Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Aug 11.
Published in final edited form as: Alcohol Clin Exp Res (Hoboken). 2023 Apr 28;47(4):713–723. doi: 10.1111/acer.15036

Evidence for an alcohol-related “harm paradox” in individuals with internalizing disorders: Test and replication in two independent community samples

Justin J Anker 1, Paul Thuras 2, Ruichong Shuai 3, Lee Hogarth 3, Matt G Kushner 1
PMCID: PMC10416809  NIHMSID: NIHMS1907127  PMID: 37115410

Abstract

Background:

Internalizing (anxiety and mood) disorders (INTD) commonly co-occur (are “comorbid”) with alcohol use disorder (AUD). The literature suggests that excessive alcohol use aimed at coping with INTD symptoms is, at best, a partial explanation for the high comorbidity rates observed. We hypothesized that individuals with INTD experience greater susceptibility to developing AUD symptoms due to the partially shared neurobiological dysfunctions underlying both conditions. We probe this hypothesis by testing the prediction that, after accounting for the volume of alcohol intake, individuals with INTD experience higher levels of alcohol-related symptoms.

Methods:

Data from the National Epidemiological Survey on Alcohol-Related Conditions (NESARC) Wave 3 were used for the primary analyses, and NESARC Wave 1 data were used for independent replication analyses. Individuals who reported any alcohol use in the prior year were categorized as: (1) never having had an INTD diagnosis (“INTD-Never”); (2) having a remitted INTD diagnosis only (“INTD-Remitted”); or (3) having current INTD diagnosis (“INTD-Current”). Between-group contrasts of alcohol-related symptoms controlled for total alcohol intake (past year), drinking pattern (e.g., binging) and variables previously shown to mark exaggerated AUD symptoms relative to drinking amount (e.g., SES, gender, and family history).

Results:

With all covariates in the model, individuals in the INTD-Current group and the INTD-Remitted group reported significantly greater alcohol-related symptoms than those in the INTD-Never group but did not themselves differ in level of alcohol-related symptoms. These results were replicated in the NESARC 1 dataset.

Conclusions:

Individuals with INTD experience more alcohol-related symptoms than those who drink at the same level. While considering other explanations, we argue that this “harm paradox” is best explained by the view that INTD confers a neurobiologically mediated susceptibility to the development of AUD symptoms.

Keywords: alcohol use disorder, anxiety disorder, comorbidity, depression, NESARC

BACKGROUND

The prevalence of alcohol use disorder (AUD) among those with an anxiety disorder or major depression disorder (collectively referred to as internalizing disorder; INTD) ranges from 20% to 40% (Hunt et al., 2020; Lai et al., 2015) compared to around 5% in the general population (SAMHSA, 2020). While a great deal is known about this association—commonly referred to as “comorbidity”—the field’s understanding of its root cause(s) remains tentative and incomplete (Kushner, 2014; Kushner et al., 1990, 2000). A common explanation for the high rate of comorbidity is that the elevated risk of AUD among those with INTD emerges from excessive alcohol use motivated by the reduction of negative affect (“self-medication”). However, findings reported in the literature do not uniformly support this explanation. It does appear that a high percentage of individuals with INTD-AUD comorbidity report that obtaining relief from negative affect serves as a primary motive for their drinking (Anker et al., 2016; Menary et al., 2011). However, results from studies comparing drinking levels between those with and without INTD among those who have not (yet) developed AUD are mixed. For example, previous work has shown that alcohol use among those with an INTD is higher than (Boschloo et al., 2011; Fröjd et al., 2011; Grant, 1996; Mueller et al., 1994), lower than (Davidson & Blackburn, 1998; Kranzler et al., 1996; Pardini et al., 2007), and the same as (Abram et al., 2015; Marmorstein, 2015; Wolitzky-Taylor et al., 2012) those without an INTD. Furthermore, it is known that INTD can onset after AUD in a sizable minority of comorbid cases (Fergusson et al., 2009; Kushner et al., 1990; McHugh & Weiss, 2019; Schuckit & Hesselbrock, 1994), an observation that is inconsistent with “self-medication” as the exclusive cause of comorbidity. These and related findings call into question the view that higher levels of alcohol consumption fully account for the large increased risk for AUD conferred by INTD.

Because the neuro-dysregulations underlying AUD and INTD overlap in some important ways, it is reasonable to consider how the common cooccurrence of these conditions may be an outgrowth of shared neurobiology (Anker & Kushner, 2019; Koob & Le Moal, 2008; Markou et al., 1998). Anker and Kushner (2019) describe how the overlapping neurobiological architecture of INTD and AUD, particularly concerning systems that regulate stress and reward such as the hypothalamic pituitary adrenal system (HPA) and the limbic system (Gilpin et al., 2015; Vinkers et al., 2021), may be relevant to the high rate of comorbidity between these conditions. For example, dysfunction in the central amygdala, which serves as an integrative hub for stress and reward systems and consequently plays a major role in emotional processing (Agoglia & Herman, 2018; Breese et al., 2011; Gilpin et al., 2015; Noronha et al., 2014; Roberto & Gilpin, 2014), is a neurobiological correlate of both INTD and AUD (Agoglia & Herman, 2018; Breese et al., 2011; Gilpin et al., 2015). Also, the central amygdala has been shown to regulate depressive- and anxiety-like states in both humans and rodents (Han et al., 2018; Koob, 2013; Koob & Le Moal, 2001; VanElzakker et al., 2014) and serves a functional role in the acquisition and maintenance of alcohol self-administration in nondependent rodents (Dyr & Kostowski, 1995; Hyytiä & Koob, 1995; Möller et al., 1997). Furthermore, amygdala volume has been shown to be reduced in nondependent individuals with a family history of AUD (Hill et al., 2001).

Anker and Kushner (2019) marshal data supporting the hypothesis that INTD-related dysregulations in mood and stress brain systems that are also linked to chronic alcohol use (e.g., Koob & Le Moal, 2008) result in a neurobiologically mediated susceptibility to the development of AUD among those with an INTD. Consistent with this, we found that individuals with an INTD reported a significantly shorter (“telescoped”) period from the age at which they began regular drinking (Kushner et al., 2011) or nicotine use (Kushner et al., 2012a) to the age at which they first met dependence criteria for these substances. Notably, Kushner et al. (2012a) showed that the INTD telescoping effect in nicotine dependence was not due to a higher level of nicotine use/exposure. These findings describe a “harm paradox” for those with INTD; that is, harms from a given level of substance use within a defined group that are in excess of those experienced by individuals outside that group (Boyd et al., 2021; Erskine et al., 2010; Harrison & Gardiner, 1999; Jones et al., 2015; Mäkelä, 1999; Randall et al., 1999). However, Kushner et al. (2011) could not rule out heavier drinking as an explanation for the condensed time frame in which those with an INTD developed alcohol dependence. Moreover, both studies only included individuals who were already substance dependent, which might not generalize to others with INTD.

