Drinking to Cope Mediates the Association between Dyadic Conflict and Drinking Behavior: A Study of Romantic Couples during the COVID-19 Pandemic

Abstract

The COVID-19 pandemic spurred public health measures to reduce viral spread. Concurrently, increases in alcohol consumption and conflict in romantic partnerships were observed. Pre-pandemic research demonstrated a bidirectional association between couples’ conflict and drinking. Recent research shows one’s drinking motives (proximal predictors of drinking behavior) can influence another person’s drinking in close relationships. It is possible that individuals are drinking to cope with distress following romantic conflict. The current study examined 348 cohabitating couples during the first lockdown in the spring of 2020. Our analyses examined coping motives as a mediator between dyadic conflict and drinking behavior using actor–partner interdependence models. Results showed that conflict was associated with greater reports of own drinking in gendered (distinguishable) and nongendered (indistinguishable) analyses through coping motives. Further, in mixed-gender couples, men partners’ coping motives predicted less drinking in women, while women partners’ coping motives predicted marginally more drinking in men. Partner effects may have been observed due to the increased romantic partner influence during the COVID-19 lockdown. While these results suggest that men’s coping motives may be protective against women’s drinking, more concerning possibilities are discussed. The importance of considering dyadic influences on drinking is highlighted; clinical and policy implications are identified.

Full text

Citation: Hagen, A.E.F.; Rodriguez,
L.M.; Neighbors, C.;
Nogueira-Arjona, R.; Sherry, S.B.;
Lambe, L.; Deacon, S.H.; Meier, S.;
Abbass, A.; Stewart, S.H. Drinking to
Cope Mediates the Association
between Dyadic Conflict and
Drinking Behavior: A Study of
Romantic Couples during the
COVID-19 Pandemic. Int. J. Environ.
Res. Public Health 2023,20, 6332.
https://doi.org/10.3390/
ijerph20146332
Academic Editor: Elizabeth C. Penick
Received: 31 May 2023
Revised: 3 July 2023
Accepted: 4 July 2023
Published: 10 July 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
International Journal of
Environmental Research
and Public Health
Article
Drinking to Cope Mediates the Association between Dyadic
Conflict and Drinking Behavior: A Study of Romantic Couples
during the COVID-19 Pandemic
Amanda E. F. Hagen 1,*, Lindsey M. Rodriguez 2, Clayton Neighbors 3, Raquel Nogueira-Arjona 4,
Simon B. Sherry 1, Laura Lambe 1, S. Hélène Deacon 1, Sandra Meier 5, Allan Abbass 5and Sherry H. Stewart 5, 1,*
1Department of Psychology & Neuroscience, Dalhousie University, Halifax, NS B3H 4R2, Canada;
simon.sherry@dal.ca (S.B.S.); laura.lambe@dal.ca (L.L.); helene.deacon@dal.ca (S.H.D.)
2
Department of Psychology, University of Florida, Gainesville, FL 33701, USA; lindsey.rodriguez1@gmail.com
3Department of Psychology, University of Houston, Houston, TX 77204, USA; cneighbors@uh.edu
4School of Psychology, University of Sussex, Brighton BN1 9RH, UK; r.nogueira-arjona@sussex.ac.uk
5Department of Psychiatry, Dalhousie University, Halifax, NS B3H 2E2, Canada;
sandra.meier@iwk.nshealth.ca (S.M.); allan.abbass@dal.ca (A.A.)
*Correspondence: m.hagen@dal.ca (A.E.F.H.); sstewart@dal.ca (S.H.S.)
Abstract:
The COVID-19 pandemic spurred public health measures to reduce viral spread. Con-
currently, increases in alcohol consumption and conflict in romantic partnerships were observed.
Pre-pandemic research demonstrated a bidirectional association between couples’ conflict and drink-
ing. Recent research shows one’s drinking motives (proximal predictors of drinking behavior) can
influence another person’s drinking in close relationships. It is possible that individuals are drinking
to cope with distress following romantic conflict. The current study examined 348 cohabitating
couples during the first lockdown in the spring of 2020. Our analyses examined coping motives
as a mediator between dyadic conflict and drinking behavior using actor–partner interdependence
models. Results showed that conflict was associated with greater reports of own drinking in gendered
(distinguishable) and nongendered (indistinguishable) analyses through coping motives. Further,
in mixed-gender couples, men partners’ coping motives predicted less drinking in women, while
women partners’ coping motives predicted marginally more drinking in men. Partner effects may
have been observed due to the increased romantic partner influence during the COVID-19 lockdown.
While these results suggest that men’s coping motives may be protective against women’s drinking,
more concerning possibilities are discussed. The importance of considering dyadic influences on
drinking is highlighted; clinical and policy implications are identified.
Keywords: COVID-19; alcohol use; couples’ conflict; drinking motives
1. Introduction
In the spring of 2020, businesses, travel, workplaces, and schools were closed across
Canada and many jurisdictions around the world in response to the declaration of the
COVID-19 virus as a pandemic [
1
]. As such, many cohabitating romantic couples found
themselves spending extended periods of time in the home together. Couples also had
highly limited opportunities to socialize with others or participate in activities outside
the home. Romantic couples also faced additional pandemic-specific stressors such as
shifts to working from home, mandatory homeschooling, and financial and/or health-
related concerns.
The effects of the COVID-19 pandemic and associated public health measures have
resulted in deleterious consequences to individuals’ well-being, including increases in
anxiety and depressive symptoms, increased loneliness, and decreased positive emotions
and life satisfaction [
2
,
3
]. Further, some evidence suggested increased individual alcohol
Int. J. Environ. Res. Public Health 2023,20, 6332. https://doi.org/10.3390/ijerph20146332 https://www.mdpi.com/journal/ijerph

Int. J. Environ. Res. Public Health 2023,20, 6332 2 of 19
use and alcohol-related problems during COVID-19 lockdowns [
4
,
5
], particularly among
those with higher levels of pandemic-related distress [
6
,
7
]. The population’s increase in
alcohol use is of great public health importance due to the range of adverse consequences
(e.g., injury, medical illness, suicide, and violence) associated with heavier drinking [
8
].
Despite increases in reports of stressful life events frequently associated with increases in
romantic conflict (e.g., job loss and health changes) [
9
], there is less information available
on the psychological impact of the pandemic on romantic couples. There is a particular
dearth of studies using dyadic methods where the perspectives of both couple members
are included.
1.1. Romantic Partnerships and Conflict in COVID
Romantic relationships are typically the largest source of social interaction and one of
the most influential social relationships [
10
,
11
]. Indeed, romantic relationship satisfaction
is a robust predictor of overall life satisfaction [
12
]. Research has begun to identify the
impacts of the pandemic on romantic relationship functioning including effects on couples’
well-being and relationship satisfaction. For instance, married individuals experienced a
steeper decline in well-being than unmarried people during the COVID-19 outbreak [13].
The COVID-19 pandemic and associated stressors (e.g., mandatory homeschooling)
have increased ‘couples’ conflict’ [
14
]. This dyadic construct includes a variety of negative
communication behaviors that arise out of disagreement or incompatibility including in-
terruptions, contempt, rejection, condemnation, and communication avoidance [
15
]. At
the extremes of conflict, reports of intimate partner violence roughly doubled in frequency
and severity during the pandemic [
16
]. Moreover, couples who reported more COVID-19-
related stress also reported more conflict and lower relationship satisfaction [
17
]. Increases
in romantic conflict resulted in decreased intimate sexual behavior within the partner-
ship during COVID-19 [
18
], further undermining relationship quality. The importance
of examining romantic conflict during the COVID-19 pandemic is further underscored
by the known deleterious effects of romantic conflict on life and relationship satisfaction
over time [19–21].
1.2. Romantic Partnership Conflict and Drinking
A systematic review found the effects of drinking on marriages are overwhelmingly
negative [
22
]. Indeed, heavy alcohol use and alcohol-related problems co-occur with marital
conflict [
22
]. Other research suggests the relation between drinking and marital satisfaction
is more nuanced. First, there are reported positive effects of matched substance use behavior
in romantic couples, regardless of the quantity of use [
23
,
24
]. Additionally, a bidirectional
relation exists between conflict and drinking in couples. Not only does drinking before a
conflict result in more aggressive conflicts [
25
], but some research suggests conflict results
in more alcohol use or alcohol-related problems [
11
,
26
]. For example, a daily diary study
found that romantic conflict predicted increases in next-day drinking behavior [
24
]. This
study also showed that the tendency to drink in response to earlier romantic conflict was
stronger in women than in men [24].
Cooper’s drinking motives theory [
27
,
28
], describes how individuals may be moti-
vated to drink to achieve different desired outcomes. These different motives are outlined
in terms of the type of reinforcement desired (i.e., positive or negative reinforcement) and
the source of reinforcement (i.e., internal or external to the individual). Crossing these two
dimensions of desired outcomes yields four distinct drinking motives: social (positive-
external motives; to increase social affiliation), enhancement (positive-internal motives; to
heighten pleasurable sensations or affect), conformity (negative-external motives; to allevi-
ate/avoid social rejection), and coping (negative-internal motives; to reduce undesirable
internal states) [
27
]. Of these four motives, coping motives appear the riskiest or among the
riskiest [
29
]; they consistently predict heavy drinking and alcohol-related problems both
cross-sectionally [30,31] and longitudinally [28,32,33].

