Abstract
Aims
The COVID-19 pandemic presents the opportunity to learn about solitary drinking as many people were forced to spend time at home. The aim of this study is to examine the relationship between solitary drinking and living without other adults on alcohol consumption.
Methods
A longitudinal study with four survey waves (between May and November 2020) obtained seven-day drinking diary data from Australian adults living in New South Wales. In May, a convenience sample of 586 participants (Mage = 35.3, SD = 14.8; 65.3% women) completed the first wave. Participants then completed a survey in June (n = 319, 54.4% response rate), July/August (n = 225, 38.4% response rate), and November (n = 222, 37.9% response rate). Information about alcohol consumption including risky drinking (more than four drinks on one occasion), household structure, solitary drinking, and demographics were collected. We conducted random-effects panel bivariate and multivariable regression analyses predicting the number of standard drinks and risky drinking.
Results
Participants with solitary drinking occasions consumed more and had more risky drinking occasions than participants with no solitary drinking occasions, which was also found to be the case during lockdown. Living without other adults was associated with less consumption and less risky drinking than living with other adults. However, participants who lived without other adults and had frequent solitary drinking occasions (solitary drinking in >50% drinking occasions) reported more consumption than participants without a solitary drinking occasion.
Conclusions
Individuals who consume alcohol alone and live without other adults or spend long periods of time at home may be more at risk of alcohol-related harm.
Keywords: alcohol, solitary drinking, living without other adults, longitudinal study, COVID-19, Australia
Short Summary:
Solitary drinking was positively associated with alcohol consumption and risky drinking.
Living without other adults was negatively associated with alcohol consumption and risky drinking.
Positive interaction between living without other adults and frequent solitary drinking on consumption.
Positive interaction between solitary drinking and lockdown on alcohol consumption and risky drinking.
Introduction
Recent systematic reviews and meta-analyses have found that solitary drinking, that is drinking alone whilst not in the presence of others, is positively associated with increased alcohol consumption and drinking problems in adults (Skrzynski and Creswell 2021) and in adolescents and young people (Skrzynski and Creswell 2020). People who consume alcohol daily have been found to be nine times more likely to report solitary drinking in the last three months compared to those consuming alcohol once a month or less (Demers and Bourgault 1996). However, it is unclear how much alcohol was consumed every day and if those who consume alcohol daily drink alone because they drink heavily or because they live alone. Previous research has found that the majority of alcohol is consumed in the person’s own home (Callinan et al. 2016). Therefore, these results should be interpreted in conjunction with living arrangements as people who consume alcohol every day may be doing so because they live alone and not necessarily because they drink heavily.
Research examining the relationship between living arrangements and consumption have reported mixed findings. Living alone has been positively association with heavy drinking among Finnish adults (Joutsenniemi et al. 2007) and in Japanese men (Okui 2021). Meanwhile, for Japanese women living alone was associated with more moderate drinking but only for those aged 20–59 (Okui 2021). In contrast, Spaniards aged 60 years and over who lived alone were less likely to report moderate alcohol consumption compared with participants who lived with others (León-Muñoz et al. 2015). The association between living alone and consumption has been found to differ by gender and age.
In their systematic review and meta-analysis of changes in alcohol use during the COVID-19 pandemic, Acuff et al. (2022) identified no association between living alone and changes in alcohol consumption. Studies published after Acuff et al.’s (2022) paper found that during the pandemic living alone was positively associated with alcohol consumption (Villanueva-Blasco et al. 2021) and problematic alcohol use (Laghi et al. 2022), and another study found a negative association between living alone and consumption (Mangot-Sala et al. 2022). During the pandemic, one in four Australians who lived alone exceeded the low risk guidelines which was similar to those who lived in family households, but less than those who lived with non-related adults (Australian Bureau of Statistics 2022). Apart from this study, there is limited research examining living alone during the pandemic and risky drinking.
Studies conducted during the COVID-19 pandemic have found that solitary drinking increased when compared to pre-pandemic (Panagiotidis et al. 2020; Wardell et al. 2020; Clare et al. 2021), while one study found a reduction in solitary drinking (McPhee et al. 2020). However, there has been limited research exploring the relationship between solitary drinking and living alone. Literature has identified solitary drinking to be positively associated with living alone (Demers and Bourgault 1996; Stickley et al. 2015; Wardell et al. 2020). Participants living alone or living only with children are more than twice as likely to report solitary drinking compared to participants living with at least one other adult and with children (Demers and Bourgault 1996). This result is not unexpected as most solitary drinking occasions occurred at home (Demers and Bourgault 1996). What is currently unknown is if this interaction between living alone and solitary drinking is associated with heavy alcohol consumption.
