Factors associated with alcohol and tobacco consumption as a coping strategy to deal with the coronavirus disease (COVID-19) pandemic and lockdown in Spain.

AIM: To provide a population-based characterization of sociodemographic and clinical risk and protective factors associated with consumption of alcohol, tobacco, or both as a coping strategy in a sample of the Spanish general population during the early phase of the COVID-19 pandemic.
METHODS: Cross-sectional study based on an online snowball recruiting questionnaire. The survey consisted of an ad hoc questionnaire comprising clinical and sociodemographic information and the Spanish versions of the Depression, Anxiety, and Stress Scale (DASS-21) and the Impact of Event Scale (IES).
RESULTS: The final sample included 21,207 individuals [mean age (SD) = 39.7 (14.0); females: 14,768 (69.6%)]. Up to 2867 (13.5%) of participants reported using alcohol, 2545 (12%) tobacco and 1384 (6.5%) both substances as a strategy to cope with the pandemic. Sex-related factors were associated with alcohol consumption as a coping strategy [female, OR = 0.600, p < 0.001]. However, education level, work status, and income played different roles depending on the substance used to cope. Having a current mental disorder was associated only with tobacco consumption as a coping strategy [OR = 1.391, p < 0.001]. Finally, sex differences were also identified.
CONCLUSIONS: Sociodemographic, clinical, and psychological factors were associated with consumption of alcohol, tobacco, or both as a coping method for the COVID-19 pandemic and lockdown. Our findings may help develop specific intervention programs reflecting sex differences, which could minimize negative long-term outcomes of substance use after this pandemic.

Introduction

The coronavirus disease of 2019 (COVID-19) had its first epicentre in Wuhan, China and has since spread worldwide (Wang, Horby, Hayden & Gao, 2020a). The WHO Emergency Committee declared a global health emergency based on growing case notification rates. Spain has been one of most affected European countries. To prevent over-burdened health systems from collapsing, on 13 March, the government announced an official lockdown.

During disaster situations, some authors suggest that substance use could be a dysfunctional coping strategy in vulnerable populations (Hogarth, Martin & Seedat, 2019). Evidence from the severe acute respiratory syndrome (SARS) outbreak showed that people used alcohol as a coping method (Wu et al., 2008). In addition, three years later, symptoms of alcohol abuse/dependence were positively associated with having been quarantined or having worked in high-risk locations (Wu et al., 2008).

Sociodemographic factors such as income were associated with substance use as a coping method in previous pandemics (Wu et al., 2008). Education level, sex, and employment status have also been associated with substance use as a coping strategy (Ilhan et al., 2016, Stapinski et al., 2016). Recent studies already show that the COVID-19 pandemic has had negative effects on alcohol and tobacco consumption, which is increasing among university students (Romero-Blanco et al., 2020) and the general population (Hawke et al., 2020, Malta et al., 2020, Rodriguez et al., 2020, Stanton et al., 2020, Sun et al., 2020, Vanderbruggen et al., 2020). Furthermore, evidence reflects differences between the sexes: males consumed more alcohol early in the pandemic, while females were twice as likely to have stress and anxiety-related problems (Ahmed et al., 2020).

Little is known about factors that may affect the use of alcohol, tobacco, or both to cope with the pandemic and any differences by sex. Therefore, the main objectives of this study are to provide a characterization of sociodemographic and clinical risk and protective factors associated with consumption as a coping strategy in a sample of the Spanish general population during the early phase of the pandemic and lockdown.

Method

Sample

This study is a secondary analysis of a larger cross-sectional exploratory online survey designed to determine the psychological impact of the pandemic and lockdown in a sample of people aged 18 years or over living in Spain (see García-Álvarez et al., 2020). Between 19 and 26 March 2020, an anonymous questionnaire was conducted through social networks and email using a virtual snowball sampling method. Furthermore, to ensure that the same person did not answer the survey twice, before analysing the data, all data were checked based on date of birth and sex.

The total sample consisted of 21,207 individuals [mean age (SD) = 39.7 (14.0); females: 14,768 (69.6%)]. The only exclusion criterion was not providing online informed consent. In the sample, the populations of the Principality of Asturias (36.2%) and Cantabria (11.9%) were over-represented. Catalonia (4.5%) and the Valencian Community (4.3%) were under-represented, but the proportion of respondents from Madrid (10.0%) (the most affected region) was significant (García-Álvarez et al., 2020).

The study was conducted according to the Declaration of Helsinki (World Medical Association General Assembly, 2013). The Clinical Research Ethics Committee of Hospital Universitario Central de Asturias in Oviedo approved the study protocol (Ref. 2020.162) on 16 March, and online informed consent was obtained from all participants before enrolment.

Assessment

The survey consisted of an ad hoc questionnaire in which participants had to choose the response that best reflected their situation, including sociodemographic information and methods used to cope with the lockdown. Specifically, the following yes/no questions explored alcohol and/or tobacco consumption as a coping strategy: “Does alcohol consumption help you cope with the COVID-19 pandemic?” and “Does tobacco consumption help you cope with COVID-19 pandemic?” Information on physical health, COVID-19 variables (testing, symptoms, number of relatives infected and relationship to them), and past and present psychiatric history was also recorded. Finally, the Spanish versions of the Depression, Anxiety, and Stress Scale (DASS-21) (Bados et al., 2005) and the Impact of Event Scale (IES) (Baguena et al., 2001) were included.