Int. J. Environ. Res. Public Health 2023,20, 6332 3 of 19
Coping motives provide one possible explanation for drinking behavior following
conflict. Drinking to cope with interpersonal conflict (including romantic conflict) explains
36% of the variance in coping drinking motives [
34
]. This supports a coping-motivated
link in the couples’ conflict-to-drinking relation. Additionally, an individual’s drinking
motives (including their coping motives) may be influenced by the drinking motives of
close others. This has been theorized to occur via social learning: while one cannot directly
observe the motivations of others, an individual can witness their partner being reinforced
or punished, which may lead them to drink to achieve similar outcomes [
35
]. While it is
well established that romantic partners influence one another’s drinking behavior [
36
], it
has only recently been established in pre-pandemic research that an individual’s drinking
motives can influence their partner’s drinking behavior. Specifically, one partner’s coping
motives were longitudinally associated with the other partner’s drinking behavior as
mediated through the other partner’s coping motives [
37
]. This points to the importance of
examining partner effects in this field.
Another pre-pandemic study from Lambe et al. examined the links between couples’
conflict and alcohol outcomes in a sample of 100 emerging adult couples [
38
]. They
examined the individual’s own coping motives and the partner’s drinking motives as
possible mediators of the conflict—drinking outcomes link [
38
]. In indistinguishable
dyad analyses (i.e., not distinguished by gender), this study showed that the individual’s
coping-with-depression motives mediated the relation between dyadic conflict (combined
reporting by both partners) and the individual’s alcohol-related problems [
38
]. Within
the mixed-gender couples, this mediation held true for women but not men. This pattern
was consistent with prior gendered findings on conflict-related drinking [
24
]. Both the
indistinguishable dyads and distinguishable dyads (by gender) analyses found actor effects
but no partner effects. Thus, no evidence was found that the partner’s coping motives
influenced the individual’s alcohol-related problems [38].
While there were important insights offered by Lambe et al.’s study [
38
], there were
also limitations. First, this study recruited emerging adults, which may limit generalizability
to longer-term romantic relationships since partner drinking influence has been shown
to vary by stage of a relationship, increasing with marriage [
23
]. Further, with only
100 couples, their study may have been underpowered to detect smaller magnitude partner
effects [
39
]. Additionally, Lambe et al. [
38
] examined alcohol-related problems as their
outcome. Increased drinking levels are the more proximal outcome of drinking to cope,
which may lead to alcohol-related problems over time [
30
]. While examining alcohol-related
problems may have been an appropriate outcome given Lambe et al.’s [
38
] longitudinal
design, it remains important to test their model with drinking behavior as the theoretically
more proximal outcome, particularly given the adverse public health consequences of
increased heavy drinking during the pandemic [5,40].
1.3. The Current Study
Given the increases in both couples’ conflict and drinking during the COVID-19
pandemic [
4
,
5
,
17
], it is important to test the influence of the drinking motives of both
individuals in a couple in response to dyadic conflict on both couple members’ heavier
drinking behavior. In other words, it is important to test both actor and partner effects. The
current study was a conceptual replication of Lambe et al. [
38
] and an extension to a larger,
older, community-recruited sample and to drinking behavior as the outcome, conducted in
the context of stay-at-home orders during the COVID-19 pandemic. We utilized a cross-
sectional design and the Actor–Partner Interdependence Model (APIM; [
41
]). We reasoned
that results could help shed light on the consequences of the pandemic, along with the
associated public health viral containment measures put in place in the spring of 2020, on
drinking behavior in couples.
We hypothesized we would detect partner effects despite their absence in Lambe et al. [
38
]
due to our larger sample size (partner effects tend to be smaller, e.g., [
37
]) and due to the
probable greater influence of a cohabitating partner on an individual’s behavior during the

Int. J. Environ. Res. Public Health 2023,20, 6332 4 of 19
stresses and isolation of lockdown. Additionally, the various unique stressors associated
with lockdown (e.g., health, financial, homeschooling, and loss of leisure activities) could
have increased conflict (e.g., negative partner behaviors and verbal aggression) and the
influence of conflict on members of a couple. We reasoned that the increase in time
spent with one’s romantic partner and corresponding decrease in time spent with other
individuals outside the home may amplify the influence of a romantic partner on one’s
own behavior. We specifically hypothesized that:
H1:
Dyadic conflict would have an indirect effect on increasing drinking behavior through one’s own
coping motives (actor effects). This would hold true for both indistinguishable and distinguishable
(by gender) dyads analyses [38].
H2:
While both were hypothesized to be significant, we expected that mediation effects for the actor
effects would be significantly larger in women than the same mediation for men in the distinguishable
(by gender) dyads analysis [24,38].
H3:
Dyadic conflict would have an indirect effect on increasing drinking through one’s partner’s
coping motives (partner effects; [
37
]) in both indistinguishable and distinguishable dyads analyses.
2. Materials and Methods
2.1. Participants
Participants were 758 romantic cohabitating couples in Canada (N= 1516 individuals).
To meet study eligibility, Qualtrics panelists must have been living in Canada throughout
April 2020, been at least 19 years of age, been involved in a romantic relationship with
a partner who is also at least 19 years of age, cohabited with that same romantic partner
throughout April 2020, reported that they and their partner were following COVID-19-
related stay-at-home advisories in their local jurisdiction in April 2020, and confirmed that
their romantic partner was also willing and available to participate. Essential workers were
excluded as we wished to examine couples staying at home together during lockdown.
Additional descriptive statistics can be found in four other published papers from this
dataset [
14
,
42
–
44
]. This study was approved by a university research ethics board (#2020-5166).
Only couples in which both partners reported drinking on at least one occasion in
April 2020 were included in the present analyses (n= 348 couples). Non-drinkers (couples
where one or both members reported consuming alcohol zero times in the reporting period;
n= 410 couples) were excluded due to the zero-inflated distribution created by these cases for
the outcome variables and due to the absence of coping motives for non-drinkers (i.e., drinking
motives are only queried of drinkers). Demographics of the final sample are provided in
Table 1. Descriptive statistics for key variables can be found in Tables 2and 3for the total
sample and subsamples by gender, respectively. Notably, the current sample contains a
higher proportion of “binge drinkers” (i.e., heavy episodic drinkers) in both the men and the
women than pre-pandemic Canadian norms [
45
]. Additionally, men showed higher rates
of binge drinking than women in our sample for both peak and typical drinking occasions,
consistent with gender differences in binge drinking rates seen prior to the pandemic [
45
].
Chi-square analyses were conducted to compare those included in the analyses vs. those
excluded for having a non-drinker as one or both members of a couple. Notably, the
subsample used in the present analyses were more educated, reported higher income, and
more full-time employment, than those excluded from analyses for not consuming alcohol
during the period assessed. Additionally, there were significant differences in ethnicity:
those excluded from analyses for being non-drinkers were disproportionately East Asian
and South Asian. Chi-square analyses yielded no significant differences in relationship
composition, relationship status, if living with children or not, other ethnic categories,
gender, age, or relationship length between those that consumed alcohol (and thus were
included in the present analyses) vs. those that were excluded from the analyses. Lastly,
couples included in the analyses did not report significantly more romantic conflict, coping

Int. J. Environ. Res. Public Health 2023,20, 6332 5 of 19
motives, peak alcohol consumption, or typical alcohol consumption quantity than those
excluded; however, individuals in couples that were included in the analyses (in which
both partners reported drinking) reported drinking at a significantly greater frequency than
drinkers whose partner did not drink and thus were excluded from analyses (see Table 2).
Table 1. Categorical Demographic Characteristics and Comparisons with Excluded Participants.
Sample Characteristics n(%) χ2p
Gender 1.04 0.793
Women 340 (48.9)
Men 355 (51.0)
Other 1 (0.1)
Ethnicity 37.36 0.001
White 563 (80.9)
East Asian 49 (7.0)
South Asian 24 (3.4)
Arab/West Asian 6 (0.9)
Black 11 (1.6)
Latin American 3 (0.4)
Indigenous 1 (0.1)
Multiracial 19 (2.7)
Other 14 (2.0)
Unknown 5 (0.7)
Education level 14.35 0.045 * †
Elementary or some high school 40 (4.9)
High school graduate 144 (17.5)
Some college/university 130 (15.8)
University graduate 371 (45.0)
Some post-graduate 31 (3.8)
Post-graduate degree 106 (12.9)
Prefer not to answer 2 (0.2)
Income 54.24 0.001 * †
$25,000 or less per year 9 (1.3)
Between $26,000 and $50,000 62 (8.9)
Between $51,000 and $75,000 108 (15.5)
Between $76,000 and $100,000 156 (22.4)
Between $101,000 and $125,000 114 (16.4)
Between $126,000 and $150,000 72 (10.3)
$151,000 or more per year 124 (17.8)
Prefer not to answer 51 (7.3)
Employment status 15.46 0.001 * †
Unemployed 292 (42.0)
Student 5 (0.7)
Employed part-time 83 (11.9)
Employed full-time 268 (38.5)
Relationship status 0.12 0.734
Married or common law 688 (98.9)
Non-married 8 (1.1)
Living with children 3.05 0.081
Yes 170 (24.4)
Relationship composition 3.01 0.222
Mixed-gender 658 (94.5)
Same-gender 34 (4.9)
Unknown or other 4 (0.6)
Demographic characteristics of the sample were compared with excluded couples (i.e., those where one or both
partners did not consume alcohol in April 2020). * p< 0.05;
†
= indicates the given characteristic is significantly
higher/larger in the sample used in the present analyses than in excluded couples.

