Table 1.

Findings of the main studies included in the present narrative review.

Author	Year	Type	Section	Main findings	Limitations
Beseler et al. (18)	2011	Prospective	1	Drinking motives accessed a decade early predicted greater alcohol use 1 and 16 weeks after 9 /11 in individuals from New Jersey county aged 18–65 years Drinking to cope with negative affect and drinking for enjoyment were the significant variables and no interactions with proximity to the fateful event and history alcohol dependence were noted	Alcohol use was not evaluated right before 9/11
DiMaggio et al. (19)	2009	Meta-analysis	1	An increase in alcohol consumption 2 years after the traumatic event was observed in this meta-analysis that included 31 population-based studies (the majority [24] of studies was from 9/11). These results suggest the need for public health interventions on alcohol use after massive trauma	Heterogeneity of the studies, and a small number of data points inserted in the meta-regression
Hirst et al. (20)	2018	Prospective	1	Findings showed that 1.5% was hospitalized for alcohol- (0.8%) or drug-related diagnosis Participants with PTSD were more likely to have been hospitalized for an alcohol- or drug-related condition than those without PTSD during a period of up to 9 years after 9/11	The study did not include data from federal, psychiatry hospitals out of NY, and emergency department visits
Welch et al. (21)	2014	Longitudinal	1	5–6 years after 9/11, 7.8% of participants reported frequent Binge Drinking (BD) (5+ drinks per occasion, 5+ times in the last 30 days) Higher odds of frequent BD were seen in individuals who were male, young (18–29 years old), never married, smokers, with high school diploma, an income of > 50 K, high exposure of the event and PTSD	The response rate of 68% on wave 2 Self-reported alcohol use and PTSD diagnosis were performed using a self-reported instrument
Welch et al. (22)	2017	Longitudinal	1	10 years after 9/11, 24.6% of the sample reported 1+ episode of Binge Drinking (BD) in the 30 days prior, to those ~37 with high intensity of BD (8+ drinks for men, 7+, women) Higher odds of BD were found in males, younger (18–34 years old), Caucasian, with an income <75 K, higher exposure of the event, and PTSD	Self-reported alcohol use and PTSD diagnosis were performed using a self-reported instrument
Alonso et al. (23)	2017	Longitudinal	2	Data from the National Institute of Statistics (INE, n = 21.9 million; 25–64 years) evaluated Deaths Directly Attributable to Alcohol (DDA) and employment status from 2002 to 2011 After the crisis, DDA increased among the employed and decreased among the unemployed, except for men, non-married, and medium/high-wealth people	Only a few DDA were analyzed Alcohol use variables were not available and it some individuals could have history of AUD before the crisis
Ásgeirsdóttir et al. (24)	2014	Longitudinal	2	A random sample of 9,028 individuals from the national survey Health and Well-being, conducted in 2007 and 2009 Reduction (−0.027) in drinking (5+ alcoholic drinks in 1 day at least 1 time/month [past year]) after the crisis among working age population (25–64 y.o.) even after including individual factors as covariates (hours of work, real income, financial assets, mortgage debt, or mental health)	Self-report alcohol use Working age population included a wide age range (25–64 y.o.)
Ásgeirsdóttir et al. (25)	2016	Longitudinal	2	Follow up of the above-mentioned study, conducted in 2012 Reduction of 5% a year in drinking (i.e., 5+ alcoholic drinks in 1 day at least 1 time/month in the past year) during the crisis and at a slower rate (2–3% a year) during recovery among the working-age population, controlling for individual factors (i.e., hours of work, real income, financial assets, mortgage debt, and mental health).	Self-report alcohol use Working age population included a wide age range (25–64 y.o.)