In this study, we build on the ideas and work reviewed above by employing two large community-based datasets that include individuals whose drinking levels span a wide range and whose INTD diagnostic history is well-documented. We test the prediction that AUD symptoms/problems are greater in those with versus without an INTD, even after controlling for level of alcohol consumption and other known (or suspected) sources of magnified consequences from alcohol use (e.g., gender, frequency of binge drinking, education, income, and family history of AUD). To increase interpretability and scientific rigor, we tested the hypothesis in current drinkers with and without an AUD who either had a “current” INTD (met diagnostic criteria in the past 12 months), a “remitted” INTD (did not meet diagnostic criteria in the last 12 months but did prior to that), or who never had an INTD. Finally, we tested the hypothesis separately in two large independent but comparable epidemiological samples to provide a within-study replication of the results.

METHODS

Sources of data

Data for the primary analyses were drawn from Wave 3 (collected in 2012 through 2013) of the National Epidemiological Survey on Alcohol-Related Conditions (NESARC), consisting of 36,309 individuals (Grant et al., 2014). The NESARC was designed and sponsored by the National Institute on Alcohol Abuse and Alcoholism (NIAAA) to assess the prevalence of psychiatric disorders among the population of noninstitutionalized civilian adults (≥18 years of age) in the United States, including Alaska, Hawaii, and the District of Columbia. Institutional Review Board approval was provided by the US Census Bureau and US Office of Management and Budget for Wave 3 of the NESARC. The Institutional Review Board of the University of Minnesota reviewed this study and waived the requirement for additional informed consent by the participants.

Sampling in the NESARC involved a multistage probability algorithm to randomly select individuals to participate in the survey. Participants self-identified race/ethnicity, which were coded as white, black, Native American or Alaskan, Asian, Native Hawaiian or Pacific Islander, or Hispanic or Latino. The Hispanic origin variable was constructed from an algorithm developed by the US Census Bureau. African American and Hispanic households and adults who were aged 18 to 24 were oversampled, and data were adjusted to accommodate oversampling and nonresponse (household and person level). (Additional details specific to the replication analysis dataset—NESARC Wave 1—are included below.)

Participants

Primary analyses were conducted using the NESARC Wave 3 dataset as restricted to the 25,778 individuals who reported consuming alcohol in the “past year” (i.e., the year preceding their NESARC interview). Within this study sample, 54% were female, and ages ranged from 18 to 98 (M = 43.4, SD = 16.50). For this study, respondents were classified into one of the following three study groups: (1) never diagnosed with an INTD (“INTD-Never,” n = 18,297); (2) remitted INTD (met diagnostic criteria >12 months ago only; “INTD-Remitted,” n = 2818); and (3) current INTD (met diagnostic criteria within the last 12 months; “INTD-Current,” n = 4663). Detailed demographic information by study group is presented in Table 1.

TABLE 1.

Demographic characteristics of participants with current or remitted INTD and those who never had an INTD (NESARC 3, N = 25,778).

INTD-never (N = 18,297) INTD-remitted (N = 2818) INTD-current (N = 4663) Statistic, p-value
Demographic/Socioeconomic category
 Age, mean (SD) 43.7 (16.91)a 45.5 (14.96)b 41.1 (15.21)c F (2, 25,775) = 72.4, p < 0.001
 Female, no. (%) 8856 (48.4%)a 1837 (65.2%)b 3086 (66.2%)b χ2(2) = 614.0, p < 0.001
 Ethnicity (%) χ2(8) = 432.6, p < 0.001
  White 9527 (52.1%)a 1964 (69.7%)b 2807 (60.2%)c
  Black 3953 (21.6%)a 355 (12.6%)b 823 (17.6%)c
  Native American 213 (1.2%)a 47 (1.7%)b 105 (2.3%)b
  Asian 883 (4.8%)a 82 (2.9%)b 131 (2.8%)b
  Hispanic/Latino 3721 (20.3%)a 370 (13.1%)b 797 (17.1%)c
 Annual household income, no. (%) χ2(4) = 293.6, p < 0.001
  <$5000 to $29,999 7097 (38.8%)a 931 (33.0%)b 2285 (49.0%)c
  $30,000 to $89,999 7968 (43.5%)a 1313 46.6%)b 1839 (39.4%)c
  $90,000 to 200,000 or more 3232 (17.7%)a 574 (20.4%)b 539 11.6%)c
 Education no. (%) χ2(6) = 330.6, p < 0.001
  High school or less 6974 (38.1%)a 755 (26.8%)b 1788 (38.3%)a
  Some college 6321 (34.5%)a 1051 (37.3%)b 1782 (38.2%)b
  College grad 2621 (14.3%)a 435 (15.4%)a 543 (11.6%)b
  Postgrad 2381 (13.0%)a 577 (20.5%)b 550 (11.8%)c
No. AUD symptoms, (SD) 1.23 (3.04)a 1.35 (3.13)a 2.71 (5.09)b F (2, 25,775) = 333.9, p < 0.001
Average daily drinking volume (12 mo, oz) 0.66 (1.49)a 0.59 (1.37)a 0.82 (1.90)b F (2, 25,656) = 24.5, p < 0.001
Drinking behavior
 Binge drinking, no. (%) χ2(4) = 106.2, p < 0.001
  Never 10,240 (56.2%)a 1547 (55.1%)a 2263 (48.7%)b
  Less than once a week 5059 (27.8%)a 883 (31.4%)b 1505 (32.4%)b
  Once or more weekly 2917 (16.0%)a 379 (13.5%)b 880 (18.9%)c
Other alcohol variables
 Family history of AUD 1st-degree relative, no. (%) 7130 (39.0%)a 1506 (53.4%)b 2769 (59.4%)c χ2(2) = 1026.3, p < 0.001
 Years of drinking at past year level (SD) 10.3 (11.46)a 11.1 (11.24)b 8.7 (10.11)c F (2, 25,234) = 47.1, p < 0.001
DSM-V diagnoses, no. (%)
 Major depression NA 2362 (83.8%) 3304 (70.9%)
 Generalized anxiety disorder NA 434 (15.4%) 1412 (30.3%)
 Social anxiety disorder NA 137 (4.9%) 709 (15.2%)
 Panic disorder NA 323 (11.5%) 846 (18.1%)
 Alcohol use disorder 3147 (17.2%)a 561 (19.9%)b 1425 (30.6%)c χ2(2) = 415.9, p < 0.001
 Non-alcohol related substance use disorder 645 (3.5%)a 126 (4.5%)a 562 (12.1%)c χ2(2) = 554.2, p < 0.001
Open in a new tab

Note: Each subscript letter denotes a subset of INTD categories whose column proportions do not differ significantly from each other at the 0.05 level.