In short, the DASS-21 and IES were used to measure the early psychological correlates associated with the pandemic and lockdown (last seven days). The DASS-21 is a self-rated scale to assess symptoms of depression (items 3, 5, 10, 13, 16, 17, 21), anxiety (2, 4, 7, 9, 15, 19, 20), and stress (1, 6, 8, 11, 12, 14, 18) over the past week. It provides scores for each of these three subscales (range 0–7). The IES is a 15-item self-report scale assessing subjective distress related to a specific event. It provides a total score and scores for two subscales, intrusion and avoidance. Total scores and subscores were used for the data analysis. Dichotomous score variables (“not a case”/“a probable case”) for the five DASS-21 and IES subscales were analysed (see García-Álvarez et al., 2020).

Statistical analysis

Data were analysed using IBM SPSS Version 24.0 (IBM Corp., Armonk, NY, 2016). They are presented as mean (standard deviation [SD]) for numeric variables and as frequencies and percentages for categorical variables. Participants were classified into four groups based on their negative (No substance group) or positive answer to one of the questions “Does alcohol consumption help you cope with the COVID-19 pandemic?” (Alcohol group), “Does tobacco consumption help you cope with COVID-19 pandemic? (Tobacco group), or to both questions (Alcohol & Tobacco group). To identify between-group differences, we employed the Chi-square test or ANOVA depending on the type of variables. Nine logistic regression models (forward stepwise selection) were estimated to determine the independent factors associated with consumption of alcohol, tobacco, or both as a coping strategy for the whole sample and by sex. To avoid inflation of Type I error, the statistical significance level was set at α < 0.001.

Results

The final sample consisted of 21,207 individuals [mean age (SD) = 39.7 (14.0); females: 14,768 (69.6%)]. Sociodemographic and clinical characteristics of the sample are shown in Table 1, Table 2 , respectively. Note that 66.5% reported no substance use as a coping strategy, 13.5% reported alcohol consumption, 12.0% reported tobacco consumption, and 6.5% reported using both as strategies to cope with the pandemic. Results by sex are shown in Supplementary Tables 1, 2, 5, and 6.

Table 1

Sociodemographic characteristics for the whole sample and according to substance use by participants as a coping strategy.