The current study
To the best of our knowledge, no study has explored the interaction between solitary drinking and living alone in predicting alcohol consumption or risky drinking. In addition, almost all solitary drinking studies included in their systematic review and meta-analysis were cross-sectional with very few longitudinal studies (Skrzynski and Creswell 2021). The COVID-19 pandemic presents the opportunity to learn about solitary drinking as many people were forced to spend time at home. The current study took place during COVID-19 restrictions in New South Wales (NSW), Australia. In Australia, licensed premises and non-essential businesses were closed nationwide on the 23 March 2020. Lockdown restrictions were then introduced on the 30 March 2020, with only four reasons to leave the house: essential shopping, essential work, medical/health needs, and exercise. During lockdown, takeaway liquor shops remained open and alcohol delivery services could operate (Miller et al. 2021). With restrictions beginning to ease in late-April 2020, mid-May 2020 saw ten people permitted to dine at cafes and restaurants in NSW. Restrictions continued to ease in NSW with some tightening on dining limits introduced in mid-July. Further information on the restrictions in NSW can be found in the Supplementary Table 1 and elsewhere (Miller et al. 2021).
The aim of this study was to examine the relationship between solitary drinking and living without other adults on alcohol consumption using 7-day diary data obtained from Australian adults living in NSW during the first year of the COVID-19 pandemic. We hypothesized that the following:
Solitary drinking will be positively associated with higher total consumption and risky drinking in a typical week.
Living without other adults will be associated with higher total consumption and risky drinking in a typical week compared to those who live with other adults.
Solitary drinking in those living without other adults will be associated with higher total consumption and risky drinking in a typical week compared to solitary drinking in those who live with other adults.
Solitary drinking during lockdown will be positively associated with higher total consumption and risky drinking in a typical week.
Method
Participants and procedure
This study was approved by the University of New South Wales Ethics Committee (#HC200305). Participants were recruited through convenience sampling via advertisements in social media, digital noticeboards, and community emails. Eligible participants had to be over the age of 18, reside in NSW and have consumed alcohol in the last twelve months. This study comprised four survey waves that occurred between 6 May 2020 and 30 November 2020. Supplementary Table 1 displays the survey dates of the four survey waves which collected data from five time periods. All the data used in this study were collected during the pandemic, the first survey wave, completed between the 6 and 18 May 2020, included questions covering the pre-lockdown period (February 2020) and since the COVID-19 restrictions were introduced (since March 23 2020). All surveys were completed online via Qualtrics. The first survey took approximately 18 min to complete, with the subsequent survey waves taking 10 min to complete. To match participant responses across each survey wave, a self-generated unique alphanumeric identifier was used. At the conclusion of each survey, participants could enter a draw to win a $200 grocery voucher.
A total of 586 participants (65.3% women) completed the first survey. Further information about the number of participants in each survey are displayed in the Supplementary Table 1. The number of participants declined in wave 2 (n = 319, 54.4% response rate), wave 3 (n = 225, 38.4% response rate), and wave 4 (n = 222, 37.9% response rate).
Measures
In each survey, wave participants were asked to complete a detailed 7-day drinking diary with questions about their drinking in a typical week of the reference period (outlined in Supplementary Table 1). In the first survey, participants completed drinking diaries for the pre-lockdown period (February 2020) and since the COVID-19 restrictions were introduced (since March 23, 2020). In the following survey waves, participants completed one drinking diary for their respective reference period (please see Supplementary Table 1). From the drinking diaries of participants, the total number of standard drinks was obtained, if there was any risky drinking, and the proportion of solitary drinking in a typical week. In addition, participants were asked about their household structure and demographics.
Alcohol consumption
Total number of standard drinks in a typical week
In the drinking diary, participants were asked to select the days of the week (Monday to Sunday) that they would typically have a drink. Participants were then asked on each of these days of week, how many standard drinks they consumed that day. A total number of standard drinks in a typical week was calculated by summing the daily number of standard drinks together.
Risky drinking in a typical week
Using the drinking diaries, a binary variable was created to distinguish between participants who were risky drinking in a typical week and those who were not. Risky drinking was defined as consuming more than 4 standard drinks at least once in a typical week. This definition of risky drinking was based on the Australian guidelines to reduce health risks from drinking alcohol (National Health and Medical Research Council 2020).
Solitary drinking occasions in a typical week
In the drinking diaries, participants were asked who they drank with (alone, friend(s), family, work colleague(s), spouse/partner, other) on each day. The proportion of solitary drinking was obtained by dividing the number of drinking occasions where participants drank alone in the typical week with the total number of drinking occasions for that week. The solitary drinking occasions were grouped into three categories: none (0%), infrequent (participants who drank alone in >0% to 50% of drinking occasions) and frequent (participants who drank alone in more than 50% of drinking occasions). Although a fairly arbitrary decision, by splitting solitary drinking occasions into three categories we could distinguish between participants who consumed alcohol alone in more than half of their drinking occasions in a typical week (frequent) against those who consumed alcohol alone but did so infrequently or had no solitary drinking occasions.