Abbreviation: INTD, internalizing disorder.

Measures

INTD diagnoses

The Alcohol Use Disorders and Associated Disabilities Interview Schedule V (AUDADIS-V; Grant et al., 2015) was administered to all NESARC participants by a trained interviewer. The AUDADIS-V is a highly structured diagnostic interview that assesses Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5; APA) criteria to produce categorical psychiatric diagnoses. Screening items assessing for the presence of the essential feature(s) of each disorder are used to determine whether further diagnostic questions would be asked: that is, if the screening items are not endorsed, participants are classified as not having the disorder and the remaining items related to that disorder are skipped. We used the NESARC AUDADIS-V results to identify cases with current (past 12 months) or remitted (before past 12 months but not within the past 12 months) INTD diagnosis(es); that is, major depression disorder, generalized anxiety disorder, panic disorder, and social anxiety disorder.

Alcohol symptoms

As part of the AUDADIS-V, respondents were presented a list of alcohol-related symptoms/problems—primarily drawn from the DSM AUD criteria—and asked to indicate which they had experienced in the past year. We excluded four of these items that were not DSM AUD symptoms (e.g., riding as a passenger in a car where the driver was drinking) and two additional items that did not match in NESARC 1 and 3. The number of affirmative responses to the remaining 32 items (see Table S1) was summed to produce a single variable representing AUD symptom count.

Alcohol volume

The NESARC survey collected information about the types and amounts of alcoholic beverages people consumed in the past year. Respondents were shown images of different sizes of glasses and asked to report how much they typically drink. The total amount of EtOH (pure alcohol) consumed in the past year was calculated by adding up the amounts of each type of beverage. This was then divided by 365 to find the average daily intake of EtOH.

Alcohol binging

Respondents were asked to estimate the typical frequency that they binged on alcohol (consuming five or more standard alcoholic drinks on one occasion for men or four or more alcoholic drinks on one occasion for women) over the past year. Responses were categorized as: “never occurs,” “occurs less than weekly,” or “occurs once or more weekly.”

Years of drinking at the same level

After assessing drinking patterns for the past year (above), interviewers asked respondents how many years they had drank at approximately the same level and pattern as in the past 12 months.

Family history of alcohol problems

Interviewers assessed family history of alcohol-related problems by asking respondents whether any first-degree relatives (biological parents, siblings, or adult children) had ever been “problem drinkers” or “alcoholics.” Responses were pooled into a single present vs. absent variable for study analyses.

Primary analyses

SPSS (Version_27) was used to calculate chi-square tests to compare groups on categorical variables (e.g., family history of AUD), and ANOVA was used to compare groups on continuous measures (e.g., past 12-month mean daily alcohol intake). The open-source statistical platform R (Core Team, 2022) was used to conduct generalized linear models examining predictors on the primary outcome of AUD symptom count. Given that the AUD symptom count was non-normally distributed (overdispersed Poisson distribution), a negative binomial estimation parameter (theta) was included in the generalized linear models. Our objective was to determine whether INTD status predicts AUD symptom count while controlling for alcohol consumption and other established predictors of an alcohol-related harm paradox effect (e.g., gender, SES, family history of AUD, and binge drinking). An iterative stepwise process was used to introduce covariates into models. Covariates were entered in the following order and included alcohol volume, socioeconomic and demographic variables (age, gender, annual household income, level of education, and ethnicity), binge drinking frequency, and AUD-related factors (family history of AUD and number of years drinking at the past year level). (Note that the race/ethnicity category was collapsed into a binary variable, “white” vs. “other,” for the purpose of analytic efficiency.) The model was also run on a subsample that excluded participants meeting a current/past year AUD diagnosis (N = 20,066).

Replication analyses

The reliability of the final model was evaluated by conducting an identical analysis using an independent sample from NESARC Wave 1 (2001 to 2002). NESARC Wave 1 included (largely) the same interview questions as NESARC Wave 3 (described above); however, for the NESARC Wave 1 sample, diagnoses were made using the DSM-IV criteria and assessed using the AUDADIS-IV (Grant et al., 2003). The same three INTD groups identified in the primary analyses using NESARC Wave 3 were also identified in the replication analyses using NESARC Wave 1 (INTD-Never [n = 19,935], INTD-Remitted [n = 3422], and INTD-Current [n = 3293]). As in the primary analyses, two replication analyses were conducted—one with AUD cases included (N = 26,062) and one with AUD cases excluded (N = 20,066).

RESULTS

Descriptive statistics

Table 1 presents demographic/socioeconomic characteristics, AUD symptoms, alcohol intake, frequency of binge drinking, family history of alcohol problems, and frequency of current (past 12 months) DSM-5 diagnoses for each INTD group. As shown, and with respect to demographic/socioeconomic characteristics, the INTD-Current group was significantly younger, had a lower annual household income, and was less likely to have completed college or a postgrad program than the INTD-Never and INTD-Remitted groups. Moreover, there was a significantly lower proportion of women in the INTD-Never group than in the INTD-Remitted and INTD-Current groups. Table 1 also shows that while a majority of respondents in all groups self-identified as “White,” the distribution of the several ethnicity categories differed significantly between the groups. Participants in the INTD-Remitted and INTD-Never groups reported drinking at their current level for a longer period than did the INTD-Current group. A significantly greater proportion in the INTD-Current group had current AUD compared to the INTD-Never and the INTD-Remitted groups. Finally, a significantly greater proportion of participants in the INTD-Current group had a current substance use disorder diagnosis compared with the INTD-Never and the INTD-Remitted groups. Demographic variables associated with significant group differences in the NESARC 3 dataset were the same demographic factors associated with significant group differences in the NESARC 1 dataset (see Table S2).