	Total sampleN = 21207	No substanceN=14411	AlcoholN =2867	TobaccoN = 2545	Alcohol & TobaccoN=1384	Statistical test, P
Sociodemographic variables
Age [Mean (SD)]	39.7 (14.0)	39.6 (14.5)	40.4 (13.1)	40.8 (13.1)	37.2 (12.0)	21.669a, <0.001
Sex, female [n (%)]	14768 (69.6)	10258 (71.2)	1750 (61.0)	1839 (72.3)	921 (66.5)	131.024b, <0.001
Marital status [n (%)]						127.925b, <0.001
Never married	9867 (46.5)	6695 (46.5)	1261 (44.4)	1159 (45.5)	752 (54.3)
Married/Living as married	9630 (45.4)	6677 (46.3)	1401 (48.9)	1056 (41.5)	496 (35.8)
Separated/Divorced/Widowed	1710 (8.1)	1039 (7.2)	205 (7.2)	330 (13.0)	136 (9.8)
Education level [n (%)]						166.642b, <0.001
Primary	333 (1.6)	207 (1.4)	26 (0.9)	75 (2.9)	25 (1.8)
Secondary	7688 (36.3)	5062 (35.1)	769 (26.8)	1289 (50.6)	568 (41.0)
University	13186 (62.2)	9142 (63.4)	2072 (72.3)	1181 (46.4)	791 (57.2)
Work status [n (%)]						479.927b, <0.001
Unemployed	1829 (8.6)	1132 (7.9)	207 (7.2)	336 (13.2)	154 (11.1)
Working
Employed	7679 (36.2)	4909 (34.1)	1214 (42.3)	958 (37.6)	598 (43.2)
Self-employed	2048 (9.7)	1250 (8.7)	345 (12.0)	277 (10.9)	176 (12.7)
Civil servant	4099 (19.3)	2925 (20.3)	600 (20.9)	338 (15.2)	186 (13.4)
Retired	1312 (6.2)	997 (6.9)	145 (5.1)	131 (5.1)	39 (2.8)
Student/Homemaker	3392 (16.0)	2622 (18.2)	278 (9.7)	12.5 (9.3)	175 (12.6)
Other	848 (4.0)	576 (4.0)	78 (2.7)	138 (5.4)	56 (4.0)
Income (€) [n (%)]						476.536b, <0.001
No income	3349 (15.8)	2576 (17.9)	273 (9.5)	350 (13.8)	150 (10.8)
Less than 500	1462 (6.9)	970 (6.7)	139 (4.8)	226 (8.9)	127 (9.2)
500–999	2667 (12.6)	1667 (11.6)	309 (10.8)	455 (17.9)	236 (17.1)
1000–1499	4201 (19.8)	2724 (18.9)	572 (20.0)	583 (22.9)	322 (23.3)
1500–1999	3799 (17.9)	2568 (17.8)	600 (20.9)	411 (16.1)	220 (15.9)
More than 1999	4404 (20.8)	2982 (20.7)	816 (28.5)	355 (13.9)	251 (18.1)
Prefer not to answer	1325 (6.2)	924 (6.4)	158 (5.5)	165 (6.5)	78 (5.6)
Change in work status due to COVID-19 [n (%)]
No	17764 (84.7)	12285 (86.1)	2379 (83.8)	2029 (80.7)	1071 (78.8)	100.502b, <0.001
ETLA/EPLO*	1871 (8.9)	1135 (8.0)	265 (9.3)	302 (12.0)	169 (12.4)
Dismissal	390 (1.9)	231 (1.6)	62 (2.2)	61 (2.4)	36 (2.6)
Furlough	954 (4.5)	615 (4.3)	132 (4.7)	123 (4.9)	84 (6.2)
Change in income due to COVID-19 [n (%)]						132.490b, <0.001
No	15677 (73.9)	10964 (76.1)	2048 (71.4)	1733 (68.1)	932 (67.3)
Reduction, up to 25%	2292 (10.8)	1444 (10.0)	348 (12.1)	329 (12.9)	171 (12.4)
Reduction, 26-50%	1367 (6.4)	847 (5.9)	215 (7.5)	188 (7.4)	117 (8.5)
Reduction, 51-100%	1738 (8.2)	1066 (7.4)	238 (8.3)	279 (11.0)	155 (11.2)
Increase	133 (0.6)	90 (0.6)	18 (0.6)	16 (0.6)	9 (0.7)	191.272b, <0.001
Living situation [n (%)]	2580 (12.2)	1596 (11.1)	342 (11.9)	399 (15.7)	243 (17.6)
Alone	7534 (35.5)	4881 (33.9)	1163 (40.6)	921 (36.2)	569 (41.1)
Two people	10722 (50.6)	7665 (53.2)	1317 (45.9)	1185 (46.6)	555 (40.1)
Three to five	371 (1.7)	269 (1.9)	45 (1.6)	40 (1.6)	17 (1.2)
More than five						72.494b, <0.001
Dependent children [n (%)]	14207 (67.0)	9626 (66.8)	1869 (65.2)	1696 (66.6)	1016 (73.4)
None	3357 (15.8)	2221 (15.4)	459 (16.0)	490 (19.3)	187 (13.5)
One	3050 (14.4)	2135 (14.8)	460 (16.0)	299 (11.7)	156 (11.3)
Two	593 (2.8)	429 (3.0)	79 (2.8)	60 (2.4)	60 (1.8)
More than two						61.935b, <0.001
Elderly dependents [n (%)]	19203 (90.6)	13084 (90.8)	2647 (92.3)	2213 (87.0)	1259 (91.0)
None	1379 (6.5)	900 (6.2)	150 (5.2)	246 (9.7)	83 (6.0)
One	521 (2.5)	358 (2.5)	52 (1.8)	75 (2.9)	36 (2.6)
Two	104 (0.5)	69 (0.5)	18 (0.6)	11 (0.4)	6 (0.4)
More than two						108.003b, <0.001
Able to enjoy free time [n (%)]
No	1605 (7.6)	1029 (7.1)	146 (5.1)	309 (12.2)	121 (8.7)
Yes	19571 (92.4)	13367 (92.9)	2713 (94.9)	2229 (87.8)	1262 (91.3)
March survey response day [n (%)]						132.991b, <0.001
19	5763 (27.2)	4010 (27.8)	752 (26.2)	665 (26.1)	336 (24.3)
20	3735 (17.6)	2577 (17.9)	531 (18.5)	403 (15.8)	224 (16.2)
21	1640 (7.7)	1126 (7.8)	253 (8.8)	169 (6.6)	92 (6.6)
22	1432 (6.8)	871 (6.0)	234 (8.2)	196 (7.7)	131 (9.5)
23	1804 (8.5)	1146 (8.0)	217 (7.6)	272 (10.7)	169 (12.2)
24	635 (3.0)	398 (2.8)	85 (3.0)	100 (3.9)	52 (3.8)
25	1203 (5.7)	807 (5.6)	176 (6.1)	138 (5.4)	82 (5.9)
26	4995 (23.6)	3476 (24.1)	619 (21.6)	602 (23.7)	298 (21.5)

ANOVA.

Chi-square test. SD: standard deviation.

ETLA: Employee Temporary Lay Off. EPLO: Employee Permanent Lay Off.

Table 2

Clinical characteristics for the whole sample and according to substance use by participants as a coping strategy.

