Alcohol 82 (2020) 91e101

Contents lists available at ScienceDirect

Alcohol
journal homepage: http://www.alcoholjournal.org/

Consequences of alcohol use, and its association with psychological

distress, sensitivity to emotional contagion and age of onset of alcohol
use, in Uruguayan youth with or without college degree
Paul Ruiz a, **, Angelina Pilatti b, c, Ricardo Marcos Pautassi b, d, *
a
Centro de Investigacio
n Clínica, Facultad de Psicología, UdelaR, Uruguay
b
Universidad Nacional de Co
rdoba (UNC), Facultad de Psicología, Argentina
c
gicas, IIPSI-CONICET-UNC, Co
Instituto de Investigaciones Psicolo
rdoba, Argentina
d
Instituto de Investigaciones M
edicas Mercedes y Martín Ferreyra, INIMEC-CONICET-UNC, Argentina

a r t i c l e i n f o a b s t r a c t

Article history: Psychological distress can promote alcohol consumption during emerging adulthood. Still unknown is,
Received 8 November 2018 however, how predisposition to emotional contagion alters psychological distress, and how these phe-
Received in revised form nomena are affected by level of education. The present study analyzed the effect of psychological distress,
29 April 2019
age of first contact with alcohol (early, late), and predisposition to emotional contagion on alcohol-
Accepted 5 September 2019
induced negative consequences and on the volume of alcohol consumed during the last year. We also
described alcohol-use behaviors as a function of sex, maximum level of education and age of first contact
Keywords:
with alcohol, in 1505 youth from Uruguay (18e30 years). A survey measured alcohol use (Alcohol Use
Psychological distress
Alcohol use
Disorders Identification Test and ad-hoc questionnaire), negative consequences of alcohol use [young
Uruguayan youth adult alcohol consequences questionnaire (YAACQ)], psychological distress (Kessler scale) and proclivity
Emotional contagion to emotional contagion (Doherty Emotional contagion scale). The patterns of alcohol use were greater in
men vs. women and in those featuring an early age of first alcohol use, yet similar in college and non-
college graduates. Early drinkers had greater levels of psychological distress than late-onset drinkers.
There was a significant bivariate and multiple correlation between psychological distress and the number
of negative consequences of alcohol experienced during the last year, which remained significant even
after controlling for total volume of alcohol consumed. Significant associations emerged between YAACQ
scores and frequency of heavy episodic or binge drinking, and between psychological distress and
emotional contagion, but not between emotional contagion and any of the remaining variables. Psy-
chological distress was not significantly correlated with heavy episodic or binge drinking. The study
indicates that, during adolescence and youth, psychological distress is associated with experiencing
negative consequences of alcohol consumption. The study also suggested that greater levels of psy-
chological distress may underlie the facilitating effect of an early age of drinking onset upon alcohol
drinking patterns.
© 2019 Elsevier Inc. All rights reserved.

Introduction

Depression and anxiety disorders usually arise during adoles-

cence (McGorry, 2011) and exhibit comorbidity with alcohol use

n Me
dica M. y M. Ferreyra
disorders (AUD) (Chamberlain & Sahakian, 2004; Gomez-Restrepo
* Corresponding author. Instituto de Investigacio
(INIMEC e CONICET), Friuli 2434, Co
rdoba, C.P 5016, Argentina. Fax: þ54-351- et al., 2004). Subramaniam et al. (2017) reported a 20% prevalence
4695163. of problematic alcohol use in patients with depression, much

n Clínica, Facultad de Psicología,
** Corresponding author. Centro de Investigacio higher than in those with other pathologies (e.g., 6.4% in schizo-
UdelaR, Mercedes 1737 (Anexo), Depto. Montevideo, Montevideo, 11200, Uruguay. phrenia). The "tension-reduction hypothesis" (Conger, 1956; Young,
Fax: þ598-2401-1631.
Oei, & Knight, 1990)] suggests that alcohol intake mitigates feelings
E-mail addresses: [email protected] (P. Ruiz), [email protected]
(R.M. Pautassi). of anxiety or depression (adverse psychological states that we will

https://doi.org/10.1016/j.alcohol.2019.09.001
0741-8329/© 2019 Elsevier Inc. All rights reserved.
92 P. Ruiz et al. / Alcohol 82 (2020) 91e101

hereinafter refer to as psychological distress) and that this pro- Sprott, 1998). Others suggested that the educational system fos-
motes subsequent alcohol intake. Consistent with this, alcohol- ters peer interaction, and therefore makes individuals more sen-
induced amelioration of anxiety or depressive states is a signifi- sitive to the effects of social norms that stimulate consumption
cant motivation for problematic alcohol use in college students (Eisenberg, Toumbourou, Catalano, & Hemphill, 2014; Schulenberg
(Bravo & Pearson, 2017) or young adult drinkers (Goldsmith, & Maggs, 2002). A novel contribution of the present study is that
Thompson, Black, Tran, & Smith, 2012). the proposed effects were separately analyzed as a function of the
Moreover, individuals that experienced early-life stressful maximum level of education achieved by the participants, with or
events (e.g., child neglect or child abuse in early childhood) without a college degree.
exhibited an earlier drinking onset than their non-neglected peers. The relevance of assessing, in our sample, the role of educational
Notably, this first consumption was mainly aimed at reducing level achieved upon alcohol outcomes goes well beyond the criti-
depressive states (Rothman, Edwards, Heeren, & Hingson, 2008). cality of college for the establishment of patterns of alcohol con-
An early onset of alcohol drinking (ED) is, in turn, one of the most sumption. College life in Uruguay drastically differs from that found
reliable predictors of problematic alcohol use (Jackson, Barnett, in the U.S. or Europe, where the great majority of studies on youth
Colby, & Rogers, 2015; Moss, Chen, & Yi, 2014; Pilatti, Read, & drinking have been conducted. Illustrating this point, unlike their
Pautassi, 2017). The mechanisms underlying the effect of ED are US counterparts the Uruguayan students cannot enroll in frater-
still under debate (Hingson, Zha, & White, 2016; Kuntsche, Rossow, nities or sororities, which are groups that organize events where
Engels, & Kuntsche, 2016; Ystrom, Kendler, & Reichborn- alcohol consumption is prevalent and part of the rites of initiation
Kjennerud, 2014), with some suggesting that ED has a direct, and membership (McCabe, Veliz, & Schulenberg, 2018). Uruguayan
causal link upon the development of AUD (Hingson, Heeren, & college students, as many college students in South America (e.g.,
Winter, 2006a; 2006b) or alcohol-related negative consequences, Argentina) typically live with their families or with friends, rather
for instance by altering the trajectories of brain transmitter systems than in campuses. Moreover, the great majority [z80% in Uruguay
or the brain pruning that takes place at adolescence. Yet others (Lizbona & Rumeau, 2013)] matriculate in federal public univer-
indicate that ED could be a consequence of preexisting psychopa- sities close to their hometown. In summary, Spanish-speaking
thology [e.g., depression (McGue & Iacono, 2008)]. South-American youth in general, and Uruguayan youth in partic-
Under the framework of the studies reviewed, it could be ular, are not only understudied but also a rather unique population.
postulated that psychological distress or ED may act independently The need to examine determinants of alcohol use in more diverse
to promote alcohol intake. It could also be proposed that the level of socio-cultural groups has been highlighted (D'Amico, Tucker, Shih,
psychological distress is, in turn, a consequence of social contagion & Miles, 2014).
of mood. Emotional contagion is a change in the emotional state The study also described psychological distress, predisposition
that occurs because of the emotions that others convey through to emotional contagion, hazardous alcohol drinking and risk of
nonverbal behavior. Katz, Beach, and Joiner (1999) reported that alcohol abuse problems. We focused on whether age at drinking
people whose partner is depressed tend to exhibit depression onset and level of education significantly modulated these vari-
themselves, and found evidence of cross-contagion of levels of ables. One survey (Junta Nacional de Drogas, 2014) assessed alcohol
stress and emotional adjustment, in marriages where the two use in high-school students from Uruguay and reported an alarm-
members were medical students (Katz, Monnier, Libet, Shaw, & ing 68% prevalence of hazardous alcohol drinking in the last week.
Beach, 2000). The same federal agency found, in a separate survey conducted in
The present study analyzed, in a large sample (n ¼ 1505, 27% college students only, problematic use of alcohol –as defined by the
men, aged 18e30 years) of youth from Uruguay, the effect of Spanish Version of the Alcohol Use Disorder Test (S-AUDIT) – in 12%
psychological distress, age of first alcohol drink and predisposition of the students (Junta Nacional de Drogas, 2015a). Data derived
to emotional contagion on (a) the volume of alcohol consumed from a census-like survey indicated first alcohol use by age 14.3
during the last year and on (b) alcohol-induced negative conse- years (Junta Nacional de Drogas, 2015b). Despite alcohol use
quences (after controlling for the total volume of alcohol ingested). -particularly heavy drinking-being highly prevalent among Uru-
The age range under analysis largely coincides with the phase guayan youth, very few studies have described alcohol use behav-
known as emerging adulthood (Arnett, 2007), a developmental iors (or psychological distress) during the critical period of
period of high risk for the escalation of substance use. For emerging adulthood in this population. Moreover, to our knowl-
instance, marijuana use and heavy episodic drinking (HED, edge how age of onset of alcohol use modulates later alcohol use,
drinking 56/70 g of pure alcohol –women and men, respec- alcohol-related consequences and psychological distress has not
tively– in one drinking session) are more frequent among been analyzed in this population. The information of the present
emerging adults than in any other age group (Farmer et al., 2015; study will be useful to suggest avenues of intervention in this target
Patrick, Schulenberg, & O'Malley, 2016). group of this understudied population.
College attendance is a common milestone of emerging adult- Contextual variables seem to exert greater influence on the
hood, with college enrollment becoming increasingly common in consumption of women than in men (Brienza & Stein, 2002). It was
developing countries, such as Uruguay (Organizacio
n para la thus possible that educational level had a greater effect on the
Cooperacio
n y el Desarrollo Econo
mico, 2018). College education alcohol drinking patterns in women than in men. Women may be
is widely considered a promoter of personal and societal develop- particularly prone to the effect of depression on alcohol con-
ment, and early work (Droomers, Schrijvers, Stronks, van de sumption (Cranford, Eisenberg, & Serras, 2009; Pedrelli, Collado,
Mheen, & Mackenbach, 1999) found a protective effect of level of Shapero, Brill, & MacPherson, 2016), although some studies pro-
education upon problematic alcohol drinking. Yet more recent data vided opposite (Markman Geisner, Larimer, & Neighbors, 2004) or
suggests it is also associated with an increase in substance use inconclusive (Hussong, Jones, Stein, Baucom, & Boeding, 2011) ev-
behavior (Krieger, Young, Anthenien, & Neighbors, 2018; Merrill idence. Men and women react differentially to alcohol and it has
and Carey, 2016; Miech, Johnston, & O'Malley, 2017). It has been been suggested that women may progress more quickly from onset
suggested that those who enroll in college do not assume the social of alcohol use to AUD. However, and despite the growing awareness
roles and responsibilities (or do so very late) inherent of adult life of these differences and the explicit suggestion of including both
(work, family), which are incompatible with the lifestyle associated sexes (McCullough et al., 2014), women remain a neglected group
with heavy alcohol consumption (Fillmore, Carscadden, & Vogel- in alcohol-related research.
 P. Ruiz et al. / Alcohol 82 (2020) 91e101 93