INTD status as a predictor of AUD symptoms controlling for covariates

All demographic/socioeconomic variables, AUD symptoms, alcohol intake, binge drinking frequency, family history of alcohol problems, and years drinking at the same level significantly differed between groups and were thus included in the negative binomial regression analyses. The incident rate ratio (IRR) is reported and indicates the proportional change in the AUD symptom count for every unit change in the predictor. Since our primary interest was to assess the extent that INTD status contributed to AUD symptom count while controlling for non-INTD-specific harm paradox predictors, INTD status was included as a predictor in the model with other variables treated as covariates and entered in a stepwise fashion (see Methods). The results of stepwise modeling are shown in Table S3.

Table 2 displays the results of the negative binomial regression model predicting AUD symptoms from INTD group status among current drinkers. The results confirm the prediction that INTD status uniquely predicts AUD symptoms after accounting for level of drinking and other covariates. Specifically, results from the negative binomial regression model (controlling for covariates) indicated that AUD symptoms were significantly higher in the INTD-Current (IRR = 1.72, p < 0.001) and the INTD-Remitted (IRR = 1.20, p < 0.001) groups relative to the INTD-Never group. Covariates were also highly predictive of AUD symptom count. Specifically, with respect to demographic/socioeconomic variables, being a younger age, male, and graduating college or a postgraduate program (vs. high school diploma or less) was associated with a higher AUD symptom count while being white or earning $30 K-$200 K (vs. <$29 K) was associated with a lower AUD symptom count. Not surprisingly, current alcohol intake (daily ounces) was also significantly associated with AUD symptom count. In addition, compared with those who never binged, AUD symptoms were significantly higher among those who reported binging once weekly (IRR = 4.84, p < 0.001) and binging more than once weekly (IRR = 7.87, p < 0.001). Having a first-degree relative with alcohol use problems was also significantly and positively associated with AUD symptom count (IRR = 1.45, p < 0.001), while years of drinking at the level of the past year was significantly and negatively associated with alcohol dependence symptom count (IRR = 0.99, p < 0.001). With the exceptions of household income (no difference between ≥90 K vs. ≤29 K) and education (no difference between postgrad vs. high school or less), results from the replication dataset (NESARC Wave 1) were consistent with those reported here from the primary dataset (NESARC Wave 3; Table S4).

TABLE 2.

Current Drinkers (NESARC 3, N = 25,091): Results from negative binomial regression with INTD status, demographic/socioeconomic categorization, frequency of binge drinking, and history with alcohol as independent variables and AUD symptom count as the dependent variable.

CI
B SE IRR 2.50% 97.5% p
Groups (ref INTD-never)
 INTD-current 0.54 0.03 1.72 1.62 1.83 ***
 INTD-remitted 0.18 0.04 1.20 1.11 1.30 ***
Demographic/socioeconomic category
 Age −0.02 <0.01 0.98 0.98 0.98 ***
 Gender (male) 0.25 0.03 1.28 1.22 1.35 ***
 Ethnicity (white) −0.14 0.03 0.87 0.83 0.91 ***
 Household income (ref ≤$29,000)
  $30,000 to $89,999 −0.14 0.03 0.87 0.83 0.92 ***
  $90,000 to $200,000 or more −0.11 0.04 0.90 0.83 0.97 **
 Education (ref high school or less)
  Some college 0.04 0.03 1.04 0.98 1.10
  College graduate 0.10 0.04 1.12 1.03 1.21 **
  Postgraduate 0.11 0.04 1.11 1.03 1.22 **
Average daily drinking volume 0.28 0.01 1.32 1.29 1.34 ***
Drinking behavior
 Alcohol binging (ref never)
  Less than once a week 1.57 0.03 4.83 4.55 5.11 ***
  Weekly 2.07 0.04 7.95 7.39 8.57 ***
Other alcohol variables
 Years of drinking at past year level −0.01 <0.01 0.99 0.98 0.99 ***
 Family history of AUD 1st-degree relative 0.38 0.03 1.46 1.39 1.53 ***
Open in a new tab

Abbreviations: AUD, alcohol use disorder; B, regression coefficient; CI, 95% confidence interval; INTD, internalizing disorder; IRR, incident rate ratio; ref, referent; SE, standard error.

**

p-value <0.01

***

p-value <0.001.

Sensitivity analyses

Exclusion of those with current AUD

Given that AUD symptom count was understandably highly correlated with AUD diagnosis, along with the possibility that those with AUD represent a distinct population in terms of study hypotheses, we evaluated the predictive model on a subsample from NESARC 3 that excluded participants meeting a current AUD diagnosis (N = 20,528). Results from this analysis indicated that all but two predictors (ethnicity and household income) remained significant and in the same direction of association as found in the full model containing AUD cases (Table 3).

TABLE 3.

Analysis with AUD cases excluded (NESARC 3, N = 20,066): Results from negative binomial regression with INTD status, demographic/socioeconomic category, frequency of binge drinking, and history with alcohol as independent variables and AUD symptom count as the dependent variable. AUD cases removed from analysis to examine specificity.

CI
B SE IRR 2.50% 97.5% p
Groups (ref INTD-never)
 INTD-current 0.28 0.05 1.32 1.20 1.44 ***
 INTD-remitted 0.19 0.06 1.21 1.08 1.35 ***
Demographic/socioeconomic category
 Age −0.01 <0.01 0.98 0.98 0.99 ***
 Gender (male) 0.16 0.04 1.17 1.09 1.25 ***
 Ethnicity (white) −0.02 0.04 0.98 0.91 1.05
 Household income (ref ≤$29,000)
  $30,000 to $89,999 −0.05 0.04 0.94 0.88 1.02
  $90,000 to $200,000 or more <0.01 0.06 1.00 0.91 1.12
 Education (ref high school or less)
  Some college 0.04 0.04 1.04 0.96 1.13
  College grad 0.12 0.06 1.12 1.01 1.26 *
  Postgrad 0.17 0.06 1.19 1.06 1.34 **
Average daily drinking volume 0.19 0.02 1.21 1.16 1.26 ***
Drinking behavior
 Alcohol binging (ref never)
  Less than once a week 1.25 0.04 3.50 3.23 3.78 ***
  Weekly 1.29 0.06 3.64 3.22 4.11 ***
Other alcohol variables
 Years of drinking at past year level −0.01 <0.01 0.99 0.99 0.99 ***
 Family history of AUD 1st-degree relative 0.13 0.06 1.13 1.06 1.22 ***
Open in a new tab

Abbreviations: AUD, alcohol use disorder; B, regression coefficient; CI, 95% confidence interval; INTD, internalizing disorder; IRR, incident rate ratio; ref, referent; SE, standard error.