	Total sampleN = 21207	No substanceN=14411	AlcoholN =2867	TobaccoN = 2545	Alcohol & TobaccoN=1384	Statistical test, P
Coping strategies [n (%)]
Exercise, yes	12084 (57.0)	8360 (58.0)	1927 (67.2)	1052 (41.3)	745 (538)	388.428a, <0.001
Watching TV*, yes	18932 (89.3)	12743 (88.4)	2676 (93.3)	2255 (88.6)	1258 (90.9)	65.268a, <0.001
Reading COVID-19 news, yes	14252 (67.2)	9366 (65.0)	2152 (75.1)	1741 (68.4)	993 (71.7)	126.940a, <0.001
Art**, yes	18333 (86.4)	12439 (86.3)	2521 (87.9)	2143 (84.2)	1230 (88.9)	23.483a, <0.001
Cooking, yes	15036 (70.9)	9905 (68.7)	2232 (77.9)	1796 (70.6)	1103 (79.7)	152.016a, <0.001
Social media, yes	19424 (91.6)	13107 (91.0)	2699 (94.1)	2310 (90.8)	1308 (94.5)	49.398a, <0.001
Working, yes	12561 (59.2)	8498 (59.0)	1949 (68.0)	1298 (51.0)	816 (59.0)	162.709a, <0.001
Yoga/meditation, yes	4685 (22.1)	3222 (22.4)	724 (25.3)	424 (25.3)	315 (22.8)	61.223a, <0.001
Personal history of mental disorder						171.953a, <0.001
No lifetime mental disorder (NMD)	15053 (71.0)	10403 (72.2)	2142 (74.7)	1600 (62.9)	908 (65.6)
Past mental disorder (PMD)	2489 (11.7)	1585 (11.0)	253 (8.8)	458 (18.0)	193 (13.9)
Current mental disorder (CMD)	3665 (17.3)	2423 (16.8)	472 (16.5)	487 (19.1)	283 (20.4)
Physical disease and COVID-19 variables
Current physical disease*** [n (%)]						22.596a, <0.001
No	14017 (71.8)	9588 (71.9)	1962 (72.6)	1548 (67.3)	919 (71.4)
Yes	5514 (28.2)	3747 (28.1)	742 (27.4)	753 (32.7)	368 (28.6)
Days with COVID-19 symptoms
[n (%)]						27.605a, 0.006
None	18761 (88.5)	12735 (88.4)	2533 (88.4)	2289 (89.9)	1204 (87.0)
One to two days	1143 (5.4)	750 (5.2)	169 (5.9)	125 (4.9)	99 (7.2)
Three to five	600 (2.8)	439 (3.0)	72 (2.5)	58 (2.3)	31 (2.2)
Six to fourteen	559 (2.6)	389 (2.7)	79 (2.8)	51 (2.0)	40 (2.9)
More than fourteen	144 (0.7)	98 (0.7)	14 (0.5)	22 (0.9)	10 (0.7)
Taken COVID-19 test [n (%)]						23.291a, 0.006
No	20894 (98.6)	14173 (98.4)	2826 (98.6)	2519 (99.0)	1376 (99.5)
Yes, negative results	180 (0.8)	130 (0.9)	29 (1.0)	19 (0.7)	2 (0.1)
Yes, positive results	64 (0.3)	55 (0.4)	4 (0.1)	2 (0.1)	3 (0.2)
Yes, waiting for results	59 (0.3)	46 (0.3)	7 (0.2)	4 (0.2)	2 (0.1)
Family/Friends infected by COVID-19 [n (%)]						28.744a, 0.001
No	16669 (78.7)	11325 (78.7)	2216 (77.4)	2068 (81.4)	1060 (76.7)
One	2181 (10.3)	1510 (10.5)	305 (10.7)	221 (8.7)	145 (10.5)
Two	1184 (5·6)	780 (5.4)	172 (6.0)	125 (4.9)	107 (7.7)
More than two	1137 (5·4)	773 (5.4)	169 (5.9)	125 (4.9)	125 (4.9)
Living with people infected with COVID-19 [n (%)]						9.222a, 0.417
No	20848 (98.3)	14170 (98.3)	2818 (98.3)	2500 (98.2)	1360 (98.3)
One	251 (1.2)	177 (1.2)	31 (1.1)	27 (1.1)	16 (1.2)
Two	46 (0.2)	30 (0.2)	9 (0.3)	5 (0.2)	2 (0.1)
More than two	62 (0.3)	34 (0.2)	9 (0.3)	13 (0.5)	6 (0.4)

Chi-square test. SD: standard deviation.

Watching TV includes films and series.

Art: Drawing, painting, writing, reading and/or listening to music.

Physical disease includes: hypertension, diabetes, cardiovascular diseases, respiratory diseases (asthma, COPD, etc.), and cancer.

Psychological impact according to substance (alcohol, tobacco, or both) used as a coping strategy

The psychological impact of the pandemic and lockdown in the total sample and in those who reported using alcohol, tobacco, or both as a coping strategy are shown in Table 3 . The impact is shown by sex in Supplementary Tables 3 and 7.

Table 3

Psychological impact of the COVID-19 pandemic and lockdown for the whole sample and according to substance use by participants as a coping strategy.