Bor et al. (26)	2013	Longitudinal	2	National survey with >2 million individuals conducted from 2006 to 2010 Frequent binge drinking (4+ episodes in the past 30 days) had a 7% increase, and was associated with young men (< 30 y.o.), not married, non-Black, higher household income, unemployed for <1 year, and without a college degree	Self-report alcohol use
de Goeij (27)	2015	Realistic review	2	Self-medication mechanism could explain a rise in heavy drinking in the US and Spain after the crisis, and that association was stronger in men. Budgetary shortfall could explain the fall in heavy drinking in Iceland	Less evidence for microeconomic (individual) factors
de Goeij (28)	2016	Longitudinal	2	Dutch Health Interview Survey conducted between 2004 and 2013 (n = 20, 140 men and 22,394 women aged 25–64) evaluated month-to-month trends in alcohol consumption over several years (episodic [6+ glasses on 1 day 1+ day/week] and chronic drinking [>14 glasses/week for women and >21 for men]) Downward trends showed a ceasing of decline among women in general and among 35–64 and high-income men. A start of decline was observed among younger men (25–34 y.o.)	Self-report alcohol use Harmful drinking was not measured longitudinally (repeated cross-sectional data), and causal relationship cannot be inferred
de Goeij et al. (29)	2017	Longitudinal	2	Data from Dutch Health Interview Survey (N = 26,355 aged 30–64 years) collected from 2004 to 2013 Job loss duration (>6 months) was related to both episodic [OR 1.40 (95% CI 1.01–1.94)] and chronic drinking (OR 1.42 [95% CI 1.05–1.91]). Current job loss was associated with chronic drinking (OR 1.43 [95% CI 1.03–1.98]) during the post 2008 economic crisis, but not before. These associations were most clear in men and different between pre-crisis and crisis period (p interaction = 0.023 and 0.035, respectively)	Self-report alcohol use Harmful drinking was not measured longitudinally (repeated cross-sectional data), and causal relationship cannot be inferred
Gili et al. (30)	2013	Longitudinal	2	Primary care patients (N = 7,940 in 2006–07 and N = 5,876 in 2010–11) were evaluated for mental health disorders AUD diagnosis increased 4.6% (dependence) and 2.4% (abuse) after the crisis. Relative increase, in comparison to other psychiatric disorders, were greater for alcohol dependence and abuse (OR= 12.2 and 4.6, respectively)	Individuals were not evaluated longitudinally
Global Burden of Disease (31)	2016	Longitudinal	2	Between 2000 and 2016 there was a 2% increase in YLD Alcohol was the second behavioral risk factor for YLD and	Inconsistencies in registry data
Kalousova et al. (32)	2014	Longitudinal	2	Data from Michigan Recession and Recovery Study (N = 840, followed from 2009–10 to 2011) using Alcohol Use Disorder Identification Test (AUDIT) Harmful drinking was associated with perceived loss of economic resources (HR: 2.75 [95% CI 1.2–6.27] p < 0.05), whereas objective measures did not predict this outcome	Data is not nationally representative Objective measures of economic resources were self-reported
Kaplan (33)	2016	Retrospective	2	Data from the U.S. National Violent Death Reporting System (NVDRS) was used to compare heavy drinking among men who committed suicide and living men (Blood alcohol levels [BAC] ≥ 0.08 g/dl for suicide decedents; at least one binge drinking in the last 30 days for the control group) Men who committed suicide had a greater increase (8%) in heavy drinking at the onset of the recession in comparison to living men. For men, adjusted odds ratio was higher after the crisis (adOR =1.15 [95% CI 1.10–1.20; p < 0.001]) relative to the prerecession period (adOR = 0.93 [95% CI 0.90–0.97]). The same pattern was not observed in women	BAC measures do not indicate a diagnosis of AUD or harmful drinking Postmortem toxicology testing rates varied across states
Mateo-Urdiales (34)	2020	Longitudinal	2	Data from Spanish Longitudinal Mortality Study (Census) collected from 2004 to 2011 evaluated DDA in a sample of 22.2 million people Largest increase in DDA in men and women with tertiary studies (+ 25.3% and +113.8%, respectively) and smallest in those with primary studies (+6.2% and +1.5%), decreasing relative educational inequalities	Only a few DDA were analyzed Causal relationship cannot be inferred (repetitive cross-sectional data)
Yang (35)	2018	Longitudinal	2	Data from The National Survey on Drug Use and Health (N = 307,935) from 2007 to 2016 Millennials were at significantly increased risk of past month binge alcohol (AOR = 1.51; 95% CI = 1.46 ± 1.56) than Gen X, while Baby Boomers were at significantly reduced risk of all substances (AOR = 0.56; 95% CI = 0.54 ± 0.58)	Self-report alcohol use Causal relationship cannot be inferred (repetitive cross-sectional data)
Lancee (36)	2008	Retrospective	3	It was not reported any increase in alcohol intake since the SARS outbreak in Hospital Workers in Canada	Self-report alcohol use
Mak et al. (37)	2009	Retrospective	3	One-third of the sample had psychiatric disorders 30 months after SARS, the most prevalent disorders in this sample were depressive and anxiety disorders, including PTSD. However, the new incidence of AUD was not observed in this infected and hospitalized patients who survived	Small sample size Self-reported questionnaires