*

p-value <0.05.

**

p-value <0.01.

***

p-value <0.001.

Separating anxiety and depression disorders

To assess the influence of our having used the single INTD predictor, we reran the primary model with anxiety disorders and depression disorder (both current and remitted) as separate predictors. Anxiety-Remitted (IRR = 1.32), Anxiety-Current (IRR = 1.41), Depression-Remitted (IRR = 1.15), and Depression-Current (IRR = 1.56) all remained significant predictors (p < 0.001) of AUD symptoms in the full models. This shows that both INTD disorder domains contribute similarly to the predicted harm paradox effect.

Number of INTDs as a predictor

Because INTD load (defined as the number of INTD diagnoses present) has been shown to correlate positively with AUD risk (Kushner et al., 2012b), we evaluated whether the number of current INTD diagnoses had a positive association with the alcohol-related harm paradox effect in the full model (i.e., with all covariates included). Compared to those with one current INTD diagnosis, those with two diagnoses had an IRR = 1.15, those with three diagnoses had an IRR = 1.34, and those with four diagnoses had an IRR = 1.49. As a continuous count variable, the effect is IRR = 1.15 (p < 0.0001); that is, the number of AUD symptoms increased by 15% for each additional INTD diagnosis beyond one.

Gender as a possible moderator

We covaried gender in the primary models because women have been shown to experience an alcohol-related harm paradox effect relative to men (e.g., Randall et al., 1999). Additionally, because women are less likely to develop AUD and more likely to develop INTD than are men (e.g., Hasin & Grant, 2015), we considered it prudent to evaluate the interaction between INTD and gender in predicting AUD symptoms in the full model. The interaction term in this sensitivity analysis was not significant. This suggests that the effect of INTD status in predicting the harm paradox effect is similar for both men and women.

Replication analyses using NESARC wave 1

Finally, except for gender (no difference) and education (no difference between postgrad vs. high school or less), results from the replication analyses using NESARC Wave 1 conformed to results obtained in the primary analyses using NESARC Wave 3 (see Table S4—whole sample—and Table S5—excludes participants with AUD).

DISCUSSION

We hypothesized that shared neurobiological dysregulations underlying AUD and INTD contribute to the development of comorbidity. Based on this, we predicted that those with INTD experience more alcohol-related symptoms than others who drink at the same level. This phenomenon—more alcohol-related problems relative to others who drink at the same level—has been termed a “harm paradox effect.” Past studies have identified several characteristics that correlate with an alcohol-related harm paradox effect, including SES (e.g., Erskine et al., 2010; Harrison & Gardiner, 1999; Jones et al., 2015), gender (e.g., Randall et al., 1999), binge drinking (e.g., Lewer et al., 2016), and a family history of AUD (e.g., Schuckit et al., 2009). By statistically controlling these variables, we increase confidence in the conclusion that INTD is a hitherto unidentified marker of a specific alcohol-related harm paradox effect. By replicating this finding in an independent epidemiological sample, we increase confidence in the reliability of this effect. Importantly, however, the NESARC did not include data capable of directly evaluating the role of neurobiological processes in causing the harm paradox effect we demonstrated.

There is some evidence to suggest that neurobiological factors present since birth (“congenital”) confer vulnerability to developing both INTD and AUD. For example, dysregulated stress-responding linked to risk of developing INTD can also mark a risk of developing AUD (e.g., Clarke et al., 2012). Alternatively, there is some evidence to suggest that the frank manifestation of INTD can produce an enduring neurobiological change (a “scar”) that renders individuals more vulnerable to developing AUD symptoms. For example, it has been shown that having experienced acute or chronic stress (e.g., in the context of an INTD) can lead to permanently dysregulated stress responding, which, in turn, can increase the risk of AUD (Brady & Back, 2012). Notably, the “congenital” versus “scar” pathways to shared neurobiology considered here are not mutually exclusive possibilities.

Although we tested the prediction that INTD confers an alcohol-related harm paradox effect, it is important to note that the shared neurobiology hypothesis would equally imply that AUD can confer an INTD-related harm paradox effect. This understanding of our hypothesis—that both INTD and AUD mark a neurobiologically mediated susceptibility to developing the other—is consistent with retrospective studies showing that the order in which comorbid conditions onset is variable (e.g., Kushner et al., 1990, 2000), as well as prospective studies showing that having either INTD or AUD alone substantially increases the likelihood of developing the other condition in the future (e.g., Kushner et al., 1999). The “congenital” pathway to shared neurobiology (see above) is consistent with the notion that the presence of either INTD or AUD marks a neurobiological status conducive to the development of the other. Additionally, there are theoretical (e.g., Koob and Le Moal, 2008), clinical (e.g., Schuckit & Hesselbrock, 1994), and empirical (e.g., Barkell et al., 2022) observations suggesting that the neurobiological consequences of chronic alcohol use (a “scar”) can serve to increase the risk of developing INTD symptoms and possibly syndromes. While an INTD-related harm paradox effect associated with AUD is implied by our hypothesis and important if true, studying this effect is beyond the scope of the present work.

Limitations

Our use of a single INTD predictor to represent the presence of any of a variety of anxiety and depression disorders was premised on past work showing that AUD risk is associated with the shared rather than the unique variance of these internalizing conditions (Kushner et al., 2012b). However, this approach risks obscuring any important differences in the contribution of specific INTD domains to the harm paradox effect. To mitigate this risk, we conducted sensitivity analyses in which the primary model was rerun with anxiety and depression disorders entered as separate variables. These analyses showed that both anxiety disorder and depression disorder predicted the harm paradox effect when the other was effectively controlled in the model. Given this, it is not surprising that additional sensitivity analyses showed that the size of the alcohol-related harm paradox effect increases as the number of individual INTD diagnoses present increases. These findings indicate that, like the general association between INTD and AUD risk (Kushner et al., 2012b), the alcohol-related harm paradox effect is more strongly associated with overall INTD load than INTD type.

The absence of continuous measures of anxiety and depression symptoms disallowed our examination of their association with the alcohol-related harm paradox effect. Having found that INTD load (defined as the number of INTD diagnoses present) was positively associated with the alcohol-related harm paradox effect, we might expect that subclinical INTD symptom level would also correlate positively with the harm paradox effect; however, this is not certain. For example, it is possible that a relatively high threshold of INTD symptoms is required to either cause a neurobiological “scar” or mark a sufficient level of “congenital” neurobiological dysregulation necessary to initiate susceptibility to the development of AUD symptoms.