	Total sampleN = 21207	No substanceN=14411	AlcoholN =2867	TobaccoN = 2545	Alcohol & TobaccoN=1384	Statistical test, P
DASS-21 subscales [Mean (SD)]
Depression	3.6 (1.1)	3.6 (1.1)	3.6 (1.1)	3.7 (1.1)	3.9 (1.2)	38.720a, <0.001
Anxiety	1.2 (1.6)	1.1 (1.6)	1.1 (1.5)	1.4 (1.8)	1.4 (1.7)	42.072a, <0.001
Stress	2.4 (2.4)	2.3 (2.3)	2.5 (2.4)	2.6 (2.4)	2.8 (2.4)	30.078a, <0.001
DASS-21 subscales [n (%)]
Depression						144.757b, <0.001
No	442 (2.1)	319 (2.2)	65 (2.3)	36 (1.4)	442 (2.1)
Doubtful	10852 (51.2)	7579 (52.6)	1494 (52.1)	1174 (46.1)	10852 (51.2)
Mild	5940 (28.0)	3954 (27.4)	803 (28.0)	793 (31.2)	5940 (28.0)
Moderate	2655 (12.5)	1728 (12.0)	340 (11.9)	377 (14.8)	210 (15.2)
Severe	1003 (4.7)	638 (4.4)	124 (4.3)	129 (5.1)	112 (8.1)
Extremely severe	315 (1.5)	193 (1.3)	41 (1.4)	36 (1.4)	45 (3.3)
Depression						83.858b, <0.001
No	11294 (53.3)	7898 (54.8)	1559 (54.4)	1210 (47.5)	627 (45.3)
Yes	9913 (46.7)	6513 (45.2)	1308 (45.6)	1335 (52.5)	757 (54.7)
Anxiety						135.323b, <0.001
No	14825 (69.9)	10267 (71.2)	2066 (72.1)	1626 (63.9)	866 (62.6)
Doubtful	4111 (19.4)	2724 (18.9)	532 (18.6)	533 (20.9)	322 (23.3)
Mild	970 (4.6)	600 (4.2)	129 (4.5)	161 (6.3)	80 (5.8)
Moderate	672 (3.2)	406 (2.8)	81 (2.8)	115 (4.5)	70 (5.1)
Severe	384 (1.8)	252 (1.7)	39 (1.4)	65 (2.6)	28 (2.0)
Extremely severe	245 (1.2)	162 (1.1)	162 (1.1)	45 (1.8)	18 (1.3)
Anxiety						86.455b, <0.001
No	18936 (89.3)	12991 (90.1)	2598 (90.6)	2159 (84.8)	1188 (85.8)
Yes	2271 (10.7)	1420 (9.9)	269 (9.4)	386 (15.2)	196 (14.2)
Stress						108.294b, <0.001
No	9842 (46.4)	6951 (48.2)	1276 (44.5)	1088 (42.8)	527 (38.1)
Doubtful	4314 (20.3)	2867 (19.9)	633 (22.1)	503 (19.8)	311 (22.5)
Mild	1907 (9.0)	1269 (8.8)	271 (9.5)	246 (9.7)	121 (8.7)
Moderate	1814 (8.6)	1205 (8.4)	237 (8.3)	220 (8.6)	152 (11.0)
Severe	1827 (8.6)	1168 (8.1)	240 (8.4)	277 (10.9)	142 (10.3)
Extremely severe	1503 (7.1)	951 (6.6)	210 (7.3)	211 (8.3)	131 (9.5)
Stress						56.921b, <0.001
No	14156 (66.8)	9818 (68.1)	1909 (66.6)	1591 (62.5)	838 (60.5)
Yes	7051 (33.2)	4593 (31.9)	958 (33.4)	954 (37.5)	546 (39.5)
IES subscales [Mean (SD)]
Intrusion	2.12 (1.9)	2.0 (1.9)	2.1 (1.9)	2.4 (1.9)	2.5 (2.0)	43.783a, <0.001
Avoidance	3.29 (2.0)	3.2 (2.0)	3.1 (1.9)	3.5 (2.0)	3.5 (2.0)	26.746a, <0.001
Total IES	5.41 (3.4)	5.3 (3.4)	5.2 (3.3)	5.9 (3.4)	6.0 (3.5)	42.877a, <0.001
Intrusion						85.336, <0.001
No	16208 (76.4)	11227 (77.9)	2197 (76.6)	1808 (71.0)	976 (70.5)
Yes	4999 (23.6)	3184 (22.1)	1808 (23.4)	737 (29.0)	408 (29.5)
Avoidance						61.684b, <0.001
No	11806 (55.7)	8115 (56.3)	1706 (59.5)	1285 (50.5)	700 (50.6)
Yes	9401 (44.3)	6296 (43.7)	1161 (40.5)	1260 (49.5)	684 (49.4)

ANOVA;

Chi-square test; SD: standard deviation DASS-21: Depression, Anxiety and Stress Scale (No: includes No and Doubtful; Yes: includes Mild, Moderate, Severe, and Extremely Severe); IES: Impact of Event Scale

Note that 54.4%, 47.5%, and 45.3% of the sample using alcohol, tobacco, or both, respectively, as a coping method, could be considered to have depressive responses on the DASS-21. On the IES, the avoidant coping style was the most prevalent. Depressive response and avoidant style were also the most prevalent in both sexes. However, these prevalence’s were higher in females.

Factors related to alcohol, tobacco, or both as a coping strategy during the pandemic and lockdown

Alcohol consumption as a coping strategy

Logistic regression models were developed to identify variables (sociodemographic, clinical, and psychological) associated with alcohol consumption as a coping strategy (see Table 4 ).

Table 4

Variables associated with alcohol or tobacco use to cope with COVID-19 during the pandemic and lockdown.