Household structure
Participants were asked about their household structure with the following response options: person living alone, couple, single parent, non-related adults sharing house/apartment/flat, and other household type. A binary variable was created to distinguish between participants who lived with other adults and those who lived without other adults. In the current study, living without other adults included participants who responded with ‘person living alone’ or ‘single parent’. Previous research has strictly defined living alone as living without another person. However, research has explained that solitary home drinking is often the only way single parents consume alcohol at home (Cook et al. 2023). This is particularly the case during the pandemic when there were restrictions on licensed premises and visitor limits, limiting the opportunity that single parents had to consume alcohol with other adults. Therefore, in this current study, we included single parents and living alone together to form a living without another adult category.
Lockdown
A lockdown binary variable was created to indicate when participants were in lockdown (as seen in Supplementary Table 1).
Demographics
Participants’ age (18–29, 30–49, and 50+) and gender (female, male, non-binary, or prefer not to say) from the first survey were used in the analyses. For the analyses including gender, only female and male categories were used as only one participant did not select either option (non-binary).
Analysis
To explore the hypotheses, descriptive statistics, random-effect regression and logistic regression analyses were conducted using Stata version 18 (StataCorp 2023). First, we provide descriptive statistics of the sample. Next, we calculated the mean standard drinks and the proportion of any risky drinking for participants living without and with adults. Then the proportion of solitary drinking occasions in a typical week for participants living without and with adults during each survey wave was calculated.
To examine the first and second hypotheses, five random-effects panel bivariate regression models were conducted, one for each of the categorical predictor variables (solitary drinking occasions, living with other adults, lockdown, gender and age) and using the total number of standard drinks in a typical week as the outcome variable. In addition, five logistic bivariate regressions were conducted, one for each of the predictors with the binary risky drinking variable used as the outcome variable. To explore the third hypothesis, a random-effects panel multivariable regression analysis (Model 1) predicting the total number of standard drinks in a typical week was conducted using the five predictor variables and an interaction between living without other adults and solitary drinking occasions. In addition, a multivariable logistic regression analysis (Model 3) predicting risky drinking in a typical week was conducted including the five predictor variables and an interaction between living without other adults and solitary drinking occasions. To examine the fourth hypothesis, a random-effects panel multivariable regression analysis and a multivariable logistic regression analysis were conducted to predict the total number of standard drinks (Model 2) and risky drinking (Model 4) in a typical week, respectively, both using the five predictor variables and an interaction between lockdown and solitary drinking occasions. Due to the non-linear relationship with consumption, age and solitary drinking occasions were used as categorical variables in the analyses (please see Figures S1 and S2 in the Supplementary Material that demonstrate the non-linear relationship). Figures were included to explore any significant interactions from the two random-effects panel multivariable regression analyses and the two multivariable logistic regression analyses.
Missing data
Across the study period, we had a total of 1938 completed surveys from 586 participants, with 33.9% (992 surveys were not completed out a possible of 2930) of surveys not completed. There were also missing values within the completed surveys, however this was a not greater than 1.5% of additional missing data for one predictor variable across the five surveys so these participants were retained. Respondents were only excluded from analyses that included variables that they did not provide data for. Therefore, the number of participants used in the random-effects panel bivariate models varies depending on the predictor variable in the regression model (ranged from 576 to 586 participants). With consistent variables included in the multivariable models, there was 570 participants in each of these models.
Results
Descriptive statistics for participants who live without and with other adults are displayed in Table 1. Using the 95% confidence intervals, participants living with other adults were significantly younger, with almost half of all participants who reported living with other adults aged 18–29, compared to 27.3% living without other adults. While participants living without other adults were significantly older, with 28.1% of participants living without adults aged 50+ compared to 17.1% of participants living with other adults. However, based on the 95% confidence intervals, there was no significant gender differences between those who lived with and without other adults. The proportion of participants living without other adults remained fairly consistent throughout the study period.
Table 1.
Descriptive statistics for participants living without other adults and living with other adults, with 95% confidence intervals
| Living without other adults | Living with other adults | |
|---|---|---|
| Survey wave (%) | ||
| Pre-lockdown (baseline) | 22.2 (19.0, 25.8) | 77.8 (74.2, 81.0) |
| NSW lockdown (wave 1) | 21.7 (18.5, 25.3) | 78.3 (74.7, 81.5) |
| Partial re-opening (wave 2) | 19.8 (15.8, 24.6) | 80.2 (75.4, 84.2) |
| Open and shut (wave 3) | 23.8 (18.6, 29.8) | 76.2 (70.2, 81.4) |
| COVID-normal (wave 4) | 22.0 (16.9, 28.1) | 78.0 (71.9, 83.1) |
| Age at baseline | ||
| 18–29 (%) | 27.3 (20.3, 35.8) | 49.4 (44.8, 54.1) |
| 30–49 (%) | 44.5 (36.1, 53.3) | 33.4 (29.2, 37.9) |
| 50+ (%) | 28.1 (21.0, 36.6) | 17.1 (13.9, 20.9) |
| Gender at baseline (%) | ||
| Women | 71.7 (63.1, 78.9) | 63.3 (58.6, 67.7) |
| Men | 28.3 (21.1, 36.9) | 36.7 (32.3, 41.4) |
Table 2 shows the mean number of standard drinks and proportion of any risky drinking in a typical week for participants who live without and with other adults. Pre-lockdown, the mean standard drinks in a typical week between those living without and with other adults were similar. During lockdown, those living without other adults increased consumption while those living with other adults maintained their pre-lockdown consumption. Following lockdown, both groups reported less consumption, however, those living without other adults returned to their pre-lockdown consumption during COVID-normal (wave 4).