Although we controlled main effects for age and gender in the primary analyses, these variables might have more complex relationships to key study variables. The prevalence of both INTD and AUD is subject to birth cohort effects (e.g., Grant et al., 2017; Weinberger et al., 2018), and gender effects (Hasin & Grant, 2015). The latter issue is more tractable in our dataset, in which we could control gender both as a covariate (primary analyses) and in interaction with INTD (sensitivity analyses). These analyses allow us to say that the predicted harm paradox effect is neither dependent upon nor modified significantly by gender. The possibility that respondent’s birth cohort effected our results is more challenging to assess in these cross-sectional data, which confounds age and birth cohort; that is, age and birth cohort are highly correlated in these data with no way to disaggregate their variance in predictive models. Nonetheless, it was notable that the INTD-Current group was significantly younger than the INTD-Remitted group. While this could reflect a birth cohort effect, it is as or more likely a reflection of the general tendency for INTDs to onset in young adulthood. Under these conditions, somewhat younger adults in the sample would be more likely to have a current INTD, while somewhat older adults in the sample would be more likely to have a remitted INTD.

Other clinical and longitudinal parameters of INTD unavailable in the present study might also have provided useful information concerning the harm paradox effect. INTD chronicity, duration, frequency, and developmental stage at onset have all been shown to correlate with AUD risk (Boschloo et al., 2013; Essau et al., 2014; Kushner et al., 2012b) and might influence the harm paradox effect studied here. Similarly, while we controlled average daily alcohol intake level in the past year, it would have been desirable to know respondents’ lifetime patterns of onset(s), offset(s), and fluctuations in both drinking behavior and AUD symptoms, including how these variables relate to various INTD milestones. Although we did control for the number of years respondents reported drinking at a similar level to that of the “past year,” this provides no information about drinking behavior in prior years respondents viewed as dissimilar to that of the “past year.” Having had such longitudinal data would have potentially allowed for a better understanding of the pathway(s) by which INTD influences the course of AUD symptoms.

Another potential limitation of the present work (particularly in an applied sense) is that while the analyses confirmed and replicated study predictions, the effects were modest in size. This is not surprising given that the models included numerous covariates that have a strong association with AUD symptoms (e.g., family history of AUD, gender, drinking volume, and binge drinking frequency). Had we not controlled these variables, the association of INTD with AUD symptoms would likely have been much stronger, but then we would have failed to provide a scientifically rigorous test for our prediction. Nonetheless, our having demonstrated that INTD increases one’s susceptibility to the development of AUD symptoms opens the door for future work aimed at evaluating the public health/clinical implications of this effect. For example, public health policymakers might consider whether the alcohol-related harm paradox demonstrated here has implications for modified “safe” drinking level recommendations for those with an INTD, as is already the case for gender.

CONCLUSIONS

Noting neurobiological dysregulations understood to be relevant to both AUD and INTD, we hypothesized here (and elsewhere) that those with INTD are prone to developing AUD symptoms more readily than others who drink at the same level. Based on this hypothesis, we confirmed and replicated the straightforward and otherwise counter-intuitive prediction that those with INTD experience a higher level of alcohol-related symptoms than do those without INTD who drink at the same level. Given the large sample size, the rigor of the data collection method, and the inclusion of an independent replication, we feel confident in asserting the validity and reliability of the findings we report here. However, in the absence of direct neurobiological measures, confirmation of the alcohol-related harm paradox effect for those with INTD provides supportive but not definitive evidence for the shared neurobiology hypothesis on which this prediction was uniquely based. We hope these supportive findings will inspire researchers to pursue this novel neurobiological hypothesis of comorbidity with increased focus and vigor.

Supplementary Material

Supplementary Material

Funding information

National Institute on Alcohol Abuse and Alcoholism, Grant/Award Number: K01AA024805 and R01AA029077

This work was supported by NIAAA grant R01 AA029077 awarded to the last author and NIAAA K01 AA024805 awarded to the first author.

Footnotes

CONFLICT OF INTEREST STATEMENT

The authors have no conflicts of interest to report.

SUPPORTING INFORMATION

Additional supporting information can be found online in the Supporting Information section at the end of this article.