Int. J. Environ. Res. Public Health 2023,20, 6332 6 of 19
Table 2. Continuous Sample Characteristics and Comparisons with Excluded Participants.
Sample Characteristics M SD Mdn Range t p
Age 54.55 13.72 57.00 25–87 −0.01 0.990
Relationship length 27.10 14.50 25.20 1.1–62.8 0.01 0.993
Conflict 16.76 12.70 12.00 7–63 −1.59 0.112
Coping motives 2.16 2.88 0.68 0–10 −1.75 0.081
Drinking frequency 12.37 9.85 8.00 1–30 −6.40 <0.001
Peak quantity 4.63 5.57 3.00 1–25.5 −1.12 0.262
Typical quantity 2.79 3.64 2.00 1–25.5 −1.85 0.064
Independent samples t-tests were used to compare drinkers (included in main analyses) vs. non-drinkers (couples
where one or both members did not drink in April 2020 and were excluded from main analyses) on continuous
sample characteristics. A significant and negative t-value indicates the given variable was higher/larger in the
sample used in the present analyses.
Table 3. Alcohol Use by Gender.
Men Women
Sample Characteristics M SD Binge
Drinking (%) M SD Binge
Drinking (%)
Drinking frequency 13.66 10.02 n/a 11.03 9.42 n/a
Peak quantity 5.27 5.91 33.3 3.97 5.13 29.8
Typical quantity 3.15 3.77 16.9 2.43 3.46 13.5
Drinking frequency = number of drinking occasions in April 2020; peak quantity = number of drinks consumed
on heaviest drinking occasion in April 2020; typical quantity = typical number of drinks consumed per drinking
occasion in April 2020. Binge drinking was defined as drinking 5 or more drinks in a single occasion for men and 4 or
more drinks in a single occasion for women in April 2020 [
45
]. n/a = not applicable since binge drinking prevalence
can only be calculated from quantity variables. Pre-pandemic, Canadian adults (age 25+) engaged in binge drinking
(peak quantity; single occasion in a given month) at rates of 27.1% for men and 19.6% for women [45].
2.2. Measures
2.2.1. Demographics
Demographic and relationship variables were assessed such as gender, sex, age, racial
identity, employment status, educational attainment, relationship status, relationship
length, as well as adherence to stay-at-home advisories, essential worker status, and
COVID-19
infection [
42
,
43
]. Age was controlled in the main analyses. Income was con-
trolled in supplemental sensitivity analyses.
2.2.2. Dyadic Conflict
Conflict was assessed using the 7-item Partner-Specific Rejecting Behaviors Scale [
46
].
Participants were asked to report on their behavior towards their partner (e.g., “I was angry
or irritated with my partner”, “I snapped or yelled at my partner”) on a Likert scale from
1-strongly disagree to 9-strongly agree, in which higher sum scores (possible scale range 7–63)
indicated more severe conflict behavior. Consistent with previous research [
38
,
47
], scores
from each member were averaged into a dyadic conflict score for analysis. This score repre-
sented equal contributions from each partner and captured the fact that couples’ conflict is
an inherently dyadic construct. Notably, this scale measures negative conflict behaviors
(i.e., not just the presence of disagreement, but rather behaviors such as ignoring, insulting,
and criticizing) but does not measure intimate partner violence or physical aggression. This
scale has shown strong psychometric properties, i.e., internal consistency [
47
]. Additionally,
each individual’s reports on their own conflict behavior toward their partner were highly
associated within couples, r= 0.78, p< 0.001, providing psychometric support for the
averaging of the two scores into a single dyadic conflict measure. Internal consistency for
the present sample can be found in Table 4.

Int. J. Environ. Res. Public Health 2023,20, 6332 7 of 19
Table 4. Correlation Matrix.
12345
1. Conflict -- 0.50 ** 0.08 * 0.11 ** 0.20 **
2. Drinking to cope -- 0.17 ** 0.15 ** 0.28 **
3. Drinking frequency -- 0.28 ** 0.21 **
4. Peak quantity -- 0.49 **
5. Typical quantity --
Internal consistency (α) 0.95 0.91 - - -
** p< 0.01, * p< 0.05. Internal consistency in column 2 treats drinking to cope with anxiety and drinking to cope
with depression as two items of the single measure of drinking to cope. Drinking frequency = number of drinking
occasions in April 2020; peak quantity = number of drinks consumed on heaviest drinking occasion in April 2020;
typical quantity = typical number of drinks consumed per drinking occasion in April 2020.
2.2.3. Drinking to Cope
Drinking motives were measured using the two coping items from the Brief Alcohol
Motive Measure (BAMM; [
48
]). Participants were asked to report how frequently during
the month of April 2020 they drank for various reasons. Responses were made on a set
of sliding scale visual analogue scales (VAS) anchored from never to always. Each VAS
item was scored on a continuum from 0–100, divided by 10 for analyses for consistency
with Bartel et al. [
48
]. Each item represented a distinct motive for drinking. This scale has
been validated against the original multi-item Modified Drinking Motives Questionnaire-
Revised [
49
] and has been shown to have similar levels of concurrent and predictive validity
to the original measure in relation to drinking levels and alcohol-related problems [
48
]. The
drinking to cope with anxiety and drinking to cope with depression items were averaged
into a single coping motives scale [
42
] given their high correlation in the present sample
(r= 0.82, p< 0.001) and to increase reliability of measurement (Table 1).
2.2.4. Drinking Behavior
The Quantity/Frequency/Peak Alcohol Use Index [
50
] was used to assess drinking
behavior during the month of April 2020. Analyses used a composite of three measures:
drinking frequency, i.e., how many days of consuming alcoholic beverages in the month,
with 11 ordinal forced choice options ranging from “once a month” to “every day”, which
were translated to days of drinking per month (e.g., “4 times per week” was coded as “16”);
typical quantity, i.e., how many drinks were typically consumed on days with alcohol
consumption (with options ranging from 0–25+; in which “25+” was coded as 25.5); and
peak quantity, i.e., how many drinks were consumed on the single instance of heaviest
consumption (with options ranging from 0–25+, with responses of “25+” coded as 25.5).
Frequency was used to identify individuals who were non-drinkers during April 2020.
2.3. Procedure
Qualtrics Panels, a survey management service that recruits from a large pool of
potential participants, was used to gather the current data based on researcher-specified
criteria. A total of 3292 couples began the survey and were assessed on the eligibility criteria.
Participants who were eligible and provided informed consent were asked questions about
a variety of constructs relevant to the pandemic including their demographics, relationship
conflict, coping motives for alcohol use, and drinking behavior. Their partner completed
the same questions in relation to their own demographics, conflict behavior, and drinking
motives and behavior. All items were completed in early July 2020, and retrospectively
queried for the month of April 2020, during which all Canadian provinces were under
stay-at-home advisories under Public Health Emergency acts [51].
Three mechanisms were in place to screen out invalid responses: a question assessing
commitment to giving honest answers while completing the survey, filter questions for
which participants were instructed to select a specific answer to determine attention and
reading of instructions, and a speeder check to ensure adequate time was spent completing
the survey. Of those recruited to participate, couples were excluded for the following

Int. J. Environ. Res. Public Health 2023,20, 6332 8 of 19
reasons for at least one partner: partner unwilling to complete the survey or partial
completion (n= 1996), failed attention/speeder check (n= 183), or not meeting specific
eligibility criteria (n= 355). These exclusions yielded a sample of 758 couples. The current
study also required that both members of the couple consumed alcohol at least once in
April 2020, yielding our final sample (N= 348 couples). Participants were compensated
through Qualtrics according to their guidelines.
3. Results
3.1. Analysis Plan
Given the non-normal distribution and count nature of the drinking variables, we
elected to use generalized structural equation models specifying a negative binomial
distribution for the outcome variable (drinking) due to the non-normal distribution of
the data. The models were run using listwise deletion. No cases were removed from the
main analyses due to missing data. Drinking was operationalized as a latent variable
comprising drinking frequency, typical quantity, and peak quantity. APIM was used to
estimate actor and partner effects of the mediator (coping) on the outcome (drinking) [
39
].
Romantic conflict was operationalized as a between-dyads variable as it was the composite
of both partners’ reports as in Lambe et al. [
38
]. Data and syntax are available from the
corresponding authors on request.
Evidence for mediation is established when the 95% confidence interval from bias-
corrected bootstrapping with 1000 resamples does not contain the value 1. This is due to
the confidence interval being estimated around the incidence rate ratio which surrounds
1 instead of 0. To be inclusive of all participants regardless of sexual orientation, we
first tested the components of the path model and indirect effects of romantic conflict on
drinking through coping motives in the entire sample in indistinguishable dyad analyses
(i.e., where partners are not distinguished by gender). Then, using the mixed-gender
couples only, we evaluated whether each path and subsequent indirect effects, differed by
gender (i.e., whether each path was different for men and women) in distinguishable dyads
analyses, with gender as the distinguishing feature.
Specifying a negative binomial distribution for the outcome variable results in the
estimates representing log-linked coefficients. One way to return the estimates into an
interpretable format is through their exponentiated form: incidence rate ratios (IRRs). IRRs
surround the number one and represent a percent higher or lower in the outcome as a
function of a one-unit increase in the predictor. For example, an IRR of 1.22 for the effect of
conflict on drinking means that every one-unit more in conflict reported corresponds to
22% more drinking. Similarly, an IRR of 0.95 indicates that every one-unit more in conflict
reported corresponds to 5% less drinking. See Tables 5and 6for IRRs for the indistinguish-
able (non-gendered) and distinguishable (gendered) APIM models, respectively.
Table 5. Direct and Indirect Effects of Conflict on Drinking in Indistinguishable Dyads.
Effect Path b SE p IRR LCI UCI
Direct Conflict →A Drinking 0.08 0.03 0.010 1.08 1.02 1.16
Indirect Conflict →A Coping →A Drinking 0.11 0.02 <0.001 1.12 1.08 1.16
Indirect Conflict →P Coping →A Drinking −0.01 0.02 0.740 0.99 0.96 1.03
Total Conflict →A Drinking 0.20 0.02 <0.001 1.22 1.15 1.27
IRR = Incident Rate Ratio. LCI and HCI are lower and upper 95% confidence intervals from bias-corrected
bootstrapping with 1000 resamples. A = actor; P = partner.
Demographic information was probed for significant associations with drinking behav-
ior (outcome variable). Age and income were used as covariates as they had the strongest
relationships with drinking behavior. Other variables that were associated with drinking
behavior in our sample (e.g., relationship length, having children in the home or not, em-
ployment status, and education level) were associated less strongly with drinking behavior
and very highly correlated with each other (e.g., age and relationship length; income and