Method collection may (when compared to the paper-and-pencil method)

affect response rates, but not the results.
Design, participants and procedure
Measures
This was a cross-sectional study that aimed at gathering data
from young Uruguayans (i.e., in the age range 18e30 years). An Alcohol use
invitation to participate in the study was distributed through social Alcohol use was measured using methods extensively used in
media and e-mail listings. The invitation asked for youth from the our lab (Pilatti, Caneto, Garimaldi, Vera, & Pautassi, 2014). Specif-
general community, 18e30 years old, willing to participate in a ically, alcohol use was defined as drinking at least one full glass (i.e.,
study assessing alcohol intake behaviors and their relationship 250 ml) of any alcoholic beverage. Participants were asked about
with psychological variables. The invitation emphasized the ano- lifetime, last year, regular (i.e., last month), and recent (last week)
nymity and academic nature of the study and contained a link at the alcohol use. Three additional questions asked about (1) the type of
end. Clicking on the link redirected to an active consent form that alcoholic beverage usually consumed (e.g. beer, wine, vermouth,
upon completion lead to the survey, which was made using Google fernet, vodka/rum and cider) (2) last year frequency (i.e. from 1 to 4
Forms online system. There was no compensation for participating times per year to 2e3 times per week), and (3) quantity (i.e. from 1
in the study. All the procedures were approved by the internal re- to 2 glasses to >12 glasses) of usual drinking of the selected
view board of Universidad de la República. beverage. Fernet is a type of Italian amaro that contains up to 43%
Participants were asked about age, sex, department of origin and alcohol by volume and is very popular among drinkers in Uruguay.
residence (Uruguay is divided into 19 departments), and the An additional question asked age at first alcohol use (how old were
highest level of education achieved and fully completed. According you the first time you consumed one full glass or more of any
to the divisions of the Uruguayan system of education, the options alcoholic beverage?). Following the criterion outlined by Lee, Hill,
were primary, secondary, secondary with technical specialization, et al. (2012) and Lee, Young-Wolff, et al. (2012) and Pilatti et al.
and university/undergraduate or university/postgraduate. The (2017), participants were classified as late drinkers (LD) if they
form was programmed to prevent missing data (e.g., participants reported first alcohol use at age 15, or as early drinkers (ED) if
received electronic prompts for any missed items) or being filled by they their first alcohol use occurred at age 14.
participants outside of the age range required. Despite these pre- Answers to questions one and three were used to calculate the
cautions, two cases were excluded because one only provided de- grams of alcohol consumed per drinking occasion and the number
mographic information and the other was outside the required age of standard drinks consumed per drinking occasion. The grams of
range of the study. Another 14 cases exhibited responses to the alcohol consumed were determined based on known alcoholic
alcohol use questions that were inconsistent or did not match the contents in each alcoholic beverage. More in detail, glasses
responses to the AUDIT, and eight cases indicated lifetime preva- consumed were first transformed to milliliters of alcohol consumed
lence of alcohol use yet did not inform age of first alcohol use. These per drinking occasion. This step was performed individually for
cases were also excluded. The final sample was composed of 1505 each type of alcohol beverage. For mixed drinks (i.e. fernet,
participants (25% men, Mean age ¼ 23.5 ± 3.5 years). Sample vermouth), we considered the usual amount of the alcoholic
characteristics are presented in Table 1. Most of the sample (71.2%) beverage needed to prepare a 250-ml glass of the corresponding
lived in the Montevideo department (i.e., the metropolitan area drink. One glass of beer or cider was defined as having a volume of
that includes Uruguay's capital and its surroundings), which 250 ml. One glass of wine was defined as having a volume of 150 ml
roughly corresponds to the actual geographical distribution of the and one glass of vermouth as having 125 ml. One glass of fernet and
population of the country. About 60% were employed and a similar one glass of any spirituous beverages (i.e. vodka) were defined as
percentage had between 18 and 24 years. These figures were having a volume of 75 ml of the alcoholic beverage. Then, this value
similar between men and women. (in milliliters) was multiplied by the reported number of glasses
Due to the online recruitment employed, it was not possible to consumed per drinking occasion. Finally, and considering the
calculate the participation rate. However, it should be noted that alcohol content of each beverage type and the alcohol density, we
according to previous research (Cronk & West, 2002), online data calculated the number of grams of alcohol consumed per drinking
occasion. One standard drink was defined as containing 14 g of
alcohol [i.e., one SD ¼ 14 g of alcohol].
A similar 3-question set asked for the second most consumed
Table 1 alcoholic beverage. Another question asked if these two beverages
Sociodemographic characteristics of the total sample and as a function of sex. were usually consumed together or an isolation. If participants
Total (n ¼ 1505) Men (n ¼ 378) Women (n ¼ 1127) reported consuming both beverages together (within a same
drinking session), grams of the two alcoholic beverages were added
Age
Mean 23.5 ± 3.5 24.0 ± 3.6 23.3 ± 3.5 up to estimate total amount consumed per drinking session. Then,
Age range 18e24 61.3 56.6 62.9 we calculated the number of SD consumed per drinking occasion.
Age range 25 a 30 38.7 43.4 37.1 Grams of alcohol per drinking occasion and last-year frequency of
Residenec alcohol use were used to calculate total volume of alcohol
Montevideo 71.2 75.4 69.8
Other 28.8 24.6 30.2
consumed during the last 12 months.
Work Status Participants also indicated the frequency of engaging in heavy
Employed 60.5 62.2 59.9 episodic and binge drinking episodes within the previous 12
Unemployeda 39.5 37.8 40.1 months (from 0 ¼ I do not drink alcohol/I do not drink that amount
Academic Status
of alcohol to 8 ¼ four or more times per week). Answers to these
Non-college 50.4 56.6 48.4
College 49.6 43.4 51.6 questions were used to estimate the number of heavy and binge
drinking episodes within the previous year. We defined HED as
For nominal variables, data are presented as percentages of participants placed in
each category. For continuous variables, data are presented as means and standard
drinking 4/5 standard drinks (women/men, respectively) in one
deviation. drinking session; whereas binge drinking was defined as drinking
a
Also including those not looking for a job. those quantities in 2 h. Separate questions inquired about
94 P. Ruiz et al. / Alcohol 82 (2020) 91e101