Table 2.
Mean standard drinks and the proportion of any risky drinking (>4 standard drinks) occasion in a typical week for participants living without other adults and living with other adults across each survey wave, with 95% confidence intervals
| Mean number of standard drinks in a typical week | % with any risky drinking in a typical week | |||
|---|---|---|---|---|
| Living without other adults | Living with other adults | Living without other adults | Living with other adults | |
| Pre-lockdown (Baseline) | 15.5 (12.0, 19.1) | 15.8 (14.2, 17.4) | 50.0 (41.4, 58.6) | 59.9 (55.3, 64.4) |
| NSW lockdown (wave 1) | 16.8 (12.1, 21.5) | 15.6 (13.8, 17.5) | 37.9 (29.8, 46.8) | 45.0 (40.4, 49.6) |
| Partial re-opening (eave 2) | 12.4 (7.1, 17.6) | 13.0 (11.3, 14.7) | 33.9 (23.2, 46.5) | 49.8 (43.6, 56.0) |
| Open and shut (wave 3) | 12.8 (7.5, 18.2) | 12.1 (10.1, 14.1) | 34.0 (22.5, 47.6) | 44.7 (37.4, 52.2) |
| COVID-normal (wave 4) | 15.8 (9.3, 22.4) | 11.9 (9.9, 13.8) | 43.5 (30.0, 58.0) | 44.2 (36.7, 51.9) |
Regarding risky drinking, participants living with other adults reported a larger proportion of any risky drinking in a typical week across the study period than those living without other adults. During lockdown, both groups reported less risky drinking compared to pre-lockdown. In the partial re-opening phase (wave 2), there was an increase in the proportion of risky drinking among those living with other adults compared with lockdown (wave 1). In contrast, those living with adults reported a lower proportion of risky drinking in wave 2 compared to lockdown. In the final wave, both groups reported less risky drinking compared to pre-lockdown.
Table 3 displays the solitary drinking occasions in a typical week during the study period for participants who live without and with other adults. Most participants did not report any solitary drinking. As expected, solitary drinking is more common in those who lived without other adults. Broadly, solitary drinking increased during lockdown before fluctuating as NSW went out of lockdown.
Table 3.
Solitary drinking occasions in a typical week for participants living without other adults and living with other adults for each survey wave, with 95% confidence intervals
| Living without other adults (%) | Living with other adults (%) | |||||
|---|---|---|---|---|---|---|
| Solitary drinking occasions in a typical week | None | Infrequent | Frequent | None | Infrequent | Frequent |
| Pre-lockdown (Baseline) | 62.2 (53.5, 70.2) | 10.2 (6.0, 16.9) | 27.6 (20.5, 36.0) | 80.2 (76.3, 83.7) | 10.6 (8.0, 13.8) | 9.2 (6.9, 12.3) |
| NSW lockdown (wave 1) | 51.7 (42.7, 60.5) | 10.8 (6.4, 17.8) | 37.5 (29.3, 46.5) | 75.8 (71.6, 79.6) | 10.2 (7.7, 13.4) | 14.0 (11.1, 17.6) |
| Partial re-opening (wave 2) | 58.1 (45.5, 69.7) | 19.4 (11.3, 31.1) | 22.6 (13.8, 34.6) | 79.7 (74.2, 84.2) | 10.0 (6.8, 14.3) | 10.4 (7.1, 14.8) |
| Open and shut (wave 3) | 66.0 (52.4, 77.5) | 15.1 (7.7, 27.4) | 18.9 (10.4, 31.7) | 85.9 (79.8, 90.4) | 5.9 (3.2, 10.6) | 8.2 (4.9, 13.4) |
| COVID-normal (wave 4) | 67.4 (52.7, 79.3) | 6.5 (2.1, 18.4) | 26.1 (15.4, 40.6) | 87.7 (81.7, 91.9) | 6.1 (3.3, 11.0) | 6.1 (3.3, 11.0) |
Note: Infrequent = participants who drank alone in 50% or fewer drinking occasions; frequent = participants who drank alone in more than 50% of drinking occasions.