Int. J. Environ. Res. Public Health 2023,20, 6332 9 of 19
employment status). Thus, age and income were selected as covariates for parsimony in
possible confounding variables. While age was covaried in the main analysis, income was
added as a covariate in a sensitivity analysis due to the power lost when including income,
as 6% of our sample opted to not report their income.
Table 6. Direct and Indirect Effects of Conflict on Men and Women Partners’ Coping and Drinking.
Effect Path b SE p IRR LCI UCI
Direct Conflict →M Drinking 0.06 0.03 0.092 1.06 0.99 1.13
Indirect Conflict →M Coping →M Drinking 0.07 0.02 0.002 1.07 1.03 1.13
Indirect Conflict →W Coping →M Drinking 0.05 0.03 0.070 1.05 1.00 1.09
Total Conflict →M Drinking 0.18 0.03 <0.001 1.20 1.14 1.26
Direct Conflict →W Drinking 0.06 0.04 0.108 1.06 0.99 1.15
Indirect
Conflict
→
W Coping
→
W Drinking
0.17 0.03 <0.001 1.19 1.12 1.26
Indirect Conflict →M Coping →W Drinking −0.06 0.03 0.033 0.94 0.89 1.00
Total Conflict →W Drinking 0.17 0.02 <0.001 1.19 1.12 1.27
IRR = Incident Rate Ratio. LCI and HCI are lower and upper 95% confidence intervals from bias-corrected
bootstrapping with 1000 resamples. M = men; W = women.
3.2. Descriptive Statistics
Descriptive statistics are found in Tables 1–3and bivariate correlations are in Table 4.
All outcomes were positively and significantly correlated with each other.
3.3. Indistinguishable Dyadic Analysis
Conventional fit indices are not calculated for generalized structural equation mod-
elling; however, the significance of paths within the model suggested a strong fit. Stan-
dardized path coefficients are presented in Figure 1and indirect effects and IRR values in
Table 5.
Int. J. Environ. Res. Public Health 2023, 20, x FOR PEER REVIEW 10 of 20
Within the latent variable of drinking, frequency was constrained to load at b = 1.00,
while peak quantity loaded at b = 0.62, 95%CIs [0.58, 0.66], p < 0.001 and typical quantity
significantly loaded at b = 0.43, 95%CIs [0.39, 0.48], p < 0.001. Age was a significant
covariate, b = 0.03, 95%CIs [0.03, 0.03], SE = 0.001, p < 0.001.
Dyadic conflict significantly predicted greater drinking to cope motives, b = 0.90,
95%CIs [0.77, 1.04], SE = 0.07, p < 0.001. A significant positive actor effect for coping
motives to drinking behavior was observed, b = 0.12, 95%CIs [0.09, 0.16], SE = 0.02, p <
0.001. No partner effect for coping motives to drinking behavior was observed, b = −0.01,
95%CIs [−0.04, 0.03], SE = −0.33, p = 0.741. Additionally, dyadic conflict directly predicted
greater drinking behavior even after accounting for coping motives, b = 0.08, 95%CIs [0.02,
0.15], SE = 0.03, p = 0.010. Both partners’ drinking behaviors were significantly and
positively associated with each other, b = 0.55, 95%CIs [0.43, 0.67], SE = 0.06, p < 0.001
(Figure 1).
The direct, indirect, and total effects are in Table 5. The mediation model was
significant for actors: one’s own coping motives positively mediated the relation between
dyadic conflict and one’s own drinking behavior. The partner’s coping motives did not
significantly mediate the relation between dyadic conflict and the individual’s drinking
behavior. In other words, actor effects were detected but partner effects were not. Overall,
the model resulted in an IRR value of 1.22 (Table 5).
Figure 1. Generalized SEM Model with Indistinguishable Dyads. Values indicate standardized b
coefficients. *** p < 0.001, ** p < 0.01.
3.4. Distinguishable Dyads Analysis
The significance of paths within the generalized structural equation model suggested
a strong fit. Standardized path coefficients are in Figure 2 and indirect effects and IRR
values in Table 6.
Figure 1.
Generalized SEM Model with Indistinguishable Dyads. Values indicate standardized
bcoefficients. *** p< 0.001, ** p< 0.01.
Within the latent variable of drinking, frequency was constrained to load at b= 1.00,
while peak quantity loaded at b= 0.62, 95%CIs [0.58, 0.66], p< 0.001 and typical quantity
significantly loaded at b= 0.43, 95%CIs [0.39, 0.48], p< 0.001. Age was a significant covariate,
b= 0.03, 95%CIs [0.03, 0.03], SE = 0.001, p< 0.001.
Dyadic conflict significantly predicted greater drinking to cope motives, b= 0.90,
95%CIs [0.77, 1.04], SE = 0.07, p< 0.001. A significant positive actor effect for coping
motives to drinking behavior was observed, b= 0.12, 95%CIs [0.09, 0.16], SE = 0.02, p< 0.001.
No partner effect for coping motives to drinking behavior was observed, b=
−
0.01, 95%CIs
[
−
0.04, 0.03], SE =
−
0.33, p= 0.741. Additionally, dyadic conflict directly predicted greater
drinking behavior even after accounting for coping motives, b= 0.08, 95%CIs [0.02, 0.15],

Int. J. Environ. Res. Public Health 2023,20, 6332 10 of 19
SE = 0.03, p= 0.010. Both partners’ drinking behaviors were significantly and positively
associated with each other, b= 0.55, 95%CIs [0.43, 0.67], SE = 0.06, p< 0.001 (Figure 1).
The direct, indirect, and total effects are in Table 5. The mediation model was sig-
nificant for actors: one’s own coping motives positively mediated the relation between
dyadic conflict and one’s own drinking behavior. The partner’s coping motives did not
significantly mediate the relation between dyadic conflict and the individual’s drinking
behavior. In other words, actor effects were detected but partner effects were not. Overall,
the model resulted in an IRR value of 1.22 (Table 5).
3.4. Distinguishable Dyads Analysis
The significance of paths within the generalized structural equation model suggested
a strong fit. Standardized path coefficients are in Figure 2and indirect effects and IRR
values in Table 6.
Int. J. Environ. Res. Public Health 2023, 20, x FOR PEER REVIEW 10 of 20
Within the latent variable of drinking, frequency was constrained to load at b = 1.00,
while peak quantity loaded at b = 0.62, 95%CIs [0.58, 0.66], p < 0.001 and typical quantity
significantly loaded at b = 0.43, 95%CIs [0.39, 0.48], p < 0.001. Age was a significant
covariate, b = 0.03, 95%CIs [0.03, 0.03], SE = 0.001, p < 0.001.
Dyadic conflict significantly predicted greater drinking to cope motives, b = 0.90,
95%CIs [0.77, 1.04], SE = 0.07, p < 0.001. A significant positive actor effect for coping
motives to drinking behavior was observed, b = 0.12, 95%CIs [0.09, 0.16], SE = 0.02, p <
0.001. No partner effect for coping motives to drinking behavior was observed, b = −0.01,
95%CIs [−0.04, 0.03], SE = −0.33, p = 0.741. Additionally, dyadic conflict directly predicted
greater drinking behavior even after accounting for coping motives, b = 0.08, 95%CIs [0.02,
0.15], SE = 0.03, p = 0.010. Both partners’ drinking behaviors were significantly and
positively associated with each other, b = 0.55, 95%CIs [0.43, 0.67], SE = 0.06, p < 0.001
(Figure 1).
The direct, indirect, and total effects are in Table 5. The mediation model was
significant for actors: one’s own coping motives positively mediated the relation between
dyadic conflict and one’s own drinking behavior. The partner’s coping motives did not
significantly mediate the relation between dyadic conflict and the individual’s drinking
behavior. In other words, actor effects were detected but partner effects were not. Overall,
the model resulted in an IRR value of 1.22 (Table 5).
Figure 1. Generalized SEM Model with Indistinguishable Dyads. Values indicate standardized b
coefficients. *** p < 0.001, ** p < 0.01.
3.4. Distinguishable Dyads Analysis
The significance of paths within the generalized structural equation model suggested
a strong fit. Standardized path coefficients are in Figure 2 and indirect effects and IRR
values in Table 6.
Figure 2.
Generalized SEM Model with Distinguishable Dyads. Values indicate standardized
bcoefficients. *** p< 0.001, ** p< 0.01, * p< 0.05 *, † p< 0.10.
Within the latent variable of drinking, drinking frequency was constrained to load at
b= 1.00 for both genders. For men, peak quantity loaded at b= 0.65, 95%CIs [0.60, 0.69],
SE = 0.02, p< 0.001, and typical quantity significantly loaded at b= 0.44, 95%CIs [0.39, 0.49],
SE = 0.02, p< 0.001. For women, peak quantity significantly loaded at b= 0.56, 95%CIs
[0.51, 0.61], SE = 0.03, p< 0.001, and typical quantity significantly loaded at b= 0.35, 95%CIs
[0.29, 0.41], SE = 0.03, p< 0.001. Age was a significant covariate for both men, b= 0.04,
95%CIs [0.03, 0.04], SE = 0.001, p< 0.001, and women, b= 0.03, 95%CIs [0.03, 0.03], SE = 0.002,
p< 0.001.
Dyadic conflict significantly predicted greater drinking to cope motives for both men,
b= 0.98, 95%CIs [0.81, 1.13], SE = 0.08, p< 0.001, and women, b= 0.87, 95%CIs [0.69, 1.06],
SE = 0.09, p< 0.001. One’s own coping motives predicted higher levels of one’s own
drinking behavior for both men, b= 0.08, 95%CIs [0.03, 0.12], SE = 0.02, p= 0.002, and
women, b= 0.19, 95%CIs [0.14, 0.25], SE = 0.03, p< 0.001 (actor effects). Notably, the
partner paths of coping motives on drinking (i.e., partner’s coping motives predicting own
drinking behavior) were significant in predicting women’s drinking b=
−
0.06, 95%CIs
[
−
0.12,
−
0.002], SE = 0.03, p= 0.041, and marginal for men’s drinking, b= 0.05, 95%CIs
[
−
0.002, 0.11], SE = 0.03, p= 0.062. These associations were opposite in direction in men
and women: women partners’ coping motives marginally predicted greater drinking by
men, while men partners’ coping motives significantly predicted less drinking by women.
Dyadic conflict did not significantly directly predict drinking behavior for men, b= 0.06,
95%CIs [
−
0.01, 0.12], SE = 0.03, p= 0.092, or women, b= 0.06, 95%CIs [
−
0.01, 0.14], SE = 0.04,
p= 0.098, after accounting for coping motives.
The direct, indirect, and total effects are in Table 6for men and women. Women’s own
coping motives significantly mediated the relation between dyadic conflict and their own
drinking behavior but the mediation pathway for their partner’s drinking behavior was