REFERENCES

  1. Abram KM, Zwecker NA, Welty LJ, Hershfield JA, Dulcan MK & Teplin LA (2015) Comorbidity and continuity of psychiatric disorders in youth after detention: a prospective longitudinal study. JAMA Psychiatry, 72, 84–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Agoglia AE & Herman MA (2018) The center of the emotional universe: alcohol, stress, and CRF1 amygdala circuitry. Alcohol, 72, 61–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Anker JJ, & Kushner MG (2019). Co-occurring alcohol used disorder and anxiety: Bridging psychiatric, psychological and neurobiological perspectives. Alcohol Research: Current Reviews, 40, 1–12. 10.35946/arcr.v40.1.03 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Anker JJ, Kushner MG, Thuras P, Menk J & Unruh AS (2016) Drinking to cope with negative emotions moderates alcohol use disorder treatment response in patients with co-occurring anxiety disorder. Drug and Alcohol Dependence, 159, 93–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barkell GA, Parekh SV, Paniccia JE, Martin AJ, Reissner KJ, Knapp DJ et al. (2022) Chronic EtOH consumption exacerbates future stress-enhanced fear learning, an effect mediated by dorsal hippocampal astrocytes. Alcoholism, Clinical and Experimental Research, 46, 2177–2190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Boschloo L, Vogelzangs N, Smit JH, van den Brink W, Veltman DJ, Beekman ATF et al. (2011) Comorbidity and risk indicators for alcohol use disorders among persons with anxiety and/or depressive disorders: findings from The Netherlands study of depression and anxiety (NESDA). Journal of Affective Disorders, 131, 233–242. [DOI] [PubMed] [Google Scholar]
  7. Boschloo L, Vogelzangs N, van den Brink W, Smit JH, Veltman DJ, Beekman ATF et al. (2013) Depressive and anxiety disorders predicting first incidence of alcohol use disorders: results of The Netherlands study of depression and anxiety (NESDA). The Journal of Clinical Psychiatry, 74, 1233–1240. [DOI] [PubMed] [Google Scholar]
  8. Boyd J, Sexton O, Angus C, Meier P, Purshouse RC & Holmes J (2021) Causal mechanisms proposed for the alcohol harm paradox-a systematic review. Addiction, 117, 33–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Brady KT & Back SE (2012) Childhood trauma, posttraumatic stress disorder, and alcohol dependence. Alcohol Research: Current Reviews, 34, 408–413. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Breese GR, Sinha R & Heilig M (2011) Chronic alcohol neuroadaptation and stress contribute to susceptibility for alcohol craving and relapse. Pharmacology & Therapeutics, 129, 149–171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Clarke TK, Nymberg C, & Schumann G (2012). Genetic and environmental determinants of stress responding. Alcohol Health and Research World, 34, 484–494. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Team Core. (2022). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria https://www.R-project.org/
  13. Davidson KM & Blackburn IM (1998) Co-morbid depression and drinking outcome in those with alcohol dependence. Alcohol and Alcoholism, 33, 482–487. [DOI] [PubMed] [Google Scholar]
  14. Dyr W & Kostowski W (1995) Evidence that the amygdala is involved in the inhibitory effects of 5-HT3 receptor antagonists on alcohol drinking in rats. Alcohol, 12, 387–391. [DOI] [PubMed] [Google Scholar]
  15. Erskine S, Maheswaran R, Pearson T & Gleeson D (2010) Socioeconomic deprivation, urban-rural location and alcohol-related mortality in England and Wales. BMC Public Health, 10, 99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Essau CA, Lewinsohn PM, Olaya B, & Seeley JR (2014). Anxiety disorders in adolescents and psychoscocial outcomes at age 30. Journal of Affective Disorders, 163, 125–132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fergusson DM, Boden JM & Horwood LJ (2009) Tests of causal links between alcohol abuse or dependence and major depression. Archives of General Psychiatry, 66, 260–266. [DOI] [PubMed] [Google Scholar]
  18. Fröjd S, Ranta K, Kaltiala-Heino R & Marttunen M (2011) Associations of social phobia and general anxiety with alcohol and drug use in a community sample of adolescents. Alcohol and Alcoholism, 46, 192–199. [DOI] [PubMed] [Google Scholar]
  19. Gilpin NW, Herman MA & Roberto M (2015) The central amygdala as an integrative hub for anxiety and alcohol use disorders. Biological Psychiatry, 77, 859–869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Grant B (1996) The relationship between DSM-IV alcohol use disorders and DSM-IV major depression: examination of the primary-secondary distinction in a general population sample. Journal of Affective Disorders, 38, 113–128. [DOI] [PubMed] [Google Scholar]
  21. Grant BF, Chou SP, Saha TD, Pickering RP, Kerridge BT, Ruan WJ et al. (2017) Prevalence of 12-month alcohol use, high-risk drinking, and DSM-IV alcohol use disorder in the United States, 2001–2002 to 2012–2013: results from the National Epidemiologic Survey on alcohol and related conditions. JAMA Psychiatry, 74, 911–923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Grant BF, Chu A, Sigman R, Amsbary M, Kali J, Sugawara Y et al. (2014) NESARC-iii. National Epidemiologic Survey on alcohol and related conditions-III (NESARC-III): source and accuracy Statement
  23. Grant BF, Dawson DA, Stinson FS, Chou PS, Kay W & Pickering R (2003) The alcohol use disorder and associated disabilities interview schedule-IV (AUDADIS-IV): reliability of alcohol consumption, tobacco use, family history of depression and psychiatric diagnostic modules in a general population sample. Drug and Alcohol Dependence, 71, 7–16. [DOI] [PubMed] [Google Scholar]
  24. Grant BF, Goldstein RB, Smith SM, Jung J, Zhang H, Chou SP et al. (2015) The alcohol use disorder and associated disabilities interview Schedule-5 (AUDADIS-5): reliability of substance use and psychiatric disorder modules in a general population sample. Drug and Alcohol Dependence, 148, 27–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Han RT, Kim Y-B, Park E-H, Kim JY, Ryu C, Kim HY et al. (2018) Long-term isolation elicits depression and anxiety-related behaviors by reducing oxytocin-induced GABAergic transmission in central amygdala. Frontiers in Molecular Neuroscience, 11, 246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Harrison L & Gardiner E (1999) Do the rich really die young? Alcohol-related mortality and social class in Great Britain, 1988–94. Addiction, 94, 1871–1880. [DOI] [PubMed] [Google Scholar]
  27. Hasin DS & Grant BF (2015) The National Epidemiologic Survey on alcohol and related conditions (NESARC) waves 1 and 2: review and summary of findings. Social Psychiatry and Psychiatric Epidemiology, 50, 1609–1640. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hill SY, De Bellis MD, Keshavan MS, Lowers L, Shen S, Hall J et al. (2001) Right amygdala volume in adolescent and young adult offspring from families at high risk for developing alcoholism. Biological Psychiatry, 49, 894–905. [DOI] [PubMed] [Google Scholar]
  29. Hunt GE, Malhi GS, Lai HMX & Cleary M (2020) Prevalence of comorbid substance use in major depressive disorder in community and clinical settings, 1990–2019: systematic review and meta-analysis. Journal of Affective Disorders, 266, 288–304. [DOI] [PubMed] [Google Scholar]
  30. Hyytiä P & Koob GF (1995) GABAA receptor antagonism in the extended amygdala decreases EtOH self-administration in rats. European Journal of Pharmacology, 283, 151–159. [DOI] [PubMed] [Google Scholar]