lifetime, last year and last month occurrence of drunkenness, and Data analysis
number of drunkenness episodes in the previous month. The reli-
ability of the four questions that interrogate about frequency and For the purposes of data analysis and based on previous litera-
quantity of the two most consumed beverages is adequate ture that emphasizes comparisons between college and non-
(a ¼ 0.79) (Pilatti, Read, & Caneto, 2016). college students (Eisenberg et al., 2014; Schulenberg & Maggs,
S-AUDIT (Rubio Valladolid, Bermejo Vicedo, Caballero Sanchez- 2002), participants were clustered into college (COLL, 43.4% and
Serrano & Santo-Domingo Carrasco, 1998): The instrument (Babor, 51.7% of the men and women sub-samples) and non-college (NON-
2001) has 10 items with an ordinal scale of response. Seven items COLL ¼ primary, secondary or technical careers completed)
are answered with a scale of five points (from 0 to 4) and three with participants.
a scale of three points (0, 2 and 4). The Spanish version (Pilatti et al.,
2016; Rubio Valladolid et al., 1998) has shown adequate values of Descriptive analyses
internal consistency (a ¼ 0.77 to a ¼ 0.86). In the present study S- Frequency, percentage, central tendency or deviation indices
AUDIT scores were analyzed as a function of sex, age at drinking were calculated for the overall sample and separately for each in-
onset and level of education, and were also used to check the dependent variable of interest (i.e., sex, age of onset category and
consistency and reliability of the responses to questions on alcohol level of education), to describe the following dependent variables:
use. prevalence of alcohol use (lifetime or last year/month/week),
prevalence of last year HED, last year binge, drunkenness (lifetime
or last year/month) and, in last year drinkers, volume (mean grams)
Spanish young adult alcohol consequences questionnaire [S-YAACQ of alcohol ingested per drinking episode, number of negative con-
(Pilatti et al., 2016; Pilatti, Read, et al., 2014)] sequences of alcohol use and AUDIT scores.
We used the Spanish version of the original YAACQ (Read,
Kahler, Strong, & Colder, 2006), a 48-items measure to assess the Group analyses
occurrence (yes/no) of negative consequences of alcohol drinking. To facilitate the inspection of occurrence of alcohol use behav-
The S-YAACQ encompasses a broad myriad of alcohol-related iors, the differences in prevalence of alcohol use and drunkenness
negative consequences, covering from mild (e.g. having a hang- according to sex, age of onset or educational level were determined
over) to more severe (e.g. driving a car under the influence) con- (separately for each of these independent variables) using c2 tests,
sequences. Answers are added up to reflect the total number of whereas differences between these groups in grams of alcohol
consequences experienced during the past year. The scale has been ingested per drinking episode, AUDIT scores and negative conse-
mainly used in college students, yet it should be noted that the quences of alcohol use were assessed via Student's t test. More in
items also illustrate examples relevant to non-college participants detail, all variables were separately analyzed for significant women
(e.g., “I have not gone to work or missed classes at school because of vs. men, ED vs. LD drinkers and COLL vs. NON-COLL differences.
drinking…”). Previous work has found adequate values of reliability These analyses involved a total of 35 comparisons. Type 1 error
(Pilatti et al., 2016). In the present study we conducted separate inflation was controlled by decreasing the alpha value to 0.0014
hierarchical regressions for participants with or without college (i.e., the product of dividing the conventional 0.05 threshold by 35).
degree, therefore we calculated reliability (Cronbach's Alpha) One of the aims of the study was to analyze if psychological
within each subsample, and both indices were adequate (a distress, predisposition to emotional contagion, risk level for
college ¼ .93, a non-college ¼ .93). alcohol abuse problems (i.e., total AUDIT score) and hazardous
alcohol drinking behaviors (considered as dependent variables)
were independently affected by the independent variables age at
Spanish version of the Kessler Scale of psychological distress (Brenlla drinking onset and level of education. Therefore, the total scores
& Aranguren, 2010) derived from the AUDIT, Kessler and Doherty scales, and the
This instrument, the Spanish version of the original screening number of episodes of heavy and episodic drinking (within the
scale of psychological distress (Kessler et al., 2002), consists of 10 previous 12 months), were analyzed, as dependent variables,
items, with a five-point Likert type response format (from 1 ¼ none through separate factorial Analyses of Variance (ANOVAs) that
of the time to 5 ¼ all of the time), in which higher scores indicate considered Sex (male, female), Age of first alcohol use (ED, LD) and
higher levels of psychological distress. The scale assesses the risk of Education level (COLL, NON-COLL) as between-group factors. We
presenting non-specific psychological distress in the last month employed Tukey's post hoc comparisons to scrutinize significant
(Brenlla & Aranguren, 2010; Kessler et al., 2002), and for the most interactions in the five ANOVAs conducted.
part the items ask about occurrence of symptoms of anxiety or The ANOVAs were conducted in participants that reported life-
depression. The Spanish version (Brenlla & Aranguren, 2010) meets time alcohol drinking or, in the case of the number of episodes of
criteria of validity and reliability (a ¼ 0.88). In the present study, we heavy and episodic drinking, in participants that reported alcohol
calculated reliability (Cronbach's Alpha) as a function of level of drinking during the last year. The partial eta-squared (h2p) was
education, and both indices were adequate (a college ¼ .83, a non- used to describe effect sizes of these ANOVAs, which were inter-
college ¼ .85). preted using the following guidelines [small (h2p ¼ 0.01e0.05),
medium (h2p ¼ 0.06e0.13), and large (h2p ¼ 0.14) (Lakens, 2013)].
Given the difficulty of representing significant main effects or sig-
Spanish version of the Doherty Scale of Emotional Contagion nificant interactions that span several groups, the significant, main
(Gouveia, Valeschka, Santos & de Medeiros, 2007) effects or group-to-group differences have not been depicted in the
We employed the Spanish version of the Emotional Contagion figures (via stars or other signs). Instead, a brief description of these
Scale (Doherty, 1997). This instrument assesses susceptibility to significant effects and differences can be found in each figure
emotional contagion and consists of 18 questions answered with a legend.
Likert response format (from 1 ¼ always to 4 ¼ never), with lower
scores indicating higher sensitivity or predisposition to emotional Bivariate associations
contagion. Reliability values in the present study were adequate (a The Pearson coefficient was employed to analyze the association
college ¼ .79, a non-college ¼ .80). between the measures related to alcohol consumption, mood and
 P. Ruiz et al. / Alcohol 82 (2020) 91e101 95