Int. J. Environ. Res. Public Health 2023,20, 6332 11 of 19
only marginally significant. In other words, higher conflict was positively associated with
women’s coping motives which in turn predicted higher rates of women’s drinking and
marginally more drinking in their male partners. Men’s coping motives significantly medi-
ated the relation between dyadic conflict and their own drinking behavior but significantly
mediated the relation between dyadic conflict and their partner’s drinking behavior in the
inverse of the expected direction. In other words, higher conflict was associated with more
men’s coping motives which in turn predicted higher rates of men’s drinking but lower
drinking in their women partners.
Significant partner effects were found for women’s drinking and only marginal part-
ner effects were found for men’s drinking. However, the difference between these two
effects themselves was not significant as the 95% confidence intervals for the mediation
pathways overlapped.
Gender differences in actor effects were established by comparing 95% confidence
intervals for the mediation pathway of dyadic conflict and own drinking behavior through
one’s own coping motives for women versus men (Table 6). The direction of the difference
was consistent with H2, in which we hypothesized larger effect sizes predicting women’s
drinking than men’s drinking. However, the 95% confidence intervals for IRR values
(i.e., exponentiated slope value) for the mediation pathways did overlap for women and
men. Thus, while both pathways were independently statistically significant, the effect was
not significantly larger for women than for men.
3.5. Sensitivity Analysis
The above analyses were also run with income as a covariate. Notably, n= 24 couples
opted to not report income, resulting in missing data, and those that did not report income
were significantly different than those who did in several important ways. Those who did
not report income, on average, reported drinking fewer beverages per drinking occasion,
reported significantly less drinking to cope with anxiety and drinking to cope with depres-
sion motives, and were significantly older. Thus, controlling income may have introduced
systemic error into the results, as well as slightly lowered the power of the analyses.
The sensitivity analyses largely replicated the main analyses. For the distinguishable
dyad analyses, the significant direct effect from conflict to drinking behavior when con-
trolling coping motives became marginal when controlling income (b= 0.06, SE = 0.03,
p= 0.063). For the indistinguishable dyad analyses, one component of the mediation
pathway from conflict to women’s drinking via men’s coping motives (i.e., from men’s
coping motives to women’s drinking) became marginal when controlling income, b=
−
0.06,
SE = 0.03, p= 0.054. All other effects remained as reported in the main analyses (see
Supplementary Materials).
4. Discussion
Our study examined the role of coping motives as a mediator between dyadic conflict
and drinking behavior in cohabitating romantic couples during the COVID-19 lockdown
while controlling for age. Overall, both distinguishable and indistinguishable dyadic analy-
ses showed that coping motives mediated the association between conflict and drinking
behavior, such that each one unit increase in conflict would predict a 26–27% increase in
drinking behavior—a small but meaningful effect. The effect sizes found in these analy-
ses are consistent with recent arguments that most psychological phenomena of research
interest are complex, in that the causes for such phenomena are highly multifactorial in
nature and each individual cause exerts small effects. Therefore, studying and reporting
small effects contributes to a cumulative science [52].
4.1. Actor Effects
Consistent with H1 and pre-pandemic results by Lambe et al. [
38
], romantic conflict
was associated with greater reports of one’s own drinking behavior, mediated through
their own higher coping motives. We replicated this finding in both indistinguishable

Int. J. Environ. Res. Public Health 2023,20, 6332 12 of 19
dyad analyses with the full sample and in distinguishable (by gender) analyses with
mixed-gender couples. This mediation effect would be hypothesized by drinking motives
theory [
28
], providing evidence that individuals in cohabitating romantic relationships
may increase their drinking to cope with negative affect (depression and anxiety) following
couples’ conflict. This mediation suggests that coping motives could be a mechanism
by which conflict may lead to more drinking and possibly to later alcohol problems. It
is important to note that coping-related drinking is a particularly risky drinking motive
and is more strongly associated than other motives with alcohol-related problems [
30
,
53
].
Lastly, this finding may partially explain increases in drinking during the COVID-19
pandemic [
4
,
5
] as increases in couple’s conflict have been reported during lockdowns [
17
]
and in relation to the stressors faced by couples during lockdown [14].
Additionally, conflict had a direct effect on drinking behavior that was not accounted
for by (mediated through) coping drinking motives in the indistinguishable dyad analyses.
This suggests there are other mechanisms not included in the model that led members
of a couple to greater drinking behavior following conflict (e.g., drinking as a shared
activity to repair conflict [
24
]; coping with other forms of negative affect such as anger or
shame [
54
–
56
]). However, it is important to note that this direct effect was neither found in
the distinguishable (by gender) analyses nor when controlling the indistinguishable dyad
analyses for income (see Supplemental Materials). This suggests that it may be a spurious
finding—at least partially explained by variance in income—and that coping motives may
indeed fully mediate the association between romantic conflict and drinking.
Actor effects (ways in which one’s own coping motives predicted one’s own drinking
behavior) were significant for both genders. Contrary to H2, the mediation of conflict on
drinking behavior by coping motives was not significantly stronger in women than men.
While past research indicates women are more likely to modulate their drinking behavior
in response to conflict [
24
,
38
], this was not supported by our analyses. However, some
findings suggest these effects may be more likely for the quantity of alcohol consumed,
rather than the frequency of drinking occasions [
7
,
24
,
57
]. The current analyses combined
measures of alcohol use into a single latent construct, which may have obscured gender
differences in the quantity of alcohol consumed in response to distress. Separating the
various indices of alcohol use behavior in future studies may yield more consistent findings
on gender differences in alcohol use in response to romantic conflict.
Overall, these findings replicate previous findings that coping motives mediated the
association between romantic conflict and alcohol-related problems and extend the results
into a sample across adulthood, particularly with middle and late adulthood represented.
However, it is worth noting the ways in which the older age of our sample may impact
the variables assessed. There are inconsistent findings on whether verbal aggression
decreases with age [
58
,
59
], with more robust findings that physical aggression and intimate
partner violence decrease with age [
60
]. Additionally, there are several studies that indicate
that mechanisms of conflict (e.g., personality predictors) and consequences of conflict
(e.g., impact on health, depressive symptoms) are consistent across the lifespan [
61
–
64
].
Age is significantly correlated with relationship length and thus both are captured by
covarying for age in these analyses; it is possible that aggression decreases with age and/or
that conflict stabilizes over decades of partnership. Nonetheless, while rates of alcohol use
and intimate partner violence decrease with age, these remain significant public health
problems in older adults [65].
4.2. Partner Effects
Our third hypothesis [H3] predicting significant and positive partner effects was
not supported. Men reported marginally more drinking behavior associated with their
partner’s coping motives but this effect did not reach statistical significance. The inverse
pattern was true for women, in that women reported significantly less drinking associated
with their partner’s coping motives. However, these two effects were not statistically
significantly different from each other in magnitude and both likely reflected small effects.

Int. J. Environ. Res. Public Health 2023,20, 6332 13 of 19
Effect sizes are typically smaller for partner effects than actor effects in dyadic analyses.
Additionally, we did not find observable partner effects in the indistinguishable analyses.
We hypothesized an adverse partner effect for both men and women, in which one
partner’s coping motives would predict the other’s drinking behavior, but the protective
partner crossover effect observed for women’s drinking behavior was not anticipated.
While there is robust evidence of partner influence on drinking behavior [
36
,
66
], there is
emerging evidence that one’s own drinking behavior can be influenced by the drinking mo-
tives of important network members. This research has primarily focused on peers [
67
–
69
],
but a handful of studies have examined and supported this effect in romantic couples [
37
].
However, the existing literature and the partner influence hypothesis would predict a con-
sistent positive association between a partner’s coping drinking motives and an individual’s
drinking behavior regardless of gender [
66
]. In contrast, our data show higher coping mo-
tives in men were associated with less women’s drinking behavior in mixed-gender couples.
‘Emotion work’ may be one explanation for this unanticipated effect. Emotion work is a
broad concept that encompasses activities that enhance a partner’s emotional well-being
and provide emotional support, which may also explain gender inequities in relationships
in other domains [
70
]. While emotion work enhances marriage satisfaction [
71
], degree
and inequity of emotion work are negatively associated with well-being, particularly for
individuals married to a man [
72
,
73
]. Research and gender role theory converge to suggest
that women in mixed-gender relationships tend to be more concerned about and in-tune
with emotion work [74].
Thus, it is possible that married/cohabiting women abstain from increased drinking
in response to the coping-motivated drinking exhibited by their men partners following
romantic conflict to reserve internal resources for emotion work. Indeed, a similar effect
has been noted with depressive symptoms during the pandemic: women with depressive
symptoms are more likely to be concerned about the impact of their symptoms on their
spouse and to perform emotion work to modulate that impact compared to men with
depressive symptoms [
75
]. A similar compensatory effect has been reported regarding
drinking, where men’s distress was associated with a decreased likelihood of binge drinking
behavior in their women partners [
76
]. It is important to note that this pattern of decreased
consumption in response to men’s drinking to cope may be protective for women’s drinking
while simultaneously being risky for intimate partner violence. Indeed, one study found
that the highest rates of intimate partner violence occurred in couples in which the husband
was a heavy-drinker and the wife was not [
77
]. Research is required to directly test the
proposed emotion work mechanism, to probe specific motivations by women to drink less
in response to their husband exhibiting more coping-motivated drinking, and to examine
the beneficial and harmful consequences of this pattern.
Contrary to H3, no significant partner-coping mediation between dyadic conflict and
drinking behavior was found in either the indistinguishable dyadic analyses with the full
sample or for women’s coping predicting men’s drinking in the distinguishable dyadic
analyses. Nonetheless, the lower limit of the confidence interval for the partner effect
on men’s drinking included but did not go past zero (or below one for the IRR values:
i.e., a marginal effect). This finding begets further investigation, as it is possible that a
larger sample or more at-risk sample may yield the hypothesized significant partner effects
on men’s drinking. The absence of partner effects in the distinguishable dyads analysis
may have been due to the gendered partner effects being of similar magnitude but in
opposite directions. This may have resulted in the washing out of overall partner effects
when collapsing across gender. This possible masking underscores the importance of our
following up analyses in the full sample with gendered analyses.
These findings on partner effects also have theoretical implications. Both the partner
influence hypothesis and most extant evidence suggests partner influence occurs via one’s
own drinking motives (i.e., the partner’s motives influence one’s own motives, which
then proximally predict one’s own drinking) and that influence occurs in the direction of
risk (e.g., [
37
]). However, these results provide preliminary evidence of partner effects