The results of the random-effects panel bivariate and multivariable regression analyses predicting the number of standard drinks and the logistic regression analyses predicting risky drinking in a typical week are displayed in Table 4. In support of the first hypothesis, participants who reported solitary drinking occasions reported a greater number of standard drinks and were more likely to report risky drinking in a typical week compared to participants without a solitary drinking occasion.
Table 4.
Random-effects panel bivariate and multivariable regression models predicting the number of standard drinks, and bivariate and multivariable logistic regression models predicting risky drinking in a typical week
| Number of standard drinks in a typical week | Risky drinking in a typical week | ||||||
|---|---|---|---|---|---|---|---|
| BV B [CI] |
MV (Model 1) B [CI] |
MV (Model 2) B [CI] |
BV OR [CI] |
MV (Model 3) OR [CI] |
MV (Model 4) OR [CI] |
||
| Solitary drinking occasions in a typical week | None (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) |
| Infrequent | 8.49*** (6.50, 10.48) |
6.56*** (4.24, 8.88) |
5.09*** (2.88, 7.31) |
3.66*** (1.99, 6.73) |
3.34** (1.56, 7.13) |
2.35* (1.12, 4.90) |
|
| Frequent | 9.69*** (7.79, 11.59) |
7.26*** (4.81, 9.71) |
7.21*** (5.00, 9.42) |
2.93*** (1.68, 5.10) |
4.08*** (1.91, 8.73) |
2.38* (1.17, 4.83) |
|
| Household structure | Living with other adults (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) |
| Living without other adults | −0.97 (−3.22, 1.27) |
−3.74** (−6.22, −1.25) |
−2.45* (−4.67, −0.23) |
0.45** (0.25, 0.80) |
0.35** (0.17, 0.72) |
0.36** (0.19, 0.68) |
|
| Lockdown | No (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) |
| Yes | 0.84 (−0.13, 1.81) |
−0.11 (−1.07, 0.85) |
−1.64** (−2.77, −0.51) |
0.44*** (0.32, 0.60) |
0.36*** (0.26, 0.50) |
0.25*** (0.16, 0.37) |
|
| Gender | Women (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) |
| Men | 8.21*** (5.29, 11.13) |
8.02*** (5.28, 10.76) |
7.97*** (5.23, 10.71) |
3.29*** (1.80, 6.01) |
3.67*** (1.91, 7.05) |
3.76*** (1.94, 7.28) |
|
| Age | 18–29 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) |
| 30–49 | 1.98 (−1.16, 5.12) |
1.33 (−1.61, 4.27) |
1.36 (−1.58, 4.31) |
0.63 (0.33, 1.19) |
0.60 (0.30, 1.20) |
0.60 (0.30, 1.20) |
|
| 50+ | 4.50* (0.70, 8.30) |
3.42 (−0.12, 6.96) |
3.60* (0.06, 7.15) |
0.53 (0.24, 1.15) |
0.43* (0.19, 1.00) |
0.44 (0.19, 1.02) |
|
| Living without other adults x solitary drinking occasions in a typical week | Yes x none (ref) | - | 0 (ref) | - | - | 0 (ref) | - |
| Yes x infrequent | - | 1.99 (−2.29, 6.28) |
- | - | 1.52 (0.37, 6.22) |
- | |
| Yes x frequent | - | 5.15** (1.32, 8.99) |
- | - | 0.94 (0.27, 3.25) |
- | |
| Lockdown x solitary drinking occasions in a typical week | Yes x none (ref) | - | - | 0 (ref) | - | - | 0 (ref) |
| Yes x infrequent | - | - | 6.91*** (3.44, 10.38) |
- | - | 4.79** (1.47, 15.59) |
|
| Yes x frequent | - | - | 5.08*** (2.32, 7.85) |
- | - | 3.66** (1.42, 9.42) |
|
* P < 0.05
** P < 0.01
*** P < 0.001. Ref = 1. BV = bivariate; MV = multivariable; B = unstandardised regression coefficient Beta; OR = Odds Ratio; Infrequently = participants who drank alone in 50% or fewer drinking occasions; frequently = participants who drank alone in more than 50% of drinking occasions.
As seen in Table 4, the second hypothesis was rejected as there was no association between participants who live without other adults and the number of standard drinks consumed in a typical week in the bivariate regression. However, in the multivariable regressions (Model 1 and 2), there was a negative association between living without other adults and the number of standard drinks. In addition, participants who lived without other adults were less likely to report risky drinking in a typical week compared with those living with other adults.