	β	SE	Wald	df	p	OR	95% CI
Alcohol
Intersection	-4.147	0.381	118.237	1	<0.001	0.016
Sex (female)	-0.511	0.049	107.900	1	<0.001	0.6001	0.545-0.661
Income (€), reference: No income
More than 1999	0.582	0.129	20.197	1	<0.001	1.7892	1.388-2.306
Coping strategy, reference: No
Exercise	0.473	0.048	97.470	1	<0.001	1.6041	1.461-1.762
Watching TV	0.396	0.085	21.673	1	<0.001	1.4861	1.258-1.756
Reading news	0.275	0.053	27.313	1	<0.001	1.3161	1.187-1.459
Cooking	0.333	0.070	39.950	1	<0.001	1.2581	1.258-1.547
Social media	0.323	0.092	12.284	1	<0.001	1.3811	1.153-1.654
Cox & Snell R2	0.042
Nagelkerke R2	0.077
Correct predictions	86.2%
Tobacco
Intersection	-2.575	0.359	51.477	1	<0.001	0.076
Age	0.012	0.003	17.112	1	<0.001	1.0121	1.006-1.018
Mental disorder, reference: Never Mental Disorder
Current Mental Disorder (CMD)	0.330	0.074	19.745	1	<0.001	1.3911	1.202-1.609
Education level, reference: Primary
University	-0.721	0.157	20.998	1	<0.001	0.4862	0.357-0.662
Work status, reference: Retired
Unemployed	0.802	0.152	27.687	1	<0.001	2.2312	1.654-3.008
Employed	0.606	0.137	19.511	1	<0.001	1.8342	1.401-2.400
Self-employed	0.575	0.156	13.503	1	<0.001	1.7772	1.308-2.415
Other	0.630	0.165	14.538	1	<0.001	1.8772	1.358-2.595
Elderly dependents, reference: None
One	0.373	0.084	19.742	1	<0.001	1.2281	0.935-1.612
Coping strategies, reference: No
Exercise	-0.545	0.049	121.542	1	<0.001	0.5801	0.526-0.639
Yoga	-0.252	0.064	15.579	1	<0.001	0.7881	0.686-0.881
Cox & Snell R2	0.046
Nagelkerke R2	0.090
Correct predictions	88.3%
Alcohol & Tobacco
Intersection	-4.093	0.516	63.034	1	<0.001	0.017
Age	-0.016	0.004	16.962	1	<0.001	0.9841	0.976-0.992
Sex (female)	-0.370	0.069	28.732	1	<0.001	0.6911	0.603-0.791
Work status, reference: Retired
Self-Employed	0.868	0.239	13.142	1	<0.001	2.3832	1.490-3.810
Income (€), reference: No income
Less than 500	0.605	0.144	17.739	1	<0.001	1.8312	1.382-2.426
500–999	0.644	0.151	18.144	1	<0.001	1.9042	1.416-2.561
1000–1499	0.549	0.158	12.125	1	<0.001	1.7312	1.271-2.358
More than 1999	0.709	0.172	16.917	1	<0.001	2.0322	1.449-2.848
Living situation, reference: Alone
Three to five	-0.528	0.109	23.535	1	<0.001	0.5902	0.477-0.730
Coping strategies, reference: No
Reading news	0.304	0.071	18.252	1	<0.001	1.3551	1.179-1.557
Cooking	0.471	0.076	38.805	1	<0.001	1.6011	1.381-1.857
Cox & Snell R2	0.027
Nagelkerke R2	0.072
Correct predictions	93.5

Notes: DASS-21: Depression, Anxiety and Stress Scale; IES: Impact of Event Scale; SE: Standard error; df: Degrees of freedom; OR: Odds ratio; CI: Confidence interval.

Cohen’s d equivalence: insignificant (OR < 1.68);

Cohen’s d equivalence: small (OR = 1.68–3.47).

Regarding non-modifiable factors, being female versus male was a protective factor. Regarding the main risk factors, having an income of more than €1999 versus having no income was associated with alcohol consumption as a coping strategy. Doing versus not doing activities for distraction was also a risk factor.

With respect to sex, doing versus not doing exercise was associated with alcohol consumption as a coping method in both sexes. However, in females, doing versus not doing activities for distraction such as watching TV and reading news was also a risk factor. In males, income of more than €1999 versus no income was a risk factor (see Supplementary Table 4).

Tobacco consumption as a coping strategy

Table 4 shows that being unemployed versus retired was a risk factor. Likewise, older age, having one elderly dependent versus none, and having a current mental disorder versus never having any were associated with tobacco use as a coping strategy. Regarding protective factors, university versus primary education was the main protective factor against tobacco consumption as a coping strategy.

Regarding the main factors associated with tobacco consumption as a coping method by sex, doing versus not doing activities such exercise reduced the risk in both sexes. In males, having a university versus primary education was a protective factor against tobacco consumption as a coping strategy. However, in females, being unemployed or employed versus retired, having one elderly dependent versus none, and having a current mental disorder versus never having any were associated with tobacco use as a coping strategy (see Supplementary Table 8).

Alcohol and tobacco consumption as a coping strategy

We also developed a logistic regression model to identify variables associated with consumption of both substances as a coping strategy (see Table 4).

Regarding non-modifiable factors, being female versus male and older age were protective factors. Furthermore, living with 3–5 people versus living alone reduced the risk of using both substances to cope with the pandemic. Table 4 shows that being self-employed versus retired was a risk factor. Likewise, having an income of more than €1999 versus no income, doing versus not doing activities for distraction such as reading news about COVID-19 and cooking were also risk factors.

With respect to sex, in males, being younger and having an income less than €500 versus no income were associated with alcohol and tobacco consumption as a coping strategy. However, in females, having an income of more than €1999 versus no income and doing versus not doing activities for distraction such as reading news and cooking were identified as a risk factors. Furthermore, living with 3–5 people versus living alone was a protective factor (see Supplementary Table 8).

Discussion

This is the first study to provide a population-based characterization of sociodemographic, risk and protective factors associated with consumption of alcohol, tobacco, or both as a coping strategy in the Spanish population early in the COVID-19 pandemic and lockdown. We found differences between people who consumed alcohol, tobacco, or both as a strategy to cope with the pandemic and lockdown.

After examining OR values, the strength of association tends to be insignificant (OR < 1.68) or small (OR 1.68–3.47) in most cases. However, all associations have been included in the Discussion, as we felt this could potentially be of theoretical interest.