Int. J. Environ. Res. Public Health 2023,20, 6332 14 of 19
being protective. The possible, though currently untested, mechanism of emotion work
may add to the existing theory. Additionally, while the current study did not probe the
mechanism of individual motives influencing partner motives directly, the marginal effect
for men’s drinking would be consistent with the partner influence hypothesis. If replicated,
and if tested in a serial mediation model (i.e., conflict
→
partner A coping motives
→
partner B coping motives
→
partner B alcohol use), it would further strengthen the extant
literature. Furthermore, most of the literature in this area samples emerging adult couples
or youth; the present study provides preliminary evidence of drinking motives influencing
the behavior of others in middle adulthood. It is possible that partner influence decreases
in strength with age or life stage. Additionally, our larger sample may have been more
adequately powered to detect partner effects than the Lambe et al. [
38
] study, though
possibly not large enough to detect significant partner effects for men’s drinking, as partner
effects overall tend to be small after controlling for actor effects [
39
]. It is also important
to consider how COVID-19 lockdown conditions may have influenced partner effects.
Lockdown conditions may have led to more partner conflict to begin with [
17
] and/or
created substantially more opportunity for partner influence, particularly in the relative
absence of other social influences [
42
]. Ultimately, more research is required to determine
the reliability, size, and direction of partner effects in drinking behavior via drinking
motives, across samples and contexts.
4.3. Limitations
There are several limitations of note. Conclusions on temporality and causality are
limited by our cross-sectional design. However, our results largely replicate the Lambe et al.
data, which were longitudinal [
38
]. Additional longitudinal design studies are necessary
to strengthen the temporal associations. Lab-based couple studies of elicited romantic
conflict [
78
] would allow for the examination of causal effects. The literature would also
benefit from daily diary studies to assess the associations between these constructs at a
micro level, i.e., drinking the day of or the day after a romantic conflict, assessing the
intensity of conflict at the event level, and assessing motives for specific drinking episodes.
Research has often distinguished between coping with anxiety and coping with depres-
sion drinking motives as seen in the five-factor model of drinking motives [
49
], including
in Lambe et al. [
38
]. Future studies could benefit from using multi-item measures of drink-
ing to cope [
49
]. The BAMM was selected for this study to minimize participant burden
given the other constructs of interest in the larger study, particularly during an already
stressful time. But this choice did not allow for the separation of the unique contributions
of coping with anxiety and coping with depression motives to the mediation effect ob-
served in the current study. Moreover, our distinguishable dyad analyses only examined
mixed-gender couples, excluding gender- and sexuality-diverse relationships. This narrows
the generalizability of the gender moderation results to mixed-gendered couples. This
limitation is imposed by the nature of the statistical techniques available to date within the
APIM framework.
Given numerous known influences on substance use behavior, such as personality
(e.g., impulsivity [
79
], adverse childhood experiences [
80
,
81
], genetic risk [
82
]), it is impor-
tant to note that this study only examines one possible mechanism influencing alcohol use.
Despite well-established distal risk factors, proximal factors (i.e., drinking motives) are often
more robust predictors of substance behavior and may be more suitable for intervention [
83
].
Lastly, the current study conceptualized conflict as a dyadic variable and relied on
self-reports. While this is consistent with most studies [
47
] and allowed for a conceptual
replication of Lambe et al. [
38
], it limited exploration of possible effects of directionality of
conflict behaviors (e.g., one partner ignoring, getting angry with, embarrassing the other),
or divergent perceptions of conflict behavior, on subsequent coping-motivated drinking be-
havior. Self-reported conflict behaviors may result in biases [
84
] in under-reporting socially
undesirable behavior such as the perpetration of abusive conflict behaviors [
85
]. However,
there is evidence of conflict predicting subsequent drinking for both directions of conflict

Int. J. Environ. Res. Public Health 2023,20, 6332 15 of 19
behavior (i.e., perpetration and receipt of conflict). Moreover, some research suggests one’s
own perception of conflict behavior may be the stronger predictor of drinking [
86
]. Another
notable limitation is that these data relied on retrospective self-reporting. In addition to the
possibility of unsystematic error in reporting, retrospective recall may also be susceptible
to recall biases [
87
], particularly for socially undesirable behaviors. However, the novelty
of lockdowns in April 2020 may have enhanced individuals’ memories of that time period,
as novelty reliably enhances memory of behaviors occurring contiguously with, before, or
after a novel event [88].
5. Conclusions
There are several implications of these results for research, theory, and practice. Re-
search found increased drinking during the pandemic, which was consistent with the binge
drinking rates in the current sample [
6
]. Our findings suggest couples’ conflict and coping
motives are important factors related to greater drinking behavior during the pandemic
for both men and women. These results contribute to the existing evidence to encourage
more mental health and social supports being allocated for couples cohabitating during the
pandemic or who are otherwise socially isolated. This should include increased supports
for those experiencing the extremes of a couple’s conflict (i.e., intimate partner violence),
which has been shown to be elevated in prevalence during the pandemic [
16
]. Additionally,
these data provide support for the allocation of resources to at-risk couples, given that
higher rates of both men and women in our sample were observed to be engaged in risky
binge drinking during the pandemic compared to pre-pandemic adult norms [
45
]. Recent
research has evidenced the effectiveness in reducing conflict intensity of brief cognitive
reappraisal interventions for couples following conflict during the pandemic [
89
]; prior
research has shown this intervention’s effects on reducing alcohol use [
90
]. Lastly, the
presence of the protective partner effect on women’s drinking behavior via men’s drinking
motives has implications: undertreated men’s distress may have downstream consequences
for the well-being of their women partners who may compensate to provide emotional
support. These results underscore the need to consider the relationship context in individ-
ual or couples’ psychological interventions for risky drinking, in addition to examining
individual motives for drinking. Overall, this study adds to the growing evidence of the
social influence of drinking motives on other individuals’ drinking behavior, including
possible protective effects on drinking behavior.
Supplementary Materials:
The following supporting information can be downloaded at: https://
www.mdpi.com/article/10.3390/ijerph20146332/s1, Figure S1: Generalized SEM Model with Indis-
tinguishable Dyads Controlling for Income; Figure S2: Generalized SEM Model with Distinguishable
Dyads Controlling for Income; Table S1: Direct and Indirect Effects of Conflict on Drinking in Indis-
tinguishable Dyads Controlling for Income; Table S2: Direct and Indirect Effects of Conflict on Men
and Women Partners’ Coping and Drinking Controlling for Income.
Author Contributions:
Conceptualization, L.L., A.E.F.H. and S.H.S.; methodology, L.M.R., R.N.-A.,
S.B.S., S.H.D., S.M., A.A., L.L., A.E.F.H. and S.H.S.; software, L.M.R., C.N. and S.H.S.; validation, C.N.,
L.M.R. and A.E.F.H.; formal analysis, C.N., L.M.R. and A.E.F.H.; investigation, L.M.R. and S.H.S.;
resources, S.H.S.; data curation, A.E.F.H. and S.H.S.; writing—original draft preparation, A.E.F.H.;
writing—review and editing, A.E.F.H. and S.H.S.; visualization, A.E.F.H.; supervision, S.H.S.; project
administration, S.H.S. and A.E.F.H.; funding acquisition, L.M.R., R.N.-A., S.B.S., S.H.D., S.M., A.A.
and S.H.S. All authors have read and agreed to the published version of the manuscript.
Funding:
This research was funded by the Social Sciences and Humanities Research Council of
Canada Partnership Engage Grant (1008-2020-0220; PI: S.H.S); the Dalhousie Psychiatry Research
Fund (P.I.: S.H.S.); the Social Sciences and Humanities Research Council of Canada, Canada Graduate
Scholarship-Doctoral (to A.E.F.H). S.H.S. is supported through a Tier 1 Canada Research Chair in
Addictions and Mental Health. L.L. was supported by a Canadian Institute of Health Research
postdoctoral fellowship.