emotional contagion. Specifically, productemoment correlations group of men, z7 SU). COLL and NON-COLL participants exhibited
were calculated between YAACQ scores, AUDIT scores, emotional similar rates of alcohol use. These results are shown in Table 2. As
contagion scores, psychological distress scores, volume of alcohol shown in Supplementary Table 1, the majority of last year drinkers
(grams) drank in the last year, grams per drinking episode, fre- engaged in heavy episodic or binge drinking during the last year
quency of heavy episodic and frequency of binge drinking. These (83.3 and 65.4%, respectively), whereas 15 and 10.4% of these par-
analyses were conducted separately for COLL and NON-COLL par- ticipants engaged in heavy episodic drinking and binge drinking,
ticipants, thus a total of 28 correlations were conducted. These respectively, at least once per week.
large number of comparisons considerable inflates Type 1 error. Almost one fourth of the participants exhibited scores 8 in the
Thus, a more stringent alpha value was employed: 0.018 (i.e., 0.05/ AUDIT, which is the threshold for a presumptive diagnosis of
28). problems with alcohol (Babor, 2001). Among these, late drinkers
exhibited the lowest (16.3%). while men and early drinkers
Multivariate associations exhibited the highest (30%), occurrence of this hazardous drink-
Multiple hierarchical regression analyses were performed, ing pattern.
separately for COLL and NON-COLL participants reporting last year The ANOVA that considered psychological distress as dependent
alcohol consumption. These analyses aimed at assessing if the level variable (Fig. 1, top panel) yielded significant main effects of Sex
of emotional contagion explained, with psychological distress and (F1,1456 ¼ 36.70, p < 0.001, h2p ¼ 0.02), Maximum level of education
the age of onset of alcohol drinking (in this case, the exact age of (F1,1456 ¼ 5.22, p < 0.05, h2p ¼ 0.004) and Age of onset of alcohol use
onset, in years, reported by each participant), the total number of (F1,1456 ¼ 5.47, p < 0.05, h2p ¼ 0.003). Women, Non-college par-
alcohol-related negative consequences (after controlling for the ticipants, and participants with ED exhibited greater levels of
total volume of alcohol ingested during the last 12 months) or the psychological distress than their respective counterparts. Sensi-
volume of alcohol consumed during the last year. It should be tivity to emotional contagion, in turn, was lower in women (F1,
1456 ¼ 69.41, p < 0.001, h p ¼ 0.05) and in those that had achieved a
2
noted, however, that the final regression model was built based on
the outcomes of the bivariate correlations. Moreover, due to the college degree (F1, 1456 ¼ 8.01, p < 0.01, h2p ¼ 0.01) (Fig. 1, bottom
significant variations found in alcohol consumption between men panel). None of the interactions, neither for psychological distress
and women, sex was also included in the regressions. nor for contagion scores, achieved significance.
Specifically, in the first step of the model predicting alcohol- The number of episodes of heavy episodic drinking was signif-
related consequences we entered Sex and Total volume of alcohol icantly greater in men than in women (significant main effect of
ingested during the last 12 months. Age of onset and Psychological Sex, F1,1357 ¼ 4.25, p < 0.05, h2p ¼ 0.003) and in ED than in LD
distress were introduced in the second step. It is important to participants (significant main effect of Age of onset, F1,1357 ¼ 9.02,
remark that, in COLL and in NON-COLL participants, the bivariate p < 0.005, h2p ¼ 0.01); whereas frequency of binge drinking was
association between emotional contagion and alcohol-related also greater in ED than in LD drinkers (F1,1357 ¼ 5.00, p < 0.05,
consequences was not significant. Therefore, emotional contagion h2p ¼ 0.003), yet similar in men and women. Level of Education
was not included in the regression. A similar regression model was was included in these analyses as a between-group factor (i.e., an
used to predict volume of alcohol ingested during the last 12 independent variable) yet it did not exert significant main effects
months. Specifically, Sex was entered first, followed by Age of onset and none of the interaction terms achieved significance. These re-
and Psychological distress. Emotional contagion was not signifi- sults are depicted in Fig. 2.
cantly associated at the bivariate level with the outcome (i.e., vol- Mean AUDIT scores across groups ranged from 4 (LD and
ume of alcohol ingested during the last 12 months), thus was not women) to z 7 (men), which is the threshold for a presumptive
included in the model. We calculated regular multiple correlation diagnosis of problems with alcohol [(score  8  20) (Babor, 2001)].
coefficients (R2) and standardized regression coefficients, with an The ANOVA that included AUDIT scores as dependent variable
alpha of 0.05. (Fig. 3) revealed a significant main effect of age at drinking onset,
with ED participants scoring significantly higher than LD peers
Results (F1,1456 ¼ 56.90, p < 0.001, h2p ¼ 0.04), a significant main effect of
Sex (F1,1456 ¼ 54.52, p < 0.001, h2p ¼ 0.04), and a significant
Descriptive results interaction between Sex and Maximum educational level achieved
(F1,1456 ¼ 6.38, p < 0.05, h2p ¼ 0.004). The post-hoc tests revealed
As shown in Table 2, lifetime alcohol consumption was almost that COLL and NON-COLL women exhibited similar AUDIT scores,
normative (97.3%), yet markedly similar in men and women. Men yet AUDIT scores were significantly higher in COLL men than in
exhibited significantly higher prevalence of alcohol use than NON-COLL men.
women in the last 7 (53.7% vs. 41.7%, respectively) days, but not in
the last year or month. Mean also exhibited, when compared to Bivariate correlations
women, significantly greater prevalence of drunkenness in all the
time-frames measured and in prevalence of heavy episodic and Table 3 illustrates the bivariate correlations between YAACQ
binge drinking and, probably because of these differences, drank scores, AUDIT scores, emotional contagion scores, psychological
significantly more grams of alcohol per drinking episode (99.5 vs. distress scores, volume of alcohol (grams) drank in the last year,
72.7 g). grams per drinking episode and frequency of heavy episodic or
binge drinking episodes in the last year; for COLL and NON-COLL
Group differences participants.
In both samples there was a positive and significant correlation
Across alcohol use indicators, ED participants exhibited signifi- between psychological distress and the negative consequences of
cantly higher scores than LD peers. Notably, the prevalence of drinking (i.e., YAACQ scores), and between the latter variable and
alcohol consumption in the last 7 days was 37.8% in LD vs. 56.2 in the annual frequency of heavy episodic or binge drinking. This is,
ED, and 30% of ED reported drunkenness in the last month (against those who experience more psychological distress or who exhibi-
19% of LD). Across conditions, the average alcohol drinking (grams) ted greater frequency of risky drinking behaviors experienced more
per episode ranged from 70.1 (LD group, z5 SU) to 99.5 (overall negative consequences associated with alcohol consumption.
96 P. Ruiz et al. / Alcohol 82 (2020) 91e101

Table 2
Alcohol outcomes for the total sample and as a function of sex, age of alcohol drinking onset and level of education.

Total sample Ever alcohol users Level of education

Total Men Women c2/t ED LD c2/t Non-Coll Coll c2/t

Alcohol Use
Lifetime 97.3 97.1 97.3 .07 100 100 e 97 97.6 0.5
12M 90.7 90.7 90.7 .01 96.5 90.7 19.4 89.9 91.6 1.3
30D 73.4 78.8 71.6 7.6 82.6 69.7 32.5 72.7 74 .38
7D 44.7 53.7 41.7 16.5 56.2 37.8 49.6 42.2 47.3 4.1
Last year HED 75.6 79.6 74.3 4.4 84.9 72 35.3 75.5 75.7 0.1
Last year binge 59.3 63.5 57.9 3.6 67.9 55.5 23.4 57.7 61 1.7
Grams per episode1 79.5 ± 64.7 99.5 ± 78.6 72.9 ± 57.9 6.7 90.6 ± 68.7 70.1 ± 59.6 5.9 80.4 ± 66.9 78.7 ± 62.5 0.5
Drunkenness
Lifetime 75.7 80.4 74.1 158.7 86.6 70.6 53.7 75.4 76 0.1
12M 48.9 57.9 45.9 16.5 58.7 43.5 33.1 49.5 48.3 0.3
30D 23.3 30.2 21 13.2 29.8 19.3 21.8 23.2 23.5 0.02
Neg Cons1 7.9 ± 7.6 9.9 ± 8.4 7.3 ± 7.20 5.4 9.8 ± 8.30 6.4 ± 6.6 8.4 7.9 ± 7.6 7.9 ± 7.7 0.04
AUDIT1 5.3 ± 4.6 7.01 ± 5.3 4.8 ± 4.20 7.9 6.5 ± 4.90 4.4 ± 4.0 8.7 5.4 ± 4.4 5.4 ± 4.8 0.1
0 3.4 1.2 4.2 1.4 5.2 3.4 3.5
1e7 73.2 62.4 76.8 66.9 78.6 73.9 72.5
8e15 18.8 28.9 15.5 25 13.6 19.1 18.6
16 4.5 7.6 3.5 6.7 2.7 3.7 5.4

LD ¼ late drinkers; ED ¼ early drinkers; COLL ¼ participants with complete graduate or post-graduate education; NON-COLL ¼ participants with primary or secondary
education degrees; Lifetime ¼ lifetime use; 12M ¼ last year occurrence; 30D ¼ last month occurrence; 7D ¼ last week occurrence; HED ¼ heavy episodic drinking (4/5
standard drinks in one drinking session); F ¼ frequency; Binge ¼ binge drinking (4/5 standard drinks in 2 h); Q D 30D ¼ number of drunkenness episodes within the previous
month; Neg Cons ¼ negative alcohol-related consequences (total score obtained in the Young Adult Alcohol Consequences Questionnaire). 1These results correspond to last
year drinkers. Significant results (p < .0014) are in bold typeface for emphasis.