Phua et al. (38)	2005	Cross-sectional	3	The use of alcohol and drugs was not observed as a coping mechanism in Healthcare Workers (HCW) in Singapore. Authors stated that cultural e religious factors could contribute to that finding	Small sample size
Wu et al. (39)	2008	Cross-sectional	3	Increased AUD symptoms in hospital employees were related to being male, in quarantining, having a higher household income, working at high-risk locations, high PTS symptoms and depression, hyper-arousal, and drinking to cope, 3 years after the SARS outbreak in China.	It is not possible to determine whether AUD symptoms started before or after the SARS outbreak
Ammar (40)	2020	Cross-sectional	4	An online survey performed in different regions (Europe, Africa, Asia, and Americas) involving 35 institutions showed a decrease in binge drinking during quarantine (p < 0.001, d = 0.58) comparing data from 2019 and 2020	Lack of inclusion and exclusion criteria Data from a convenience sample recruited online
Ahmed et al. (41)	2020	Cross-sectional	4	In the overall sample, about one third affirmed the occurrence of anxiety symptoms and 37.1% of depressive symptoms Additionally, 29.1% of the participants reported hazardous drinking, 9.5% harmful drinking, and 1.6% alcohol dependency. Individuals ages 21–40 were more vulnerable to alcohol use Hubei had significantly higher proportions of hazardous drinking (33.5% in Hubei and 21.5% in others); harmful drinking (11.1 vs. 1.9%) and alcohol dependence (6.8% vs. 1.0%).	Self-report scales More than 50% of the sample was from Wuhan province
Fiocruz (42)	2020	Cross-sectional	4	In Brazil, participants were selected using a Respondent-Driven Sampling (RDS) method from April 24th and May 8th. There was a subject perception of increased use of alcoholic beverages by 18% of the respondents, individuals from 30 to 39 years old showed a higher increase. Alcohol intake was associated with feeling sad/depressed (reaching 46.9% of the participants who reported feeling sometimes [22.5%] or always [24.4%] sad/depressed during the pandemic)	Self-report alcohol use Subjective perception of an increase in consumption Data from a convenience sample recruited online
Lee et al. (43)	2020	Cross-sectional	4	Examining the validation of the Obsession COVID-19 Scale (OCS) in the U.S. population, proposing a cutoff point of 7. Findings also showed that higher scores of OCS were correlated with alcohol and drug use to cope	Does not quantify alcohol use
Liang et al. (44)	2020	Cross-sectional	4	Two weeks after the outbreak of the COVID-19 in China, 40.4% of the sample was prone to psychological problems, and 14.4% PTSD symptoms. Among these young adults (age 14–35), those with more negative coping strategies (including alcohol use) had a higher chance of having psychological problems	Self-report Snowball sampling approach
Nanos Research (45)	2020	Cross-sectional	4	Participants were selected using a Random Digit Dialed in April and May. In this Canadian study, among individuals who affirmed staying home more due to COVID-19 (90% of the sample), 20% of those reported increased alcohol, and 21% said they were drinking more often as well. The main reasons for drinking more: lack of a regular schedule, boredom, and stress	Non-standard questionnaire Descriptive analysis
Newby et al. (46)	2020	Cross-sectional	4	Participants filled out an online questionnaire through (March 27th and April 7th), respondents were mostly females (86%). About three-quarters of the subjects said that their mental health was worse. Levels of distress, anxiety, and fears were higher in the respondents with a mental health diagnosis 52.7% declared a hazardous pattern of alcohol use in the prior month (scores ≤ 3/women; ≤ 4 men in AUDIT-C)	Self-report
Stanton et al. (47)	2020	Cross-sectional	4	An online survey carried out between April 9–19, participants were on average 50 years old (SD 14.9) and 67% were women. 22.3% of the respondents affirmed using alcohol 4+ occasions/week, and 26.6% said there was an increase in alcohol use Higher anxiety, depression, and stress levels were noted in individuals aged 18–45 and were related to more elevated alcohol use	Self-report
Sidor and Rzymski (48)	2020	Cross-sectional	4	An online study conducted between 17 April and 1 May (period of national quarantine) in Poland observed an increase of 14.6% in alcohol use. Additionally, individuals who recognized themselves with an AUD reported more frequent alcohol use	Self-report
Sun et al. (49)	2020	Cross-sectional	4	An online survey carried out from March 24–31 observed relapses and an increase in alcohol use during COVID-19 in China. Respondents were on average 28 years old (SD 9) and the distribution of males and females was similar. Results revealed that 32.1% of regular drinkers increased alcohol intake, 18.7% ex-drinkers relapsed, and 1.7% non-drinkers initiated the use of alcohol	Data from a convenience sample recruited online Self-report
Zhang et al. (50)	2020	Cross-sectional	4	The study was conducted between February 28 and March 02 in Wuhan. Participants were submitted to the Perceived Stress Scale (PSS), daily routine, and habits. Results showed that more than 80% reported elevated perceived stress levels. Also, females who were regular alcohol drinkers had more elevated perceived stress levels	Self-report

1: Terrorist attacks; 2: Economic crises; 3: SARS; 4 Covid-19.