Int. J. Environ. Res. Public Health 2023,20, 6332 16 of 19
Institutional Review Board Statement:
The study was conducted in accordance with the Declaration
of Helsinki and approved by the Institutional Review Board (“Research Ethics Board”) of Dalhousie
University (#2020-5166).
Informed Consent Statement:
Informed consent was obtained from all participants involved in
the study.
Data Availability Statement:
The data presented in this study are available on reasonable request
from the corresponding author.
Acknowledgments:
The authors would like to acknowledge the contributions of Pamela Collins for
administrative support, as well as Sean MacKinnon for consultation in adapting the statistical syntax.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
World Health Organization. Coronavirus Disease 2019 (COVID-19): Situation Report, 79; World Health Organization: Geneva,
Switzerland, 2020.
2.
Rosenberg, M.; Luetke, M.C.; Hensel, D.; Kianersi, S.; Fu, T.-C.; Herbenick, D. Depression and loneliness during COVID-19
restrictions in the United States, and their associations with frequency of social and sexual connections. Soc. Psychiatry Psychiatr.
Epidemiol. 2021,56, 1221–1232. [CrossRef] [PubMed]
3. Vindegaard, N.; Benros, M.E. COVID-19 pandemic and mental health consequences: Systematic review of the current evidence.
Brain Behav. Immun. 2020,89, 531–542. [CrossRef] [PubMed]
4.
Kim, J.U.; Majid, A.; Judge, R.; Crook, P.; Nathwani, R.; Selvapatt, N.; Lovendoski, J.; Manousou, P.; Thursz, M.; Dhar, A.; et al.
Effect of COVID-19 lockdown on alcohol consumption in patients with pre-existing alcohol use disorder. Lancet Gastroenterol.
Hepatol. 2020,5, 886–887. [CrossRef] [PubMed]
5.
Pollard, M.S.; Tucker, J.S.; Green, H.D., Jr. Changes in adult alcohol use and consequences during the COVID-19 pandemic in the
US. JAMA Netw. Open 2020,3, e2022942. [CrossRef]
6.
Schmits, E.; Glowacz, F. Changes in alcohol use during the COVID-19 pandemic: Impact of the lockdown conditions and mental
health factors. Int. J. Ment. Health Addict. 2021,20, 1147–1158. [CrossRef]
7.
Rodriguez, L.M.; Litt, D.M.; Stewart, S.H. Drinking to cope with the pandemic: The unique associations of COVID-19-related perceived
threat and psychological distress to drinking behaviors in American men and women. Addict. Behav. 2020,110, 106532. [CrossRef]
8.
Rossow, I.; Mäkelä, P. Public health thinking around alcohol-related harm: Why does per capita consumption matter? J. Stud.
Alcohol Drugs 2021,82, 9–17. [CrossRef]
9.
Howe, G.W.; Levy, M.L.; Caplan, R.D. Job loss and depressive symptoms in couples: Common stressors, stress transmission, or
relationship disruption? J. Fam. Psychol. 2004,18, 639–650. [CrossRef]
10.
Umberson, D.; Crosnoe, R.; Reczek, C. Social relationships and health behavior across the life course. Annu. Rev. Sociol.
2010
,
36, 139–157. [CrossRef]
11.
Whisman, M.A.; Sheldon, C.T.; Goering, P. Psychiatric disorders and dissatisfaction with social relationships: Does type of
relationship matter? J. Abnorm. Psychol. 2000,109, 803–808. [CrossRef]
12.
Roberson, P.N.E.; Norona, J.C.; Lenger, K.A.; Olmstead, S.B. How do relationship stability and quality affect wellbeing?: Romantic
relationship trajectories, depressive symptoms, and life satisfaction across 30 years. J. Child Fam. Stud. 2018,27, 2171–2184. [CrossRef]
13.
Yang, H.; Ma, J. How an epidemic outbreak impacts happiness: Factors that worsen (vs. protect) emotional well-being during the
coronavirus pandemic. Psychiatry Res. 2020,289, 113045. [CrossRef] [PubMed]
14.
Basso, N.; Lambe, L.J.; Kim, A.J.; Rodriquez, L.; Deacon, H.; Nogueria-Arjona, R.; Abbass, A.; Stewart, S.H. Effects of mandatory
homeschooling during COVID-19 on conflict in romantic couples. J. Fam. Psychol. 2023,37, 275–281. [CrossRef] [PubMed]
15.
Oetzel, J.G.; Ting-Toomey, S. The SAGE Handbook of Conflict Communication: Integrating Theory, Research, and Practice; Sage
Publications: Thousand Oaks, CA, USA, 2006.
16.
Gosangi, B.; Park, H.; Thomas, R.; Gujrathi, R.; Bay, C.P.; Raja, A.S.; Seltzer, S.E.; Balcom, M.C.; McDonald, M.L.; Orgill, D.P.; et al.
Exacerbation of physical intimate partner violence during COVID-19 pandemic. Radiology 2021,298, E38–E45. [CrossRef] [PubMed]
17.
Balzarini, R.N.; Muise, A.; Zoppolat, G.; Di Bartolomeo, A.; Rodrigues, D.L.; Alonso-Ferres, M.; Urganci, B.; Debrot, A.;
Pichayayothin, N.B.; Dharma, C.; et al. Love in the time of COVID: Perceived partner responsiveness buffers people from lower
relationship quality associated with COVID-related stressors. Soc. Psychol. Personal. Sci. 2023,14, 342–355. [CrossRef]
18.
Luetke, M.; Hensel, D.; Herbenick, D.; Rosenberg, M. Romantic relationship conflict due to the COVID-19 pandemic and changes in
intimate and sexual behaviors in a nationally representative sample of American adults. J. Sex Marital. Ther.
2020
,46, 747–762. [CrossRef]
19. Booth, A.; Crouter, A.C.; Clements, M. Couples in Conflict; Routledge: Milton Park, UK, 2001.
20. Cramer, D. Relationship satisfaction and conflict style in romantic relationships. J. Psychol. 2000,134, 337–341. [CrossRef]
21.
Karney, B.R.; Bradbury, T.N. The longitudinal course of marital quality and stability: A review of theory, method, and research.
Psychol. Bull. 1995,118, 3–34. [CrossRef]
22. Marshal, M.P. For better or for worse? The effects of alcohol use on marital functioning. Clin. Psychol. Rev. 2003,23, 959–997. [CrossRef]

Int. J. Environ. Res. Public Health 2023,20, 6332 17 of 19
23.
Leonard, K.E.; Homish, G.G. Predictors of heavy drinking and drinking problems over the first four years of marriage. Psychol.
Addict. Behav. 2008,22, 25–35. [CrossRef]
24.
Levitt, A.; Cooper, M.L. Daily alcohol use and romantic relationship functioning: Evidence of bidirectional, gender-, and
context-specific effects. Personal. Soc. Psychol. Bull. 2010,36, 1706–1722. [CrossRef]
25.
Stets, J.E.; Henderson, D.A. Contextual factors surrounding conflict resolution while dating: Results from a national study. Natl.
Counc. Fam. Relat. 1991,40, 29–36. [CrossRef]
26.
Maisto, A.; O’Farrel, T.O.; Connors, J.; McKay, R.; Pelcovits, M. Alcoholics’ attributions of factors affecting their relapse to drinking
and reasons for terminating relapse episodes. Addict. Behav. 1988,13, 79–82. [CrossRef]
27.
Cooper, M.L. Motivations for alcohol use among adolescents: Development and validation of a four-factor model. Psychol. Assess.
1994,6, 117–128. [CrossRef]
28.
Cooper, M.L.; Kuntsche, E.; Levitt, A.; Barber, L.L.; Wolf, S. Motivational Models of Substance Use: A Review of Theory and
Research on Motives for Using Alcohol, Marijuana, and Tobacco. In The Oxford Handbook of Substance Use and Substance Use
Disorders; Kenneth, J.S., Ed.; Oxford University Press: Oxford, UK, 2016; pp. 375–421.
29.
Bresin, K.; Mekawi, Y. The “why” of drinking matters: A meta-analysis of the association between drinking motives and drinking
outcomes. Alcohol. Clin. Exp. Res. 2021,45, 38–50. [CrossRef] [PubMed]
30.
Cooper, M.L.; Frone, M.R.; Russell, M.; Mudar, P. Drinking to regulate positive and negative emotions: A motivational model of
alcohol use. J. Personal. Soc. Psychol. 1995,69, 990–1005. [CrossRef] [PubMed]
31.
Park, C.L.; Levenson, M.R. Drinking to cope among college students: Prevalence, problems, and coping processes. J. Stud. Alcohol
2002,63, 486–497. [CrossRef] [PubMed]
32.
Mackinnon, S.P.; Kehayes, I.-L.; Clark, R.; Sherry, S.B.; Stewart, S.H. Testing the four-factor model of personality vulnerability to
alcohol misuse: A three-wave, one-year longitudinal study. Psychol. Addict. Behav. 2014,28, 1000–1012. [CrossRef]
33.
Stevenson, B.L.; Dvorak, R.D.; Kramer, M.P.; Peterson, R.S.; Dunn, M.E.; Leary, A.V.; Pinto, D. Within- and between-person
associations from mood to alcohol consequences: The mediating role of enhancement and coping drinking motives. J. Abnorm.
Psychol. 2019,128, 813–822. [CrossRef] [PubMed]
34.
Carrigan, G.; Samoluk, S.B.; Stewart, S.H. Examination of the short form of the Inventory of Drinking Situations (IDS-42) in a
young adult university student sample. Behav. Res. Ther. 1998,36, 789–807. [CrossRef]
35.
Bandura, A. Vicarious and Self-Reinforcement Processes. In The Nature of Reinforcement; Robert, G., Ed.; Academic Press:
New York, NY, USA, 1971.
36.
Muyingo, L.; Smith, M.M.; Sherry, S.B.; McEachern, E.; Leonard, K.E.; Stewart, S.H. Relationship on the rocks: A meta-analysis of
romantic partner effects on alcohol use. Psychol. Addict. Behav. 2020,34, 629–640. [CrossRef] [PubMed]
37.
Kehayes, I.-L.; Mackinnon, S.P.; Sherry, S.B.; Leonard, K.E.; Stewart, S.H. Drinking motives and drinking behaviors in romantic
couples: A longitudinal actor-partner interdependence model. Psychol. Addict. Behav. 2019,33, 208–220. [CrossRef] [PubMed]
38.
Lambe, L.; Mackinnon, S.P.; Stewart, S.H. Dyadic conflict, drinking to cope, and alcohol-related problems: A psychometric study
and longitudinal actor-partner interdependence model. J. Fam. Psychol. 2015,29, 697–707. [CrossRef] [PubMed]
39. Kenny, D.A.; Kashy, D.A.; Cook, W.L. Dyadic Data Analysis; Guilford Press: New York, NY, USA, 2020.
40.
Sohal, A.; Khalid, S.; Green, V.; Gulati, A.; Roytman, M. The pandemic within the pandemic: Unprecedented rise in alcohol-related
hepatitis during the COVID-19 pandemic. J. Clin. Gastroenterol. 2022,56, e171–e175. [CrossRef]
41.
Kenny, D.A.; Ledermann, T. Detecting, measuring, and testing dyadic patterns in the actor–partner interdependence model.
J. Fam. Issues 2010,24, 359–366. [CrossRef]
42.
Deacon, S.H.; Rodriguez, L.M.; Elgendi, M.; King, F.E.; Nogueira-Arjona, R.; Sherry, S.B.; Stewart, S.H. Parenting through a
pandemic: Mental health and substance use consequences of mandated homeschooling. Couple Fam. Psychol. Res. Pract.
2021
,
10, 281–293. [CrossRef]
43.
DesRoches, D.I.; Deacon, S.H.; Rodriguez, L.M.; Sherry, S.B.; Nogueira-Arjona, R.; Elgendi, M.M.; Meier, S.; Abbass, A.; King, F.E.;
Stewart, S.H. Homeschooling during COVID-19: Gender differences in work–family conflict and alcohol use behaviour among
romantic couples. Soc. Sci. 2021,10, 240. [CrossRef]
44.
Kim, A.J.; Smith, M.M.; Sherry, S.B.; Rodriguez, L.M.; Meier, S.M.; Nogueira-Arjona, R.; Deacon, S.H.; Abbass, A.; Stewart, S.H.
Depressive symptoms and conflict behaviors: A test of the stress generation hypothesis in romantic couples during the COVID-19
pandemic. J. Soc. Clin. Psychol. 2022,41, 517–540. [CrossRef]
45.
Canadian Centre on Substance Use and Addiction. Canadian Drug Summary: Alcohol; Canadian Centre on Substance use and
Addiction (CCSA): Ottawa, ON, Canada, 2019.
46.
Murray, S.L.; Rose, P.; Griffin, D.W.; Bellavia, G.M. Calibrating the sociometer: The relational contingencies of self-esteem.
J. Personal. Soc. Psychol. 2003,85, 63–84. [CrossRef]
47.
Mackinnon, S.P.; Sherry, S.B.; Antony, M.M.; Stewart, S.H.; Sherry, D.L.; Hartling, N. Caught in a bad romance: Perfectionism,
conflict, and depression in romantic relationships. J. Fam. Psychol. 2012,26, 215–225. [CrossRef]
48.
Bartel, S.; Sherry, S.; Mahu, I.; Stewart, S. Development of brief alcohol and cannabis motives measures: Psychometric evaluation
using expert feedback and longitudinal methods. Cannabis 2023,6, 34–49. [CrossRef] [PubMed]
49.
Grant, V.V.; Stewart, S.H.; O’Connor, R.M.; Blackwell, E.; Conrod, P.J. Psychometric evaluation of the five-factor Modified
Drinking Motives Questionnaire-Revised in undergraduates. Addict. Behav. 2007,32, 2611–2632. [CrossRef] [PubMed]