Fig. 1. Total scores obtained in the Spanish version of the Kessler Scale of psychological distress (upper panel) or in the Spanish version of the Doherty Scale of Emotional Contagion
(lower panel), as a function of maximum level of education achieved (COLL: college; NON-COLL: non-college), sex and early or late onset of alcohol use. Participants were classified
as late drinkers (LD) if they reported first alcohol use at age 15, or as early drinkers (ED) if they their first alcohol use occurred at age 14. The ANOVAs revealed that women, Non-
COLL participants, and participants with ED exhibited significantly greater levels of psychological distress than their respective counterparts; and that sensitivity to psychological
contagion was lower in women than in men, and in COLL compared to NON-COLL participants. The vertical bars indicate SEM.
 P. Ruiz et al. / Alcohol 82 (2020) 91e101 97

Fig. 2. Frequency of heavy episodic or binge drinking in last-year drinkers, as a function of maximum level of education achieved (COLL: college; NON-COLL: non-college), sex and
early or late onset of alcohol use. Participants were classified as late drinkers (LD) if they reported first alcohol use at age 15, or as early drinkers (ED) if they their first alcohol use
occurred at age 14. Heavy episodic drinking was defined as drinking 4 or 5 standard drinks (women and men, respectively) in one drinking session; whereas binge drinking was
defined as drinking those quantities in 2 h. The ANOVAs revealed the number of episodes of heavy episodic drinking was significantly greater in men than in women and in ED
than in LD participants; whereas frequency of binge drinking was also greater in ED than in LD drinkers, yet similar in men and women (p > 0.05). Level of Education did not exert
significant main effects and none of the interaction terms achieved significance. The vertical bars indicate SEM.

Psychological distress was not significantly associated with heavy episodic and binge drinking, and with AUDIT and YACCQ scores, in
episodic or binge drinking, yet total volume of alcohol consumed in NON-COLL but not in COLL participants.
COLLEGE, but not in NON-COLLEGE, students exhibited a negative We also observed a significant negative correlation between
and significant correlation with psychological distress. Total vol- psychological distress and emotional contagion scores, which
ume of alcohol consumed was positively associated with heavy indicated that those more prone to emotional contagion (i.e. lower

Fig. 3. Total scores obtained in the Spanish Version of the Alcohol Use Disorder Test [S-AUDIT], as a function of maximum level of education achieved (COLL: college; NON-COLL:
non-college), sex and early or late onset of alcohol use. Participants were classified as late drinkers (LD) if they reported first alcohol use at age 15, or as early drinkers (ED) if they
their first alcohol use occurred at age 14. The ANOVA and subsequent Tukey post-hoc tests revealed that ED participants scored significantly higher than LD participants. These
analyses also indicated that COLL and NON-COLL women exhibited similar AUDIT scores, yet AUDIT scores were significantly higher in COLL men than in NON-COLL men. The
vertical bars indicate SEM.
98 P. Ruiz et al. / Alcohol 82 (2020) 91e101

Table 3 p  .001) were significantly associated with greater number of

Bivariate correlations between emotional contagion and psychological discomfort alcohol-related consequences.
with diverse alcohol outcomes in college and non-college youth.
NON-COLL: In the first step, the volume of alcohol consumed,
1 2 3 4 5 6 7 8 but not the sex of the participants, modestly (R2 ¼ .06) but signif-
1.Grams .02 .30 .25 .51 .55 .06 .03 icantly explained the variance of alcohol-related negative conse-
2.Total Volume .33 .01 .01 -.04 .02 .07 -.12 quence [F change (2,679) ¼ 21.88 p  .001]. The greater the volume of
3.F HED .36 .27 .71 .37 .42 .03 .05 alcohol consumed, the greater the number of alcohol-related con-
4.F binge .24 .22 .71 .30 .34 .02 .07
sequences (b ¼ .23, t ¼ 6.21, p  .001). Age of onset and psycho-
5.Neg Cons .52 .24 .47 .33 .78 .01 .15
6.AUDIT .54 .23 .47 .30 .77 .02 .04 logical distress significantly increased, in the second step, the
7.Em Contagion -.01 .02 -.02 -.02 -.04 .02 -.14 explained variance to 14% [F change (2,677) ¼ 31.70 p  .001]. A
8.Psych distress .08 .01 .06 .06 .24 .14 -.09 greater level of psychological distress (b ¼ .24, t ¼ 6.52, p  .001)
The upper triangle presents results among COLL participants. The lower triangle and a lower age of onset (b ¼ .14, t ¼ 3.77, p  .001) were
presents results among NON-COLL participants; Total Volume ¼ total volume of associated with greater number of alcohol-related negative
alcohol consumption during the last year; F HED ¼ frequency of heavy drinking; F consequences.
Binge ¼ frequency of binge drinking; Em Contagion ¼ total scores obtained in the
Doherty scale; Neg Cons ¼ negative alcohol-related consequences (total score ob-
tained in the Young Adult Alcohol Consequences Questionnaire); Psych Discussion
Disc ¼ Psychological Discomfort (total scores obtained in the Kessler scale). Sig-
nificant results (p < .018) are in bold typeface for emphasis. Our findings indicate that, either at the bivariate or multivariate
level, a higher level of psychological distress among emergent
adults from Uruguay was significantly associated with a greater
scores on the Doherty scale indicate greater propensity to
number of alcohol-related negative consequences. Psychological
emotional contagion), are more likely to experience psychological
distress, however, was not significantly correlated with heavy
distress. In NON-COLL participants, AUDIT scores were significantly
episodic or binge drinking, nor greater distress levels promoted
and positively associated with all the alcohol use indicators, and
greater volume of alcohol use during the last year. This dissociation
with YACCQ and Kessler scores. The latter pattern was similarly
suggests that, in our sample, psychological distress is associated
exhibited by COLL participants, yet they did not exhibit a significant
with alcohol-related negative consequences, but not with a heavy
correlation between AUDIT scores and total volume of alcohol
episodic pattern of alcohol consumption.
consumed or Kessler scores.
Drinking motives are considered the last link of the etiological
chain leading to alcohol intake and include, among other,
Regression analyses on volume of alcohol consumed during the last enhancement, conformity, coping with anxiety or depression and
year social order motives (Lannoy, Billieux, Poncin, & Maurage, 2017).
Drinking motives were not assessed in the present work. Despite
COLL: In the first step, the entering of sex increased 2% the this limitation, the lack of positive associations between psycho-
explained variance [F change (1,681) ¼ 12.28 p  .001]. Being male logical distress and alcohol use behaviors suggest that, in our
(b ¼ .13, t ¼ 3.50, p  .001) was associated with greater volume sample, motivations other than coping with psychological distress
of alcohol consumed during the last year. The entering of psycho- underlied the occurrence of heavy episodic or binge drinking.
logical discomfort and age of onset, in the second step, contributed Future studies akin to the present should consider the inclusion of
a mere additional 1% to the explained variance to the explained an explicit measurement of drinking motives (e.g., the Drinking
variance [F change (2,679) ¼ 3.86 p  .002]. A lower level of psy- Motives Questionnaire) (Kuntsche, Stewart, & Cooper, 2008).
chological discomfort (b ¼ .10, t ¼ 2.65, p  .001) was associated It is also possible that psychological distress was associated with
with greater volume of alcohol consumed during the last year. drinking behaviors not measured in this work. For example, the
NON-COLL: The level of explained variance significantly implementation of protective behavioral strategies (PBS) against
increased to 3% after entering sex on the first step [F change alcohol's effects (Pearson, 2013) allows drinkers to reduce the
negative consequences associated with alcohol consumption, even
(1,680) ¼ 22.31 p  .001]. Being male (b ¼ .18, t ¼ 4.72, p  .001)
was significantly associated with greater volume of alcohol when drinking has already occurred or is about to occur. Drinkers
consumed during the last year. The entering of psychological can plan in advance when to consume (to avoid affecting later work
discomfort and age of onset, in the second step, increased the or academic activity) or avoid exceeding a certain number of drinks
variance explained to 4% [F change (2,678) ¼ 3.36 p  .05]. A lower or ask for help from friends to mitigate the likelihood of experi-
age of onset (b ¼ .09, t ¼ 2.32, p  .05), but not psychological encing negative consequences (e.g., physical attacks, involvement
distress, was associated with greater volume of alcohol consumed in unwanted sexual behavior) of alcohol (Martens et al., 2009).
during the last year. Interestingly, it has been suggested that psychological distress
moderates the relationship between PBS and alcohol consequences
(Jordan, Villarosa-Hurlocker, Ashley, & Madson, 2018). Therefore, it
Regression analyses on alcohol-related negative consequences could be hypothesized that e in our sample – individuals with high
(YAAC-Q scores) psychological distress did not have the capacity to implement these
strategies and therefore were more prone to suffer the negative
COLL: Sex and volume of alcohol consumed during the last year consequences of drinking. The available literature, however, does
explained, in the first step, 5% of the variance of alcohol-related not support the hypothesis. Studies conducted in college students
negative consequence [F change (2,680) ¼ 17.06 p  .001]. Being reported that the negative association between PBS and alcohol-
male (b ¼ .18, t ¼ 5.01, p  .001) was associated with a greater related consequences or binge drinking was strengthened among
number of alcohol-related negative consequences. The entering of those with high levels of psychological distress (Kenney & LaBrie,
psychological distress and age of onset of drinking significantly 2013; LaBrie, Kenney, Lac, Garcia, & Ferraiolo, 2009).
increased the level of explained variance to 14% [F change (2, At the bivariate level, we found a positive, significant association
678) ¼ 36.61 p  .001]. Greater scores in the Kessler (b ¼ .19, t ¼ 5.05, between sensitivity to emotional contagion and psychological
p  .001) and a lower age of drinking onset (b ¼ .24, t ¼ 6.71, distress. This suggests that greater sensitivity to emotional
 P. Ruiz et al. / Alcohol 82 (2020) 91e101 99