The third hypothesis was partially supported as there was a significant positive association between frequent solitary drinking occasions and living without other adults on the number of standard drinks in typical week (Table 4, Model 1). However, the interaction between living without other adults and solitary drinking on risky drinking in a typical week was not significant (Table 4, Model 3). To explore the significant interaction in Model 1, the estimated marginal means were produced in Figure 1(a). As displayed in Figure 1(a), regardless of the household structure, participants with any solitary drinking occasions reported more alcohol consumption than those with no solitary drinking occasions.
Figure 1.
Predictive estimated marginal means of the number of standard drinks in a typical week by a) living without other adults and solitary drinking occasions, b) lockdown and the solitary drinking, and c) the proportion of risky drinking in a typical week by lockdown and solitary drinking occasions in a typical week, with 95% confidence intervals.
Lastly, the fourth hypothesis was supported as there was a significant positive interaction between participants in lockdown with solitary drinking occasions on the number of standard drinks in typical week (Table 4, Model 2). In Figure 1(b), the interaction between solitary drinking and lockdown from Model 2 is shown with the estimated marginal means of the number of standard drinks. As seen in Figure 1(b), participants without a solitary drinking occasion reported fewer number of standard drinks in a typical week regardless of lockdown than participants who reported any solitary drinking occasion.
There was also a significant interaction between lockdown and the proportion of solitary drinking occasions on risky drinking in a typical week (Table 4, Model 4). Participants who reported infrequent and frequent solitary drinking occasions during lockdown were five and four times as likely, respectively, to report risky drinking in a typical week than those without a solitary drinking occasion. To explore this significant interaction, the estimated marginal means of the proportion of risky drinking on lockdown was produced in Figure 1(c). During lockdown, participants without a solitary drinking occasion reported less risky drinking than those who reported solitary drinking occasions.
Discussion
This study examined the relationship between solitary drinking and living without other adults on alcohol consumption during the first year of the COVID-19 pandemic. Supporting the first hypothesis, participants who reported any solitary drinking occasion consumed more alcohol and had higher odds of risky drinking in a typical week compared to those without solitary drinking occasions. This is consistent with results found in a systematic review and meta-analysis by Skrzynski and Creswell (2021) and in studies conducted during the pandemic (Panagiotidis et al. 2020; Wardell et al. 2020). This result adds to the limited longitudinal solitary drinking literature (Skrzynski and Creswell 2021).
In the current study, participants living without other adults consumed fewer standard drinks and had lower odds of risky drinking in a typical week compared to those who live with other adults, rejecting the second hypothesis. This is in contrast to pre-pandemic (Joutsenniemi et al. 2007; Okui 2021) and pandemic (Villanueva-Blasco et al. 2021; Laghi et al. 2022) research that found a positive association between living alone and alcohol consumption. The discrepancy in results may be due to participants in the current study who live without other adults may have reported less consumption because the venues where they drink, such as licensed premises, were closed or were open with restrictions (e.g. venue caps and density limits; Miller et al. 2021). A Victorian study found that participants who lived alone were more likely to indicate that the main reason they drank less during the first lockdown was due to the closure of places where they usually drink (VicHealth 2020). In addition, the current study’s definition of living without other adults (those who live alone and single parents) was different to the previously cited papers (Joutsenniemi et al. 2007; Okui 2021; Villanueva-Blasco et al. 2021; Laghi et al. 2022) who examined participants who lived alone.
The third hypothesis was partially supported as we only found a significant positive interaction between frequent solitary drinking occasions and living without other adults on alcohol consumption. Previous research has found a positive relationship between solitary drinking and living alone (Demers and Bourgault 1996; Stickley et al. 2015; Wardell et al. 2020), but there has been no research exploring how this relationship can influence alcohol consumption for people who live with and without other adults. The results from the current study demonstrate that living without other adults is associated with fewer standard drinks in a typical week compared to people living with other adults. However, if participants lived without other adults and also had more than half of their drinking occasions alone, then they reported more standard drinks in typical week compared to those who had no solitary drinking occasions. This result has implications for health professionals as it highlights the importance of drinking context among people who live without other adults, as those with frequent solitary drinking may be at greater risk of alcohol-related harm.
Lastly, the fourth hypothesis was supported as solitary drinking during lockdown was positively associated with consumption and risky drinking. This result is consistent with previous studies that found solitary drinking increased during the pandemic compared to pre-pandemic (Panagiotidis et al. 2020; Wardell et al. 2020; Clare et al. 2021). This result is important for health professionals to consider if they work with patients who tend to spend longer periods at home, such as people with chronic illness, particularly if they report solitary drinking as they may require further support to reduce their odds of risky drinking.
Strengths and limitations
The strengths of the current paper include the use of the detailed drinking diary that asked about daily alcohol consumption during a typical week for five different time periods. This allowed us to examine a more detailed set of data that would not be captured in the typical frequency/quantity items which generally asks participants about their alcohol consumption over a period of time (e.g. in the past year). By using the drinking diary, we were able to examine the proportion of solitary drinking occasions that took place in a typical week via a categorical variable (none, infrequent, and frequent), as suggested by Skrzynski and Creswell (2021) rather than the frequency of solitary drinking.
