcohol and Alcohol-Related Diseases. Berlin, Springer Nature, 2023. pp 623–642
17.
Fridberg DJ, Lee Z, Fischer AM, et al: Assessing real-time positive subjective effects of alcohol using high-resolution ecological momentary assessment in risky versus light drinkers. Alcohol Clin Exp Res (Hoboken) 2023; 47:1570–1581
18.
Piasecki TM: Assessment of alcohol use in the natural environment. Alcohol Clin Exp Res 2019; 43:564–577
19.
Fridberg DJ, Cao D, King AC, et al: Alcohol subjective responses in heavy drinkers: measuring acute effects in the natural environment versus the controlled laboratory setting. Alcohol Clin Exp Res 2021; 45:1287–1297
20.
King AC, Vena A, Howe MM, et al: Haven’t lost the positive feeling: a dose-response, oral alcohol challenge study in drinkers with alcohol use disorder. Neuropsychopharmacology 2022; 47:1892–1900
21.
Posner K, Brown GK, Stanley B, et al: The Columbia–suicide severity rating scale: initial validity and internal consistency findings from three multisite studies with adolescents and adults. Am J Psychiatry 2011; 168:1266–1277
22.
Rueger SY, Trela CJ, Palmeri M, et al: Self-administered web-based timeline followback procedure for drinking and smoking behaviors in young adults. J Stud Alcohol Drugs 2012; 73:829–833
23.
Babor TF, Higgins-Biddle JC, Saunders JB, et al: AUDIT: The Alcohol Use Disorders Identification Test: Guidelines for Use in Primary Health Care, 2nd ed. Geneva, World Health Organization, 2001
24.
Beck AT, Ward CH, Mendelson M, et al: An inventory for measuring depression. Arch Gen Psychiatry 1961; 4:561–571
25.
Spielberger CD, Gorusch RL, Lushene R, et al: Manual for the State-Trait Anxiety Inventory. Palo Alto, CA, Consulting Psychologists Press, 1983
26.
First M, Williams J, Karg R, et al: User’s Guide for the Structured Clinical Interview for DSM-5 Disorders, Research Version (SCID-5-RV). Washington, DC, American Psychiatric Association, 2015
27.
Turner JC, Bauerle J, Shu J, et al: Estimated blood alcohol concentration correlation with self-reported negative consequences among college students using alcohol. J Stud Alcohol 2004; 65:741–749
28.
Fischer AM, King AC, Cao D, et al: Drinking context, alcohol use, and subjective responses during binge drinking episodes measured by high-resolution ecological momentary assessment (HR-EMA). Psychol Addict Behav 2023; 37:258–266
29.
Matthews DB, Miller WR: Estimating blood alcohol concentration: two computer programs and their applications in therapy and research. Addict Behav 1979; 4:55–60
30.
Piasecki TM, Jahng S, Wood PK, et al: The subjective effects of alcohol–tobacco co-use: an ecological momentary assessment investigation. J Abnorm Psychol 2011; 120:557–571
31.
Rueger SY, McNamara PJ, King AC, et al: Expanding the utility of the biphasic alcohol effects scale (BAES) and initial psychometric support for the brief-BAES (B-BAES). Alcohol Clin Exp Res 2009; 33:916–924
32.
Hedeker D, Mermelstein RJ, Demirtas H, et al: An application of a mixed-effects location scale model for analysis of ecological momentary assessment (EMA) data. Biometrics 2008; 64:627–634
33.
Morean ME, de Wit H, King AC, et al: The drug effects questionnaire: psychometric support across three drug types. Psychopharmacology (Berl) 2013; 227:177–192
34.
Penning R, McKinney A, Bus LD, et al: Measurement of alcohol hangover severity: development of the Alcohol Hangover Severity Scale (AHSS). Psychopharmacology (Berl) 2013; 225:803–810
35.
2022 National Survey on Drug Use and Health: Detailed Tables. Rockville, MD, Substance Abuse and Mental Health Services Administration (US), 2023. https://www.samhsa.gov/data/report/2022-nsduh-detailed-tables
36.
National Institute on Alcohol Abuse and Alcoholism: Drinking Levels Defined. https://www.niaaa.nih.gov/alcohol-health/overview-alcohol-consumption/moderate-binge-drinking
37.
Patrick ME: A call for research on high-intensity alcohol use. Alcohol Clin Exp Res 2016; 40:256–259
38.
Grodin EN, Kady A, Ray LA, et al: The future of reward and relief drinking profiles: considerations for social motives, stability of profiles, and tailored interventions. Alcohol Clin Exp Res (Hoboken) 2024; 48:462–465
39.