contagion may promote negative alterations in mood. Although The issue just raised pinpoints a limitation of the present work.
highly speculative, this hypothesis is consistent with a clinical The cross-sectional nature of the study prevents establishing,
study which found a positive correlation between emotional beyond speculation, cause-effect relationships or assessing for
contagion and depressive symptoms (Sturm et al., 2013). We did "true mediation". Establishment of the latter requires a temporal
not find, however, significant correlations between sensitivity to ordering of the independent variable, mediator and dependent
emotional contagion and alcohol outcomes (YAACQ scores and variable. Other limitations are the use of self-report measures,
frequency of heavy or binge drinking), suggesting that emotional which may have resulted in distorted responses due to social
contagion is not involved in the effect of psychological distress on desirability or because of failures in the retrieval of information
the negative consequences of alcohol consumption. A limitation of (e.g., age of first drink is affected by telescopic bias). Also, the online
our conclusions is, however, that the level of adjustment of the recruitment impeded calculating the participation rate of the study
multivariate models was low. and may have introduce unwanted bias in the sample. For instance,
A second objective of this study was to analyze how the we only reached Uruguayans with regular exposure to internet.
occurrence of alcohol drinking behaviors was affected by sex, an Likewise, our sample was composed primarily of women and res-
early age of first drink and the educational level of the participants. idents of the Department of Montevideo, so the results should be
Regarding sex, we observed that men and women exhibit statisti- extrapolated with caution to populations with better sex balance
cally similar consumption in many drinking indicators (e.g., life- and to the other departments of Uruguay.
time, last year and last month drinking). These results are We did not observe differences between COLL and NON-COLL in
consistent with several studies (Grucza, Norberg, & Bierut, 2009; the patterns of alcohol consumption measured or, at a descriptive
Keyes, Li, & Hasin, 2011) that suggest a shrinking of the gap in level, in the negative consequences of this consumption. These
alcohol consumption between men and women. It is important to findings disagree with those derived from studies reporting that
note, however, that we still found differences in favor of men when university students exhibit exacerbated alcohol consumption than
focusing on high-risk drinking behaviors, such as recent drinking or same-age counterparts not attending college (Huckle, You, &
prevalence of binge drinking. Some of these differences were sub- Casswell, 2010; Lui & Zamboanga, 2018a, 2018b). Most of the
stantial (e.g., drinking in the last week: 54% vs. 42%; occurrence of studies indicating this association, however, have been conducted
intoxication in the last year: 58% vs. 46%), and the occurrence of in countries whose college differ substantially from that found in
heavy episodic drinking in the last year were also significantly Uruguay. The vast majority of Uruguayan college students do not
higher in men than in women. This pattern is reminiscent of that pay tuition fees, nor live in campus nor have the possibility to
found with a sample of Argentinian freshman university students endorse fraternities or sororities. Future studies should benefit
(Pilatti et al., 2017). In that work, men and women showed statis- from comparing objective aspects and internalized values of college
tically similar rates of lifetime and last-year prevalence of alcohol life (Osberg et al., 2016) in diverse socio-cultural groups. A limita-
consumption, yet men exhibited significantly greater levels of tion of our conclusions, however, is that the distinction between
heavy episodic drinking than women. "COLL" and "NON-COLL" participants was based on completion of
It has been suggested that contextual variables affect alcohol the college degree. It is this possible that some of those in the
consumption to a greater extent in women than in men (Brienza & "NON-COLL" group could have been enrolled in college before,
Stein, 2002). Consistent with this hypothesis we recently reported thereby it is possible that the two groups are not mutually exclusive
that women, but not men, exhibit a positive association between from each other.
greater availability of alcohol outlets and alcohol-related negative It is difficult to compare the occurrence of alcohol drinking
consequences (Montejano, Pilatti, Pitt & Pautassi 2019). We thus found in the present results with those reported by national sur-
expected that maximum level of education achieved would affect veys conducted in Uruguay, which sampled either younger or older
the levels of alcohol drinking behaviors, psychological distress or populations, or only focused in university students, and have sub-
sensitivity to emotional contagion more in women than in men. stantial methodological differences with our study (Junta Nacional
This was not the case. Moreover, sex did not affect the promoting de Drogas, 2014; 2015a; 2015b). The present study is methodo-
effect of an early onset of alcohol consumption on heavy episodic or logically akin to a study in a large sample of Argentinean university
binge drinking. ED facilitated - as was our hypothesis - all drinking students (Pilatti et al., 2017). We observed a high prevalence of
indicators, in some cases dramatically: e.g., drinking in the last 7 risky alcohol consumption. Approximately 40% of the participants
days increased from 37.8% in LD to 56.2% ED. Also important is that, reported heavy episodic drinking at least once a month, about one
like previous research, we found greater sensitivity to emotional third reported binge drinking once a month, and three out of four
contagion in women than in men (Wild, Erb, & Bartels, 2001). indicated lifetime drunkenness. These figures are lower than those
The results confirm - for this understudied Uruguayan popula- reported by Pilatti et al. (2017), who observed 60 and 42% of high
tion - the facilitating effect of the early onset of alcohol on the and intensive episodic prevalence, respectively, in the last month. It
consumption patterns of this drug in adolescents and young adults. should be considered, however, that our sample had a higher than
As has been reported across samples and cultures (Lee, Young- that of Pilatti et al. (2017) and that our work recruited the general
Wolff, Kendler, & Prescott, 2012; McGorry, Purcell, Goldstone, & population, not only university students. Despite these differences,
Amminger, 2011; Mooney, Fromme, Kivlahan, & Marlatt, 1987; the present results highlight the extent of alcohol-related behaviors
Pilatti, Cupani, & Pautassi, 2015) ED is one of the best predictors of and consequences among Uruguayan youth. The findings suggest
later alcohol consumption. An interesting result was that ED par- that interventions aimed at managing psychological distress or
ticipants exhibited significantly higher levels of psychological those that try to delay alcohol initiation should be beneficial in this
distress than their LD peers. This result is consistent with the hy- population.
pothesis (Buchmann et al., 2009) that an early onset of alcohol Overall, the present study cements the notion that, during
consumption is an emergent of genetic risk or underlying psycho- adolescence and youth, psychological distress is associated with
pathology. It is also possible, however, that the greater psycholog- experiencing negative consequences of alcohol consumption. Pro-
ical distress of ED participants is actually a consequence of their pensity to emotional contagion was not associated with alcohol
longer time of exposure to drinking (i.e., the interval between the drinking nor with alcohol-related negative consequences, although
onset of drinking and the measurement of psychological distress is emotional contagion and psychological distress were significantly
greater in ED than in LD participants). associated. The study also suggested that greater levels of
100 P. Ruiz et al. / Alcohol 82 (2020) 91e101

psychological distress may underlie the facilitating effect of an early Grucza, R. A., Norberg, K. E., & Bierut, L. J. (2009). Binge drinking among youths and
young adults in the United States: 1979-2006. Journal of the American Academy
age of drinking onset upon alcohol drinking patterns.
of Child & Adolescent Psychiatry, 48(7), 692e702.
Hingson, R. W., Heeren, T., & Winter, M. R. (2006a). Age at drinking onset and
Declaration of interest alcohol dependence: Age at onset, duration, and severity. Archives of Pediatrics
and Adolescent Medicine, 160(7), 739e746.
Hingson, R. W., Heeren, T., & Winter, M. R. (2006b). Age of alcohol-dependence
None. onset: Associations with severity of dependence and seeking treatment. Pedi-
atrics, 118(3), e755ee763.
Hingson, R., Zha, W., & White, A. (2016). The usefulness of 'age at first drink' as a
Acknowledgements concept in alcohol research and prevention. Addiction, 111(6), 968e970.
Huckle, T., You, R. Q., & Casswell, S. (2010). Socio-economic status predicts drinking
patterns but not alcohol-related consequences independently. Addiction, 105(7),
This work was supported by PICT 2015-0325 to RMP, and 1192e1202.