Factors associated with alcohol and/or tobacco consumption during the pandemic and lockdown

In our study, 13.5% of the sample reported alcohol consumption as a distraction during the lockdown and pandemic. These rates are higher than those obtained by Wu et al., 2008, where approximately 6% of the sample reported using alcohol to cope with unpleasant feelings during the SARS outbreak. The fact that alcohol consumption is very widespread in Spain but less so in China could explain these differences (Hao et al., 2004). Furthermore, recent findings reflect increased alcohol consumption (Malta et al., 2020, Sun et al., 2020, Vanderbruggen et al., 2020) and use as a coping strategy during the COVID-19 pandemic (Hawke et al., 2020).

Regarding non-modifiable factors, being female reduced the risk. This is consistent with data from one national survey (Plan Nacional de Drogas, 2018) where males had a higher prevalence of alcohol consumption. Recent findings also reflect that males consumed more alcohol early in the pandemic (Ahmed et al., 2020).

Socioeconomic characteristics such as upper-middle income increased the risk for alcohol consumption as a coping method. This is in line with a previous study on the SARS outbreak (Waldrop et al., 2007) and with a large longitudinal study in the US general population (Erschens et al., 2018, Lui et al., 2018). According to Devaux and Sassi (2016), people with the highest income tend to consume more alcohol more frequently to cope with jobs with high responsibility and stress. However, in our sample, higher income impacted alcohol consumption as coping strategy only in males. This may be because males are still more likely to have jobs that involve high responsibility.

Some authors distinguish three coping styles: avoidant, emotion-focused, and problem-focused (Mihashi et al., 2009). People are more likely to use emotional or avoidance strategies in this situation because they cannot change it. Alcohol could be one of these strategies. However, we found no association between avoidance responses on the IES, while distractive behaviours did show such associations. Alcohol consumption could be an avoidance strategy not assessed by the IES. Females did more distractive activities than males. It has been suggested that females tend to use more distractive strategies as an avoidant coping style (Tabernero et al., 2019). Another explanation could be that alcohol consumption is perceived as a reward for hard work or good athletic performance (Vlahov et al., 2002). During lockdown, there are fewer positive reinforcement methods, and one of the most readily available is alcohol.

In our study, 12.0% of the sample reported using tobacco to cope with the pandemic and lockdown. This prevalence was similar to alcohol, albeit higher in the general population (Plan Nacional de Drogas, 2018). Many people in this study used tobacco as a coping strategy. Previous studies analysing the effect of the pandemic on this behaviour have found inconsistent results (Malta et al., 2020, Sun et al., 2020, Vanderbruggen et al., 2020). However, tobacco use increases after exposure to traumatic events. After the terrorist attack of 9/11, in a random telephone survey of Manhattan residents, Vlahov et al. (2002) found that 9.7% of participants reported increased alcohol, tobacco, and marijuana consumption. These rates are lower than in Spain, but the situation is different, as the pandemic involves isolation and social distancing.

Regarding non-modifiable factors, being older increased the risk. This is consistent with data from one national survey (Plan Nacional de Drogas, 2018) where the highest prevalence of tobacco consumption was in the 45–54 years age group. In fact, being older was a risk factor in females, consistent with the survey showing that older females consume more tobacco.

Sociodemographic characteristics such as being unemployed, employed, or self-employed were risk factors for tobacco use to cope with the pandemic. There is growing evidence that socioeconomic environment influences this behaviour (Chen et al., 2019, Hiscock et al., 2012). Living in an area with high unemployment and lower income constitutes a high risk for tobacco use (Daponte-Codina, Bolívar-Muñoz, Ocaña-Riola, Toro-Cárdenas & Mayoral-Cortés, 2009). These sociodemographic characteristics have more impact in females. The relationship between tobacco consumption, being female, and being socially disadvantaged has been reported in previous studies (Schiaffino et al., 2003).

However, having a university education was a protective factor. Higher education is normally associated with greater knowledge/concern about health and therefore lower risk of tobacco use (Mansouri et al., 2019). However, our findings reflect that a university education was a protective factor only in males. In a study by Jiménez-Rodrigo (2010), males were more likely to quit smoking because it affected their academic performance.

Having one elderly dependent was also a risk factor. Being a caregiver is associated with detrimental health behaviour such as smoking (Gallant & Connell, 1998). The literature also reflects that females are more often caregivers and experience a greater burden than their male counterparts (Xiong et al., 2020). This could explain why having one elderly dependent was associated with tobacco consumption as a coping method only in females.

Activities like exercise and yoga were protective factors. These behaviours, in addition to being incompatible with tobacco use, are associated with promoting and improving physical health, as well as health consciousness incompatible with smoking.

The authors suggest that tobacco use is one emotion-focused coping strategy that mitigates symptoms in the short term but fails to address the main source of the problem. In the long term, it can inhibit psychological adjustment and cause psychiatric disorders (Bazrafshan, Jahangir, Mansouri & Kashfi, 2014). In that regard, current mental disorder was a risk factor associated with smoking to cope with the pandemic. In a study by McClave et al. (2010), participants with life-time and current mental disorders had higher tobacco use rates than those who reported never having a mental illness. Thus, a current mental disorder could increase vulnerability to the use of dysfunctional coping strategies. In addition, early in the COVID-19 pandemic in China, females experienced a more significant psychological impact (Wang et al., 2020b). In our sample, females tended to report a current mental disorder more frequently than males, and this could be related to tobacco consumption as a coping strategy.