Tabakoff B, Cornell N, Hoffman PL: Alcohol tolerance. Ann Emerg Med 1986; 15:1005–1012
40.
Didier N, Vena A, Feather AR, et al: Holding your liquor: comparison of alcohol‐induced psychomotor impairment in drinkers with and without alcohol use disorder. Alcohol Clin Exp Res (Hoboken) 2023; 47:1156–1166
41.
Heilig M, MacKillop J, Martinez D, et al: Addiction as a brain disease revised: why it still matters, and the need for consilience. Neuropsychopharmacology 2021; 46:1715–1723
42.
Cyders MA, Plawecki MH, Corbin W, et al: To infuse or ingest in human laboratory alcohol research. Alcohol Clin Exp Res 2020; 44:764–776
43.
Chutuape MA, de Wit H: Preferences for ethanol and diazepam in anxious individuals: an evaluation of the self-medication hypothesis. Psychopharmacology (Berl) 1995; 121:91–103
44.
D’Souza DC, Gil RB, Madonick S, et al: Enhanced sensitivity to the euphoric effects of alcohol in schizophrenia. Neuropsychopharmacology 2006; 31:2767–2775
45.
Lippard ETC, Kirsch DE, Kosted R, et al: Subjective response to alcohol in young adults with bipolar disorder and recent alcohol use: a within-subject randomized placebo-controlled alcohol administration study. Psychopharmacology (Berl) 2023; 240:739–753
46.
Yip SW, Doherty J, Wakeley J, et al: Reduced subjective response to acute ethanol administration among young men with a broad bipolar phenotype. Neuropsychopharmacology 2012; 37:1808–1815
47.
Mellick WH, Tolliver BK, Brenner HM, et al: Alcohol cue processing in co-occurring bipolar disorder and alcohol use disorder. JAMA Psychiatry 2023; 80:1150–1159
48.
Kwako LE, Spagnolo PA, Schwandt ML, et al: The corticotropin releasing hormone-1 (CRH1) receptor antagonist pexacerfont in alcohol dependence: a randomized controlled experimental medicine study. Neuropsychopharmacology 2015; 40:1053–1063
49.
Fridberg DJ, Cao D, King AC, et al: Integrating alcohol response feedback in a brief intervention for young adult heavy drinkers who smoke: a pilot study. Drug Alcohol Depend 2015; 155:293–297
50.
Nisbett RE, Wilson TD: Telling more than we know: verbal reports on mental processes. Psychol Rev 1977; 84:231–259
51.
Bohm MK, Liu Y, Esser MB, et al: Binge drinking among adults, by select characteristics and state – United States, 2018. MMWR Morb Mortal Wkly Rep 2021; 70:1441–1446
52.
Richards VL, Wang Y, Porges EC, et al: Using alcohol biosensors and biomarkers to measure changes in drinking: associations between transdermal alcohol concentration, phosphatidylethanol, and self-report in a contingency management study of persons with and without HIV. Exp Clin Psychopharmacol 2023; 31:991–997
53.
Wang Y, Fridberg DJ, Leeman RF, et al: Wrist-worn alcohol biosensors: strengths, limitations, and future directions. Alcohol 2019; 81:83–92
Information & Authors
Information
Published In
History
Received: 24 January 2024
Revision received: 3 April 2024
Revision received: 23 May 2024
Accepted: 9 July 2024
Published online: 1 February 2025
Published in print: February 01, 2025
Keywords
Authors
Competing Interests
Dr. Fridberg reports serving as a consultant on the development for Click Therapeutics, Inc. The remaining authors report no financial relationships with commercial interests.
Funding Information
This research was supported by grant R01-AA013746 (AK) and R21-AA029746 (AK, DF) from the National Institute on Alcohol Abuse and Alcoholism.
Metrics & Citations
Metrics
Citations
Export Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
View Options
View options
PDF/EPUB
View PDF/EPUBLogin options
Already a subscriber? Access your subscription through your login credentials or your institution for full access to this article.
Personal login Institutional Login Open Athens loginNot a subscriber?
PsychiatryOnline subscription options offer access to the DSM-5-TR® library, books, journals, CME, and patient resources. This all-in-one virtual library provides psychiatrists and mental health professionals with key resources for diagnosis, treatment, research, and professional development.
Need more help? PsychiatryOnline Customer Service may be reached by emailing PsychiatryOnline@psych.org or by calling 800-368-5777 (in the U.S.) or 703-907-7322 (outside the U.S.).