n e Innovacio
Agencia Nacional de Investigacio
n, Grant/Award Hussong, A. M., Jones, D. J., Stein, G. L., Baucom, D. H., & Boeding, S. (2011). An
Number: POS_EXT_2014_1_105 877. internalizing pathway to alcohol use and disorder. Psychology of Addictive Be-
haviors, 25(3), 390e404.
Jackson, K. M., Barnett, N. P., Colby, S. M., & Rogers, M. L. (2015). The prospective
Appendix A. Supplementary data association between sipping alcohol by the sixth grade and later substance use.
Journal of Studies on Alcohol and Drugs, 76(2), 212e221.
Jordan, H. R., Villarosa-Hurlocker, M. C., Ashley, A. L., & Madson, M. B. (2018). Pro-
Supplementary data related to this article can be found at tective behavioral strategies and hazardous drinking among college students:
https://doi.org/10.1016/j.alcohol.2019.09.001. The moderating role of psychological distress. Journal of Drug Education,
47237918800505.
Junta Nacional de Drogas. (2014). Sexta encuesta nacional sobre consumo de drogas
References en estudiantes de ensen ~ anza media.
Junta Nacional de Drogas. (2015a). I estudio piloto sobre consumo de drogas en
Arnett, J. J. (2007). Emerging adulthood: What is it, and what is it good for? Child estudiantes universitarios de Uruguay.
Development Perspectives, 1(2), 68e73. Junta Nacional de Drogas. (2015b). VI Encuesta nacional en hogares sobre consumo de
Babor, T. F. (2001). Audit : The alcohol use disorders identification test : Guidelines for drogas.
use in primary health care. Geneva. Katz, J., Beach, S. R. H., & Joiner, T. E., Jr. (1999). Contagious depression in dating
Bravo, A. J., & Pearson, M. R. (2017). In the process of drinking to cope among college couples. Journal of Social and Clinical Psychology, 18(1), 1e13.
students: An examination of specific vs. global coping motives for depression Katz, J., Monnier, J., Libet, J., Shaw, D., & Beach, S. R. (2000). Individual and crossover
and anxiety symptoms. Addictive Behaviors, 73, 94e98. effects of stress on adjustment in medical student marriages. Journal of Marital
Brenlla, M. E., & Aranguren, M. (2010). Adaptacio
n Argentina de la Escala de Mal- and Family Therapy, 26(3), 341e351.
estar psicolo
gicode Kessler (K10). Revista de Psicología, 28(2), 309e340. Kenney, S. R., & LaBrie, J. W. (2013). Use of protective behavioral strategies and
Brienza, R. S., & Stein, M. D. (2002). Alcohol use disorders in primary care: Do reduced alcohol risk: Examining the moderating effects of mental health,
gender-specific differences exist? Journal of General Internal Medicine, 17(5), gender, and race. Psychology of Addictive Behaviors, 27(4), 997e1009.
387e397. Kessler, R. C., Andrews, G., Colpe, L. J., Hiripi, E., Mroczek, D. K., Normand, S. L., et al.
Buchmann, A. F., Schmid, B., Blomeyer, D., Becker, K., Treutlein, J., (2002). Short screening scales to monitor population prevalences and trends in
Zimmermann, U. S., et al. (2009). Impact of age at first drink on vulnerability to non-specific psychological distress. Psychological Medicine, 32(6), 959e976.
alcohol-related problems: Testing the marker hypothesis in a prospective study Keyes, K. M., Li, G., & Hasin, D. S. (2011). Birth cohort effects and gender differences
of young adults. Journal of Psychiatric Research, 43(15), 1205e1212. in alcohol epidemiology: A review and synthesis. Alcoholism: Clinical and
Chamberlain, S. R., & Sahakian, B. J. (2004). Cognition in mania and depression: Experimental Research, 35(12), 2101e2112.
Psychological models and clinical implications. Current Psychiatry Reports, 6(6), Krieger, H., Young, C. M., Anthenien, A. M., & Neighbors, C. (2018). The epidemiology
451e458. of binge drinking among college-age individuals in the United States. Alcohol
Conger, J. J. (1956). Alcoholism: Theory, problem and challenge. II. Reinforcement Research: Current Reviews, 39(1), 23e30.
theory and the dynamics of alcoholism. Quarterly Journal of Studies on Alcohol - Kuntsche, E., Rossow, I., Engels, R., & Kuntsche, S. (2016). Is 'age at first drink' a
Part A, 17(2), 296e305. useful concept in alcohol research and prevention? We doubt that. Addiction,
Cranford, J. A., Eisenberg, D., & Serras, A. M. (2009). Substance use behaviors, mental 111(6), 957e965.
health problems, and use of mental health services in a probability sample of Kuntsche, E., Stewart, S. H., & Cooper, M. L. (2008). How stable is the motive-alcohol
college students. Addictive Behaviors, 34(2), 134e145. use link? A cross-national validation of the drinking motives questionnaire
Instruments, & Computers Cronk, B. C., & West, J. L. (2002). Personality research on revised among adolescents from Switzerland, Canada, and the United States.
the internet: A comparison of web-based and traditional instruments in take- Journal of Studies on Alcohol and Drugs, 69(3), 388e396.
home and in-class settings. Behavior Research Methods, 34(2), 177e180. Journal LaBrie, J. W., Kenney, S. R., Lac, A., Garcia, J. A., & Ferraiolo, P. (2009). Mental and
Article. social health impacts the use of protective behavioral strategies in reducing
D'Amico, E. J., Tucker, J. S., Shih, R. A., & Miles, J. N. (2014). Does diversity matter? risky drinking and alcohol consequences. Journal of College Student Develop-
The need for longitudinal research on adolescent alcohol and drug use trajec- ment, 50(1), 35e49.
tories. Substance Use & Misuse, 49(8), 1069e1073. Lakens, D. (2013). Calculating and reporting effect sizes to facilitate cumulative
Doherty, R. W. (1997). The emotional contagion scale: A measure of individual science: A practical primer for t-tests and ANOVAs. Frontiers in Psychology, 4,
differences. [journal article]. Journal of Nonverbal Behavior, 21(2), 131e154. 863.
Droomers, M., Schrijvers, C. T., Stronks, K., van de Mheen, D., & Mackenbach, J. P. Lannoy, S., Billieux, J., Poncin, M., & Maurage, P. (2017). Binging at the campus:
(1999). Educational differences in excessive alcohol consumption: The role of Motivations and impulsivity influence binge drinking profiles in university
psychosocial and material stressors. Preventive Medicine, 29(1), 1e10. students. Psychiatry Research, 250, 146e154.
Eisenberg, M. E., Toumbourou, J. W., Catalano, R. F., & Hemphill, S. A. (2014). Social Lee, J. O., Hill, K. G., Guttmannova, K., Bailey, J. A., Hartigan, L. A., Hawkins, J. D., et al.
norms in the development of adolescent substance use: A longitudinal analysis (2012). The effects of general and alcohol-specific peer factors in adolescence
of the International youth development study. Journal of Youth and Adolescence, on trajectories of alcohol abuse disorder symptoms from 21 to 33 years. Drug
43(9), 1486e1497. and Alcohol Dependence, 121(3), 213e219.
Farmer, R. F., Seeley, J. R., Kosty, D. B., Gau, J. M., Duncan, S. C., Lynskey, M. T., et al. Lee, L. O., Young-Wolff, K. C., Kendler, K. S., & Prescott, C. A. (2012). The effects of age
(2015). Internalizing and externalizing psychopathology as predictors of at drinking onset and stressful life events on alcohol use in adulthood: A
cannabis use disorder onset during adolescence and early adulthood. Psychol- replication and extension using a population-based twin sample. Alcoholism:
ogy of Addictive Behaviors, 29(3), 541e551. Clinical and Experimental Research, 36(4), 693e704.
Fillmore, M. T., Carscadden, J. L., & Vogel-Sprott, M. (1998). Alcohol, cognitive Lizbona, A., & Rumeau, D. (2013). La educacio
n universitaria en el Uruguay: Con-
impairment and expectancies. Journal of Studies on Alcohol, 59(2), 174e179. tinuidades, cambios y desafíos (1960-2012). Informe del Instituto de Evaluacio
n
Goldsmith, A. A., Thompson, R. D., Black, J. J., Tran, G. Q., & Smith, J. P. (2012). Educativa del Uruguay.
Drinking refusal self-efficacy and tension-reduction alcohol expectancies Lui, P. P., & Zamboanga, B. L. (2018a). Acculturation and alcohol use among Asian
moderating the relationship between generalized anxiety and drinking be- Americans: A meta-analytic review. Psychology of Addictive Behaviors, 32(2),
haviors in young adult drinkers. Psychology of Addictive Behaviors, 26(1), 59e67. 173e186.
Gomez-Restrepo, C., Bohorquez, A., Pinto Masis, D., Gil Laverde, J. F., Rondon Lui, P. P., & Zamboanga, B. L. (2018b). A critical review and meta-analysis of the
Sepulveda, M., & Diaz-Granados, N. (2004). The prevalence of and factors associations between acculturation and alcohol use outcomes among hispanic
associated with depression in Colombia. Revista Panamericana de Salud Públic, Americans. Alcoholism: Clinical and Experimental Research, 42(10), 1841e1862.
16(6), 378e386. Markman Geisner, I., Larimer, M. E., & Neighbors, C. (2004). The relationship among
Gouveia, V. V., Rildesia, G. S. V., Valeschka, M. G., Santos, W. S., & Dio
genes de alcohol use, related problems, and symptoms of psychological distress: Gender
Medeiros, E. (2007). Midiendo contagio emocional: Adaptacio
n de la escala de as a moderator in a college sample. Addictive Behaviors, 29(5), 843e848.
Doherty. Revista de Psicología Social, 22(2), 99e112.
 P. Ruiz et al. / Alcohol 82 (2020) 91e101 101