There were a few limitations that may have influenced our results. Firstly, participants were not asked if they consumed alcohol while socializing online. Therefore, it is not clear if participants would classify this interaction with others online as solitary drinking or drinking with others. Research has reported on the use of technology to consume alcohol with others virtually increased during the pandemic (McPhee et al. 2020). To be eligible for this study, participants had to report consuming alcohol in the past year prior to the first survey. Therefore, this study did not capture individuals who may have started drinking alcohol during the pandemic.
This study used a convenience sample obtained by via advertisements in social media, digital noticeboards, and community emails. Therefore, the results cannot be generalized to the NSW population. In addition, our results may be influenced by recall bias and social desirability bias. Specifically, in the first survey as participants were asked about their alcohol consumption across two time periods (pre-pandemic and since the pandemic started) which may have affected participants recall. However, due to the unprecedented nature of the pandemic, this was the most feasible way to obtain pre-pandemic alcohol consumption data. Due to the longitudinal nature of the study conducted during the pandemic, the number of participants who completed the survey declined each wave. We were also not able to capture the long-term changes in consumption once all the COVID-19 restrictions were removed. We also did not analyse if other factors such as stress, mental health and income during the pandemic influenced consumption among participants.
Conclusion
The current study is the first to explore the interaction between solitary drinking and living without other adults on alcohol consumption. Living without other adults was negatively associated with alcohol consumption; however, if these individuals also had frequent solitary drinking occasions they reported more consumption compared to those living without other adults and had no solitary drinking occasions. Meanwhile, participants with any solitary drinking occasions were four times as likely to report risky drinking during lockdown than those without a solitary drinking occasion. It is recommended that public health messaging target individuals who consume alcohol alone, particularly if they live without other adults or spend long periods of time at home.
Supplementary Material
Acknowledgements
The data collection for this study was supported by the UNSW COVID-19 Rapid Response Research Initiative.
Yvette Mojica-Perez, MPsych, Research Officer; Bree Willoughby, BPsychSc(Hons), Research Officer; Dan Anderson-Luxford, MCrim, Research Officer, Claire Wilkinson, PhD, Senior Research Fellow; Emmanuel Kuntsche, PhD, Director; Sarah Callinan, PhD, Senior Research Fellow; Alison Ritter, PhD, Director.
Contributor Information
Yvette Mojica-Perez, Centre for Alcohol Policy Research, La Trobe University, Melbourne, VIC 3086, Australia.
Bree Willoughby, Centre for Alcohol Policy Research, La Trobe University, Melbourne, VIC 3086, Australia.
Dan Anderson-Luxford, Centre for Alcohol Policy Research, La Trobe University, Melbourne, VIC 3086, Australia.
Claire Wilkinson, Drug Policy Modelling Program, Social Policy Research Centre, UNSW Sydney, Sydney, NSW 2052, Australia.
Emmanuel Kuntsche, Centre for Alcohol Policy Research, La Trobe University, Melbourne, VIC 3086, Australia; Institute of Psychology, Eötvös Loránd University, Budapest 1088, Hungary.
Sarah Callinan, Centre for Alcohol Policy Research, La Trobe University, Melbourne, VIC 3086, Australia.
Alison Ritter, Drug Policy Modelling Program, Social Policy Research Centre, UNSW Sydney, Sydney, NSW 2052, Australia.
Author contributions
Yvette Mojica-Perez (Formal analysis-Lead, Writing—original draft-Lead, Writing—review & editing-Equal), Bree Willoughby (Writing—review & editing-Equal), Dan Anderson-Luxford (Writing—review & editing-Equal), Claire Wilkinson (Investigation-Equal, Writing—review & editing-Equal), Emmanuel Kuntsche (Writing—review & editing-Equal), Sarah Callinan (Conceptualization-Equal, Methodology-Lead, Writing—review & editing-Equal), Alison Ritter (Conceptualization-Equal, Investigation-Equal, Writing—review & editing-Equal).
Conflict of interest statement: None declared.
Data Availability
The data underlying this article cannot be shared publicly as the authors do not have permission to share the data.