Int. J. Environ. Res. Public Health 2023,20, 6332 18 of 19
50.
Dimeff, L.A. Brief Alcohol Screening and Intervention for College Students (BASICS): A Harm Reduction Approach; Guilford Press:
New York, NY, USA, 1999.
51.
A Timeline of Events in Canada’s Fight against COVID-19; The Canadian Press: Toronto, ON, Canada, 2020; Available online:
https://www.thestar.com/news/canada/2020/06/18/a-timeline-of-events-in-canadas-fight-against-covid-19.html (accessed on
24 March 2021).
52.
Götz, F.M.; Gosling, S.D.; Rentfrow, P.J. Small effects: The indispensable foundation for a cumulative psychological science.
Perspect. Psychol. Sci. 2022,17, 205–215. [CrossRef]
53.
Kuntsche, E.; Knibbe, R.; Gmel, G.; Engels, R. Why do young people drink? A review of drinking motives. Clin. Psychol. Rev.
2005,25, 841–861. [CrossRef]
54.
Bilevicius, E.; Single, A.; Bristow, L.A.; Foot, M.; Ellery, M.; Keough, M.T.; Johnson, E.A. Shame mediates the relationship between
depression and addictive behaviours. Addict. Behav. 2018,82, 94–100. [CrossRef]
55.
Kelly, J.F.; Stout, R.L.; Tonigan, J.S.; Magill, M.; Al, K.E.T. Negative affect, relapse, and alcoholics anonymous (AA): Does AA
work by reducing anger? J. Stud. Alcohol Drugs 2010,71, 434–444. [CrossRef] [PubMed]
56.
Rabinovitz, S. Drowning your sorrows? Social exclusion and anger effects on alcohol drinking. Addict. Res. Theory
2014
,
22, 363–370. [CrossRef]
57.
Rodriguez, L.M.; Gius, B.K.; Derrick, J.L.; Leonard, K.E. A dyadic approach to attachment anxiety and avoidance, marital conflict,
and drinking. J. Soc. Pers. Relatsh. 2020,37, 2386–2408. [CrossRef]
58. Bookwala, J.; Jacobs, J. Age, marital process, and depressed affect. Gerontologist 2004,44, 328–338. [CrossRef]
59.
Straus, M.A.; Sweet, S. Verbal/symbolic aggression in couples: Incidence rates and relationship to personal characteristics.
J. Marriage Fam. 1992,54, 346–357. [CrossRef]
60.
Mouton, C.; Rodabough, R.; Rovi, S.; Hunt, J.; Talamantes, M.; Brzyski, R.; Burge, S. Prevelance and 3-year incidence of abuse
amongst postmenopausal women. Res. Pract. 2004,94, 605–612.
61.
Iveniuk, J.; Waite, L.J.; McClintock, M.K.; Teidt, A.D. Marital Conflict in Older Couples: Positivity, Personality, and Health.
J. Marriage Fam. 2014,76, 130–144. [CrossRef] [PubMed]
62.
Schreiber, E.; Georgia Salivar, E. Using a vulnerability-stress-adaptation framework to model intimate partner violence risk
factors in late life: A systematic review. Aggress. Violent Behav. 2021,57, 101493. [CrossRef]
63.
South, S.C.; Boudreaux, M.J.; Oltmanns, T.F. Personality disorders and intimate partner aggression: A replication and extension
in older, married couples. Pers. Disord. 2021,12, 70–80. [CrossRef] [PubMed]
64.
Whisman, M.A.; Uebelacker, L.A. Prospective associations between marital discord and depressive symptoms in middle-aged
and older adults. Psychol. Aging 2009,24, 184–189. [CrossRef]
65. Sorocco, K.H.; Ferrell, S.W. Alcohol use among older adults. J. Gen. Psychol. 2006,133, 453–467. [CrossRef]
66.
Mushquash, A.R.; Stewart, S.H.; Sherry, S.B.; Mackinnon, S.P.; Antony, M.M.; Sherry, D.L. Heavy episodic drinking among dating
partners: A longitudinal actor-partner interdependence model. Psychol. Addict. Behav. 2013,27, 178–183. [CrossRef]
67.
Hussong, A. Social influences in motivated drinking among college students. Psychol. Addict. Behav.
2003
,17, 142–150. [CrossRef]
68.
Kuntsche, E.; Stewart, S.H. Why my classmates drink: Drinking motives of classroom peers as predictors of individual drinking
motives and alcohol use in adolescence – a mediational model. J. Health Psychol. 2009,14, 536–546. [CrossRef]
69.
Stewart, S.H.; Castellanos-Ryan, N.; Vitaro, F.; Conrod, P.J. Why my friends drink is why I drink. Alcohol. Clin. Exp. Res.
2014
,
38, 187a.
70. Erickson, R.J. Why emotion work matters: Sex, gender, and the division of household labor. J. Marriage Fam. 2005,67, 337–351. [CrossRef]
71.
Holm, K.E.; Werner-Wilson, R.J.; Cook, A.S.; Berger, P.S. The association between emotion work balance and relationship
satisfaction of couples seeking therapy. Am. J. Fam. Ther. 2001,29, 193–205. [CrossRef]
72.
Duncombe, J.; Marseden, D. Love and intimacy: The gender division of emotion and ‘emotion work’. Sociology
1993
,27, 221–241. [CrossRef]
73.
Umberson, D.; Thomeer, M.B.; Pollitt, A.M.; Mernitz, S.E. The psychological toll of emotion work in same-sex and different-sex
marital dyads. J. Marriage Fam. 2020,82, 1141–1158. [CrossRef]
74.
Elliott, S.; Umberson, D. The performance of desire: Gender and sexual negotiations in long-term marriages. J. Marriage Fam.
2008,70, 391–406. [CrossRef]
75.
Calarco, J.M.; Meanwell, E.; Anderson, E.; Knopf, A. “My husband thinks I’m crazy”: COVID-19-related conflict in couples with
young children. SocArXiv 2020. [CrossRef]
76.
Holway, G.V.; Umberson, D.; Thomeer, M.B. Binge drinking and depression: The influence of romantic partners in young
adulthood. Soc. Ment. Health 2017,7, 36–49. [CrossRef]
77.
Quigley, B.M.; Leonard, K.E. Alcohol and the continuation of early marital aggression. Alcohol. Clin. Exp. Res.
2000
,24, 1003–1010. [CrossRef]
78.
Denson, T.F.; Spanovic, M.; Miller, N. Cognitive appraisals and emotions predict cortisol and immune responses: A meta-analysis
of acute laboratory social stressors and emotion inductions. Psychol. Bull. 2009,135, 823–853. [CrossRef]
79.
Coskunpinar, A.; Dir, A.L.; Cyders, M.A. Multidimensionality in impulsivity and alcohol use: A meta-analysis using the UPPS
model of impulsivity. Alcohol. Clin. Exp. Res. 2013,37, 1441–1450. [CrossRef]
80.
Anda, R.F.; Whitfield, C.L.; Felitte, V.J.; Chapman, D.; Edwards, V.J.; Dube, S.R.; Williamson, D.F. Adverse childhood experiences,
alcoholic parents, and later risk of alcoholism and depression. Psychiatr. Serv. 2002,53, 1001–1009. [CrossRef]
81.
Lee, R.D.; Chen, J. Adverse childhood experiences, mental health, and excessive alcohol use: Examination of race/ethnicity and
sex differences. Child Abus. Negl. 2017,69, 40–48. [CrossRef] [PubMed]

Int. J. Environ. Res. Public Health 2023,20, 6332 19 of 19
82.
Kendler, K.S.; Myers, J.; Dick, D.; Prescott, C.A. The relationship between genetic influences on alcohol dependence and on
patterns of alcohol consumption. Alcohol. Clin. Exp. Res. 2010,34, 1058–1065. [CrossRef] [PubMed]
83.
Canale, N.; Vieno, A.; Santinello, M.; Chieco, F.; Andriolo, S. The efficacy of computerized alcohol intervention tailored to drinking
motives among college students: A quasi-experimental pilot study. Am. J. Drug Alcohol Abus. 2015,41, 183–187. [CrossRef] [PubMed]
84.
Mathie, N.L.; Wakeling, H.C. Assessing socially desirable responding and its impact on self-report measures among sexual
offenders. Psychol. Crime Law 2011,17, 215–237. [CrossRef]
85.
Dutton, D.G.; Hemphill, K.J. Patterns of socially desirable responding among perpetrators and victims of wife assault. Violence
Vict. 1992,7, 29–39. [CrossRef]
86.
Derrick, J.L.; Testa, M. Temporal effects of perpetrating or receiving intimate partner aggression on alcohol consumption: A daily
diary study of community couples. J. Stud. Alcohol Drugs 2009,78, 213–221. [CrossRef]
87.
Patrick, M.E.; Lee, C.M. Comparing numbers of drinks: College students’ reports from retrospective summary, followback, and
prospective daily diary measures. J. Stud. Alcohol Drugs 2010,71, 554–561. [CrossRef]
88.
Lorents, A.; Ruitenberg, M.F.L.; Schomaker, J. Novelty-induced memory boosts in humans: The when and how. Heliyon
2023
,
9, E14410. [CrossRef]
89.
Rodriguez, L.M.; Stewart, S.H.; Neighbors, C. Effects of a brief web-based interpersonal conflict cognitive reappraisal expressive-
writing intervention on changes in romantic conflict during the COVID-19 quarantine. Couple Fam. Psychol. Res. Pract.
2021
,
10, 212–222. [CrossRef]
90.
Rodriguez, L.M.; Dell, J.B.; Lee, K.D.M.; Onufrak, J. Effects of a brief cognitive reappraisal intervention on reductions in alcohol
consumption and related problems. Psychol. Addict. Behav. 2019,33, 637–643. [CrossRef] [PubMed]
Disclaimer/Publisher’s Note:
The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.