  31. Jones L, Bates G, McCoy E & Bellis MA (2015) Relationship between alcohol-attributable disease and socioeconomic status, and the role of alcohol consumption in this relationship: a systematic review and meta-analysis. BMC Public Health, 15, 400. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Koob GF (2013) Theoretical frameworks and mechanistic aspects of alcohol addiction: alcohol addiction as a reward deficit disorder. Current Topics in Behavioral Neurosciences, 13, 3–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Koob GF & Le Moal M (2001) Drug addiction, dysregulation of reward, and allostasis. Neuropsychopharmacology, 24, 97–129. [DOI] [PubMed] [Google Scholar]
  34. Koob GF, & Le Moal M (2008). Addiction and the brain antireward system. Annual Review of Psychology, 59, 29–53. [DOI] [PubMed] [Google Scholar]
  35. Kranzler HR, Del Boca FK & Rounsaville BJ (1996) Comorbid psychiatric diagnosis predicts three-year outcomes in alcoholics: a posttreatment natural history study. Journal of Studies on Alcohol, 57, 619–626. [DOI] [PubMed] [Google Scholar]
  36. Kushner MG (2014) Seventy-five years of comorbidity research. Journal of Studies on Alcohol and Drugs. Supplement, 75(Suppl 17), 50–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Kushner MG, Abrams K & Borchardt C (2000) The relationship between anxiety disorders and alcohol use disorders: a review of major perspectives and findings. Clinical Psychology Review, 20, 149–171. [DOI] [PubMed] [Google Scholar]
  38. Kushner MG, Maurer E, Menary KR, & Thuras P (2011). Vulnerbility to the rapid (“telescoped”) development of alcohol depdendence in individuals with anxiety disorder. Journal of Studies on Alcohol and Drugs, 72, 1019–1027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Kushner MG, Sher KJ & Beitman BD (1990) The relation between alcohol problems and the anxiety disorders. The American Journal of Psychiatry, 147, 685–695. [DOI] [PubMed] [Google Scholar]
  40. Kushner MG, Sher KJ & Erickson DJ (1999) Prospective analysis of the relation between DSM-III anxiety disorders and alcohol use disorders. The American Journal of Psychiatry, 156, 723–732. [DOI] [PubMed] [Google Scholar]
  41. Kushner MG, Menary KR, Maurer EW, & Thuras P (2012a) Greater elevation in risk for nicotine dependence per pack of cigarettes smoked among those with anxiety disorder. Journal of Studies on Alcohol and Drugs, 73, 920–924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Kushner MG, Wall MM, Krueger RF, Sher KJ, Maurer E, Thuras P et al. (2012b) Alcohol dependence is related to overall internalizing psychopathology load rather than to particular internalizing disorders: evidence from a national sample. Alcoholism, Clinical and Experimental Research, 36, 325–331. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Lai HMX, Cleary M, Sitharthan T & Hunt GE (2015) Prevalence of comorbid substance use, anxiety and mood disorders in epidemiological surveys, 1990–2014: a systematic review and meta-analysis. Drug and Alcohol Dependence, 154, 1–13. [DOI] [PubMed] [Google Scholar]
  44. Lewer D, Meier P, Beard E, Boniface S & Kaner E (2016) Unravelling the alcohol harm paradox: a population-based study of social gradients across very heavy drinking thresholds. BMC Public Health, 16, 599. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Mäkelä P (1999) Alcohol-related mortality as a function of socioeconomic status. Addiction, 94, 867–886. [DOI] [PubMed] [Google Scholar]
  46. Menary KR, Kushner MG, Maurer E, & Thuras P (2011). The prevalence and clinical implications of self-medication among individuals with anxiety disorder. Journal of Anxiety Disorders, 25, 335–339 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Markou A, Kosten TR & Koob GF (1998) Neurobiological similarities in depression and drug dependence: a self-medication hypothesis. Neuropsychopharmacology, 18, 135–174. [DOI] [PubMed] [Google Scholar]
  48. Marmorstein NR (2015) Interactions between internalizing symptoms and urgency in the prediction of alcohol use and expectancies among low-income, minority early adolescents. Subst Abuse, 9, 59–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. McHugh RK & Weiss RD (2019) Alcohol use disorder and depressive disorders. Alcohol Research: Current Reviews, 40, e1–e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Möller C, Wiklund L, Sommer W, Thorsell A & Heilig M (1997) Decreased experimental anxiety and voluntary EtOH consumption in rats following central but not basolateral amygdala lesions. Brain Research, 760, 94–101. [DOI] [PubMed] [Google Scholar]
  51. Mueller TI, Lavori PW, Keller MB, Swartz A, Warshaw M, Hasin D et al. (1994) Prognostic effect of the variable course of alcoholism on the 10-year course of depression. The American Journal of Psychiatry, 151, 701–706. [DOI] [PubMed] [Google Scholar]
  52. Noronha A, Cui C, Harris RA & Crabbe JC (2014) Neurobiology of alcohol dependence Amsterdam: Elsevier Academic Press. [Google Scholar]
  53. Pardini D, White HR & Stouthamer-Loeber M (2007) Early adolescent psychopathology as a predictor of alcohol use disorders by young adulthood. Drug and Alcohol Dependence, 88(Suppl 1), S38–S49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Randall CL, Roberts JS, Del Boca FK, Carroll KM, Connors GJ & Mattson ME (1999) Telescoping of landmark events associated with drinking: a gender comparison. Journal of Studies on Alcohol, 60, 252–260. [DOI] [PubMed] [Google Scholar]
  55. Roberto M & Gilpin NW (2014) Central amygdala neuroplasticity in alcohol dependence. Neurobiology of Alcohol Dependence. [DOI] [PMC free article] [PubMed]
  56. Schuckit MA & Hesselbrock V (1994) Alcohol dependence and anxiety disorders: what is the relationship? The American Journal of Psychiatry, 151, 1723–1734. [DOI] [PubMed] [Google Scholar]
  57. Schuckit MA, Smith TL, Danko GP, Trim R, Bucholz KK, Edenberg HJ et al. (2009) An evaluation of the full level of response to alcohol model of heavy drinking and problems in COGA offspring. Journal of Studies on Alcohol and Drugs, 70, 436–445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Substance Abuse and Mental Health Services Administration (SAMHSA). (2020) Key substance use and mental health indicators in the United States: Results from the 2019 National Survey on Drug Use and Health (HHS Publication No. PEP20-07-01, NSDUH Series H-55) Rockville MD: Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration. https://www.samhsa.gov/data/ [Google Scholar]
  59. VanElzakker MB, Dahlgren MK, Davis FC, Dubois S & Shin LM (2014) From Pavlov to PTSD: the extinction of conditioned fear in rodents, humans, and anxiety disorders. Neurobiology of Learning and Memory, 113, 3–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Vinkers CH, Kuzminskaite E, Lamers F, Giltay EJ & Penninx BWJH (2021) An integrated approach to understand biological stress system dysregulation across depressive and anxiety disorders. Journal of Affective Disorders, 283, 139–146. [DOI] [PubMed] [Google Scholar]
  61. Weinberger AH, Gbedemah M, Martinez AM, Nash D, Galea S & Goodwin RD (2018) Trends in depression prevalence in the USA from 2005 to 2015: widening disparities in vulnerable groups. Psychological Medicine, 48, 1308–1315. [DOI] [PubMed] [Google Scholar]
  62. Wolitzky-Taylor K, Bobova L, Zinbarg RE, Mineka S & Craske MG (2012) Longitudinal investigation of the impact of anxiety and mood disorders in adolescence on subsequent substance use disorder onset and vice versa. Addictive Behaviors, 37, 982–985. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplementary Material
NIHMS1907127-supplement-Supplementary_Material.docx (63.2KB, docx)

RESOURCES