Martens, M. P., Karakashian, M. A., Fleming, K. M., Fowler, R. M., Hatchett, E. S., & Pilatti, A., Caneto, F., Garimaldi, J. A., Vera Bdel, V., & Pautassi, R. M. (2014).
Cimini, M. D. (2009). Conscientiousness, protective behavioral strategies, and Contribution of time of drinking onset and family history of alcohol problems in
alcohol use: Testing for mediated effects. Journal of Drug Education, 39(3), alcohol and drug use behaviors in Argentinean college students. Alcohol and
273e287. Alcoholism, 49(2), 128e137.
McCabe, S. E., Veliz, P., & Schulenberg, J. E. (2018). How collegiate fraternity and Pilatti, A., Cupani, M., & Pautassi, R. M. (2015). Personality and alcohol expectancies
sorority involvement relates to substance use during young adulthood and discriminate alcohol consumption patterns in female college students. Alcohol
substance use disorders in early midlife: A national longitudinal study. Journal and Alcoholism, 50(4), 385e392.
of Adolescent Health, 62(3S), S35eS43. Pilatti, A., Read, J. P., & Caneto, F. (2016). Validation of the Spanish version of the
McCullough, L. D., de Vries, G. J., Miller, V. M., Becker, J. B., Sandberg, K., & young adult alcohol consequences questionnaire (S-YAACQ). Psychological
McCarthy, M. M. (2014). NIH initiative to balance sex of animals in preclinical Assessment, 28(5), e49ee61.
studies: Generative questions to guide policy, implementation, and metrics. Pilatti, A., Read, J. P., & Pautassi, R. M. (2017). ELSA 2016 cohort: Alcohol, tobacco,
Biology of Sex Differences, 5, 15. and marijuana use and their association with age of drug use onset, risk
McGorry, P. (2011). Transition to adulthood: The critical period for pre-emptive, perception, and social norms in argentinean college freshmen. Frontiers in
disease-modifying care for schizophrenia and related disorders. Schizophrenia Psychology, 8, 1452.
Bulletin, 37(3), 524e530. Pilatti, A., Read, J. P., Vera Bdel, V., Caneto, F., Garimaldi, J. A., & Kahler, C. W. (2014).
McGorry, P. D., Purcell, R., Goldstone, S., & Amminger, G. P. (2011). Age of onset and The Spanish version of the brief young adult alcohol consequences question-
timing of treatment for mental and substance use disorders: Implications for naire (B-YAACQ): A rasch model analysis. Addictive Behaviors, 39(5), 842e847.
preventive intervention strategies and models of care. Current Opinion in Psy- Read, J. P., Kahler, C. W., Strong, D. R., & Colder, C. R. (2006). Development and
chiatry, 24(4), 301e306. preliminary validation of the young adult alcohol consequences questionnaire.
McGue, M., & Iacono, W. G. (2008). The adolescent origins of substance use disor- Journal of Studies on Alcohol, 67(1), 169e177.
ders. International Journal of Methods in Psychiatric Research, 17(Suppl 1), Rothman, E. F., Edwards, E. M., Heeren, T., & Hingson, R. W. (2008). Adverse
S30eS38. childhood experiences predict earlier age of drinking onset: Results from a
Merrill, J. E., & Carey, K. B. (2016). Drinking over the lifespan: Focus on college ages. representative US sample of current or former drinkers. Pediatrics, 122(2),
Alcohol Research: Current Reviews, 38(1), 103e114. e298ee304.
Miech, R., Johnston, L., & O'Malley, P. M. (2017). Prevalence and attitudes regarding Rubio Valladolid, G., Bermejo Vicedo, J., Caballero Sanchez-Serrano &, M. C., &
marijuana use among adolescents over the past decade. Pediatrics, 140(6). Santo-Domingo Carrasco, J. (1998). Validacio
n de la Prueba para la Identi-
Montejano, G. R., Pilatti, A., Pitt, M. A., & Pautassi, R. M. (2019). Relacio
n entre la ficacio
n de Transtornos por Uso de Alcohol (AUDIT) en Atencio
n Primaria.
disponibilidad de alcohol, consumo de alcohol y problemas en jo
venes argen- Revista Clínica Espan ~ ola, 198(1), 11e14.
tinos. Health and Addicctions, 19(1), 36e46. Schulenberg, J. E., & Maggs, J. L. (2002). A developmental perspective on alcohol use
Mooney, D. K., Fromme, K., Kivlahan, D. R., & Marlatt, G. A. (1987). Correlates of and heavy drinking during adolescence and the transition to young adulthood.
alcohol consumption: Sex, age, and expectancies relate differentially to quantity Journal of Studies on Alcohol, Suppl(14), 54e70.
and frequency. Addictive Behaviors, 12(3), 235e240. Sturm, V. E., Yokoyama, J. S., Seeley, W. W., Kramer, J. H., Miller, B. L., & Rankin, K. P.
Moss, H. B., Chen, C. M., & Yi, H. Y. (2014). Early adolescent patterns of alcohol, (2013). Heightened emotional contagion in mild cognitive impairment and
cigarettes, and marijuana polysubstance use and young adult substance use Alzheimer's disease is associated with temporal lobe degeneration. Proceedings
outcomes in a nationally representative sample. Drug and Alcohol Dependence, of the National Academy of Sciences of the United States of America, 110(24),
136, 51e62. 9944e9949.
Organizacio
n para la Cooperacio
n y el Desarrollo Econo
mico. (2018). Education at a Subramaniam, M., Mahesh, M. V., Peh, C. X., Tan, J., Fauziana, R., Satghare, P., et al.
glance 2018. (2017). Hazardous alcohol use among patients with schizophrenia and
Osberg, T. M., Atkins, L., Buchholz, V., Shirsova, A., Swiatek, J., Whitely, J., et al. depression. Alcohol, 65, 63e69.
(2016). Development and validation of the college life alcohol salience scale: A Wild, B., Erb, M., & Bartels, M. (2001). Are emotions contagious? Evoked emotions
measure of beliefs about the role of alcohol in college life. Psychology of while viewing emotionally expressive faces: Quality, quantity, time course and
Addictive Behaviors, 24, 1e12. gender differences. Psychiatry Research, 102(2), 109e124.
Patrick, M. E., Schulenberg, J. E., & O'Malley, P. M. (2016). High school substance use Young, R. M., Oei, T. P., & Knight, R. G. (1990). The tension reduction hypothesis
as a predictor of college attendance, completion, and dropout: A national multi- revisited: An alcohol expectancy perspective. British Journal of Addiction, 85(1),
cohort longitudinal study. Youth & Society, 48(3), 425e447. 31e40.
Pearson, M. R. (2013). Use of alcohol protective behavioral strategies among college Ystrom, E., Kendler, K. S., & Reichborn-Kjennerud, T. (2014). Early age of alcohol
students: A critical review. Clinical Psychology Review, 33(8), 1025e1040. initiation is not the cause of alcohol use disorders in adulthood, but is a major
Pedrelli, P., Collado, A., Shapero, B. G., Brill, C., & MacPherson, L. (2016). Different indicator of genetic risk. A population-based twin study. Addiction, 109(11),
pathways explain alcohol-related problems in female and male college stu- 1824e1832.
dents. Journal of American College Health, 64(7), 535e544.