References
- Acuff SF, Strickland JC, Tucker JA. et al. Changes in alcohol use during COVID-19 and associations with contextual and individual difference variables: a systematic review and meta-analysis. Psychol Addict Behav 2022;36:1–19. 10.1037/adb0000796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Australian Bureau of Statistics . Alcohol Consumption: Contains Key Statistics and Information from Survey Data about Alcohol Consumption and its Prevalence within Australia. Canberra: ABS, 2022, Retrieved from: Australian Bureau of Statistics. [Google Scholar]
- Callinan S, Livingston M, Room R. et al. Drinking contexts and alcohol consumption: how much alcohol is consumed in different Australian locations? J Stud Alcohol Drugs 2016;77:612–9. 10.15288/jsad.2016.77.612. [DOI] [PubMed] [Google Scholar]
- Clare PJ, Aiken A, Yuen WS. et al. Alcohol use among young Australian adults in May-June 2020 during the COVID-19 pandemic: a prospective cohort study. Addiction 2021;116:3398–407. 10.1111/add.15599. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cook M, MacLean S, Callinan S. Home alone: patterns and perceptions of solitary home alcohol consumption in an Australian convenience sample. Drug Alcohol Rev 2023;42:1018–27. 10.1111/dar.13600. [DOI] [PubMed] [Google Scholar]
- Demers A, Bourgault C. Changing society, changing drinking: solitary drinking as a non-pathological behaviour. Addiction 1996;91:1505–16. [DOI] [PubMed] [Google Scholar]
- Joutsenniemi K, Martelin T, Kestilä L. et al. Living arrangements, heavy drinking and alcohol dependence. Alcohol Alcohol 2007;42:480–91. 10.1093/alcalc/agm011. [DOI] [PubMed] [Google Scholar]
- Laghi F, Di Tata D, Bianchi D. et al. Problematic alcohol use in young adults during the COVID-19 lockdown in Italy. Psychol Health Med 2022;27:139–49. 10.1080/13548506.2021.1975785. [DOI] [PubMed] [Google Scholar]
- León-Muñoz LM, Galán I, Donado-Campos J. et al. Patterns of alcohol consumption in the older population of Spain, 2008-2010. J Acad Nutr Diet 2015;115:213–24. 10.1016/j.jand.2014.08.017. [DOI] [PubMed] [Google Scholar]
- Mangot-Sala L, Tran KA, Smidt N. et al. The impact of the COVID lockdown on alcohol consumption in the Netherlands. The role of living arrangements and social isolation. Drug Alcohol Depend 2022;233:109349. 10.1016/j.drugalcdep.2022.109349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McPhee MD, Keough MT, Rundle S. et al. Depression, environmental reward, coping motives and alcohol consumption during the COVID-19 pandemic. Front Psych 2020;11:1–14. 10.3389/fpsyt.2020.574676. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Miller M, Mojica-Perez Y, Livingston M. et al. A Timeline of Alcohol-Relevant Restrictions during the COVID-19 Pandemic (Version 6), Centre for Alcohol Policy Research: Melbourne, Australia. 2021.
- National Health and Medical Research Council . Australian Guidelines to Reduce Health Risks from Drinking Alcohol: NHMRC. Commonwealth of Australia, Canberra: Australian Research Council and Universities Australia, 2020. [Google Scholar]
- Okui T. An analysis of predictors for heavy alcohol drinking using nationally representative survey data in Japan. BMC Public Health 2021;21:1–9. 10.1186/s12889-021-10382-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Panagiotidis P, Rantis K, Holeva V. et al. Changes in alcohol use habits in the general population, during the COVID-19 lockdown in Greece. Alcohol Alcohol 2020;55:702–4. 10.1093/alcalc/agaa092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Skrzynski CJ, Creswell KG. Associations between solitary drinking and increased alcohol consumption, alcohol problems, and drinking to cope motives in adolescents and young adults: a systematic review and meta-analysis. Addiction 2020;115:1989–2007. 10.1111/add.15055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Skrzynski CJ, Creswell KG. A systematic review and meta-analysis on the association between solitary drinking and alcohol problems in adults. Addiction 2021;116:2289–303. 10.1111/add.15355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- StataCorp . Stata Statistical Software: Release 18. College Station, TX: StataCorp LLC, 2023. [Google Scholar]
- Stickley A, Koyanagi A, Roberts B. et al. Male solitary drinking and hazardous alcohol use in nine countries of the former Soviet Union. Drug Alcohol Depend 2015;150:105–11. 10.1016/j.drugalcdep.2015.02.017. [DOI] [PubMed] [Google Scholar]
- VicHealth . VicHealth Coronavirus Victorian Wellbeing Impact Study. Report for survey #1:. Melbourne: Victorian Health Promotion Foundation, 2020, 10.1007/978-3-030-14449-4_33-1. [DOI] [Google Scholar]
- Villanueva-Blasco VJ, Villanueva Silvestre V, Vázquez-Martínez A. et al. Age and living situation as key factors in understanding changes in alcohol use during COVID-19 confinement. Int J Environ Res Public Health 2021;18:11471. 10.3390/ijerph182111471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wardell JD, Kempe T, Rapinda KK. et al. Drinking to cope during COVID-19 pandemic: the role of external and internal factors in coping motive pathways to alcohol use, solitary drinking, and alcohol problems. Alcohol Clin Exp Res 2020;44:2073–83. 10.1111/acer.14425. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data underlying this article cannot be shared publicly as the authors do not have permission to share the data.