In our study, 6.5% of the sample reported using alcohol and tobacco to cope with the pandemic and lockdown. Being older and female were protective factors against both substances. According to data from one national survey (Plan Nacional de Drogas, 2018), females tend to consume less alcohol and tobacco than males. Furthermore, in a study by Romero-Blanco et al. (2020) university students consumed more alcohol and tobacco during lockdown. The lockdown probably had a greater impact on young people because they were deprived of their usual social/community environment.

We also examined the role of work status and found that self-employment was a risk factor. Since the start of the lockdown, self-employed people could not open their businesses, increasing their uncertainty and concern about the future. In previous pandemics, economic instability was one of the most important factors negatively affecting mental health (Mihashi et al., 2009). Furthermore, this uncertainty and instability could have had more impact on self-employed individuals who continued to pay taxes during lockdown without an income.

Both low and high income increased the risk of alcohol and tobacco use as a coping method. There were inconsistent results in previous studies where different socioeconomic levels were associated with substance use. However, in our study, to understand these results, we must analyse the differences between the sexes. Lower income was a risk factor in males, while higher income was a risk factor in females. In that regard, a study developed to investigate alcohol consumption patterns by sex in countries that differ in income level found that women had higher alcohol consumption patterns in higher-income countries due to cultural factors (Chaiyasong et al., 2018).

Social support is a protective factor against substance use (Cornwell and Waite, 2009, Musick and Bumpass, 2012). Thus, living with people reduced the risk of substance use as a coping method, which is consistent with previous results (Vanderbruggen et al., 2020). Research suggests that the form and function of cohabitation are similar to social support from a marriage (Musick & Bumpass, 2012) and, as the current state of self-isolation and social distancing is having a strong impact on daily life, this could be alleviated by living with family or friends. However, in our study, this was true in women only, which could be related to the fact that women had higher average levels of perceived satisfaction with social support (Tinajero et al, 2015).

The coping strategies of readings news and cooking were associated with alcohol and tobacco consumption as a coping method. These activities are compatible with drinking and smoking, and people reported using them as a coping strategy to deal with discomfort, which would explain their association. Both activities were risk factors in females, which is consistent with our previous explanation, i.e. this could be an avoidance and emotional regulation strategy for them.

Finally, there are two theories about how lockdown has influenced substance use. One is associated with increased substance use during a stressful situation (Bianchini et al., 2015, Hogarth et al., 2019). But lockdown could also reduce alcohol and tobacco use as a coping method due to less availability. However, in Spain, the lockdown resulted in closure of places of consumption but not places of sale. Thus, the availability of alcohol and tobacco was relatively stable during the pandemic.

Some limitations of the study arise from its inherent methodology. The first limitation is the online snowball recruitment strategy that was used instead of random selection. Our conclusions cannot be extrapolated to the Spanish general population, as Catalonia and Valencia were clearly under-represented. This type of sample collection strategy did not reveal how many people received the questionnaire so we could not determine the participation rate. Another limitation is that individuals were asked to self-report their substance abuse, thus potentially introducing bias into the findings. In addition, the use of ad hoc questionnaires precludes certainty regarding the measurement of these variables, and we included only questions about substance use as a coping method, thus we do not know its full prevalence in our sample or the patterns of consumption that might influence the way alcohol or tobacco is used. The cross-sectional design of the study does not provide information to detect changes in mental health and coping methods over time. However, the main strength of this study consists of the non-restrictive inclusion and exclusion criteria and the large sample size. Therefore, our results are robust and reflect use of alcohol, tobacco, or both as strategies to cope with the pandemic and lockdown.

Conclusions

To our knowledge, this is the first study in Spain to provide a profile of the population that uses substances to deal with feelings caused by the lockdown. Since the state of emergency was declared, no recommendations have been made to prevent mental health problems or substance use. Among the findings of this study, it should be noted that a percentage of people report substance use as a coping strategy, so this should be taken into account when implementing preventative public health strategies for future pandemics and other stressful life events.

Our findings may help promote timely interventions, customised by sex, with the aim of alleviating the negative impact of the pandemic on substance use.

Role of the funding source

This work was partly supported by the Government of the Principality of Asturias PCTI-2018-2022 IDI/2018/235, the CIBERSAM, and Fondos Europeos de Desarrollo Regional (FEDER).

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author contributions

LG-A, LF-T, MPG-P, PAS and JB designed the study. All authors reviewed it, gave approvals, and acquired the data. LF-T, CMC and PAS conducted statistical analyses. CMC, LF-T, IM and PAS wrote the first draft of the manuscript. All authors reviewed all drafts and gave the final approval.

CRediT authorship contribution statement

Clara Martínez-Cao: Conceptualization, Data curation, Formal analysis, Methodology, Investigation, Writing - original draft. Lorena Fuente-Tomás: Data curation, Formal analysis, Methodology, Investigation, Writing - original draft. Isabel Menéndez: Data curation, Writing - review & editing. Ángela Velasco: Data curation, Writing - review & editing. Paula Zurrón-Madera: Data curation, Writing - review & editing. Leticia García-Álvarez: Conceptualization, Data curation, Supervision, Project administration, Writing - review & editing. Pilar A. Sáiz: Conceptualization, Formal analysis, Methodology, Supervision, Project administration, Writing - review & editing. Paz Garcia-Portilla: Conceptualization, Methodology, Funding acquisition, Supervision, Project administration, Writing - review & editing. Julio Bobes: Conceptualization, Methodology, Funding acquisition, Supervision, Project administration, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.