Abstract
The early phases of the coronavirus 19 disease (COVID-19) pandemic were associated with changes in psychological wellbeing and alcohol use. However, it is unclear whether these changes are artifacts of psychological wellbeing and alcohol use prior to the pandemic across different socio-demographic groups. We received surveys from 247 adult residents of Allegheny County, Pennsylvania (United States), with an oversampling of sexual and gender minority individuals. Responses included measures of psychological wellbeing, substance use, and socio-demographic characteristics. Unadjusted mean depression scores, anxiety scores, and number of drinking days increased for all age and income groups during COVID-19, while average number of drinks per drinking day and days intoxicated differentially increased or decreased by age and income groups. Using Bayesian seemingly unrelated regression, we assessed depression and anxiety symptoms and alcohol use during the early stages of the pandemic and one month before COVID-19 was first identified in Allegheny County concurrently. Those in the youngest (18–24) group drank on more days during (but not before) the pandemic than those in the 25–44 age group. Compared to cisgender women, gender minority adults had higher depression scores during the early stages of the pandemic. Employed adults had lower anxiety scores during (but not before) the pandemic than adults who were unemployed. Those with past-year annual incomes above $80,000 had fewer drinks on average drinking occasions than those in the $40,000 or below group before (but not during) the pandemic. Patterns of psychological distress and alcohol use associated with the COVID-19 pandemic differ by subgroup compared to patterns prior to the pandemic. Interventions addressing worsening mental health outcomes and shifting alcohol use patterns must be sensitive to the needs of vulnerable groups, such as younger adults and those experiencing poverty or unemployment.
Keywords: COVID-19, sexual and gender minorities, alcohol use, depression, anxiety
Introduction
The health effects of the coronavirus 19 disease (COVID-19) pandemic reach beyond COVID-19-specific mortality and morbidity and include decreased psychological wellbeing and changing substance use behaviors.1–4 In general, quarantine and social isolation are associated with negative affective symptoms, and exacerbated by stressors such as financial burdens, lack of resources, and stigma.5,6 Particularly before the availability of vaccines, many people faced vast uncertainty about their social, physical, and financial futures. Unsurprisingly, psychological distress, overall and coronavirus-specific anxiety, depression, and sleep disturbances all increased during the early phases of the COVID-19 pandemic.1–3,6–12
In Pennsylvania, efforts to contain the spread of COVID-19 began in mid-March, 2020. Non-essential businesses were required to close on March 19th, and stay-at-home orders were sequentially ordered across the state shortly thereafter. In Allegheny County, stay-at-home orders were mandated on March 23rd. This constituted the beginning of the “red phase,” which prohibited large gatherings, required masks in public indoor spaces, and permitted only life-sustaining businesses to remain open. Pennsylvania’s “Fine Wine & Good Spirits” stores, which hold on a monopoly on state liquor sales, were ordered to close. The “yellow phase,” which allowed the re-opening of schools and certain businesses (including state liquor stores) under strict safety guidelines, was instituted on May 15. This was followed by the “green phase” on June 5, 2020; at this time, most businesses were allowed to re-open with capacity limitations. In this context, it is of great interest to assess both the psychological wellbeing and alcohol use shifts of Allegheny County residents.
Many early publications identified important heterogeneities in the relationships between the COVID-19 pandemic and mental health13, with certain groups particularly vulnerable to adverse mental health experiences. Those with pre-existing anxiety and mood disorders were at especially high risk.8,14–17 Sexual and gender minority (SGM) adults (i.e., non cis-gender heterosexual adults) have historically been vulnerable to poorer mental health and substance use outcomes than cis-gender heterosexual adults, and these disparities were likely exacerbated during the COVID-19 pandemic.18,19 Other vulnerable groups included younger people, the elderly, essential and migrant workers, mothers of young children, women in general, single individuals, those experiencing pay cuts and job loss, and people who use more drugs and alcohol than the general population.1,7–11,15,16,20–26 Conversely, higher income, economic stability, older age, and social support were associated with less severe mental health hurdles.7,16,27
Nonmedical use of drugs and alcohol changed in the early phases of the COVID-19 pandemic, with studies reporting both increases and decreases across different subgroups.1,4,15,28 Alcohol availability rapidly shifted after the first wave of shelter-in-place restrictions in March 2020 in U.S. regions. The initial surge of online alcohol sales potentially offset the closure of bars, restaurants, and off-premise alcohol outlets.29 The impacts of these shifts, as well as the effects of stressors and changes in daily life, likely were not uniform. For example, females appeared to increase their drinking more than males and in response to different factors: among females, increased heavier drinking was most strongly connected to childcare stressors, while among males, it was more commonly associated with job loss or reductions in paid employment hours.4 Older adults had poorer substance use outcomes than younger adults.30 Men who have sex with men used more recreational drugs and drank more during the COVID-19 pandemic than before, thus potentially exacerbating existing disparities.31 Vulnerabilities for relapse or increased alcohol use among those with a history of substance use disorders can escalate as healthcare and addiction services become limited.32–34
Prior to the pandemic, factors related to greater risk for depression and anxiety included being female and living in poverty.35,36 Though different age groups are prone to varying stressors and vulnerabilities, younger and middle-aged adults tended to have higher rates of anxiety and depression.35,37 Of all age groups, young adults (particularly ages 21–25) tended to have the highest rates of heavy and binge drinking prior to the pandemic.38 Crucially, identifying emerging high-risk groups during the early phases of the COVID-19 pandemic requires reconciling the findings reported both during and before the pandemic. This necessitates accounting for the respondent-level correlations between outcome measures before and during the pandemic. In this paper, we explore associations between socio-demographic factors and depression, anxiety, and alcohol-related outcomes during the period following the first diagnosed COVID-19 case and associated stay-at-home orders in Allegheny County, Pennsylvania (April-June 2020). We also measure each of these outcomes in the month preceding the first COVID-19 case in Allegheny County (February 2020) and simultaneously assess correlates of mental health and alcohol use before and during the COVID-19 pandemic. Identifying subgroups at risk for psychological distress and substance use, especially those for whom the risk was exacerbated upon the onset of the COVID-19 pandemic, provides useful information for intervention development and service provision during the early phases of global health crises.
Methods
Study Design and Sample
We conducted an internet-based survey between April and June 2020 that included adults living in Allegheny County, Pennsylvania. Participants were recruited using Facebook and Instagram advertisements posted by social media accounts affiliated with the University of Pittsburgh’s Center for Social Dynamics and Community Health and Center for LGBT Health Research. One set of advertisements was targeted to a general pool of Allegheny County adult residents, while a second set was targeted to SGM residents. The goal was to recruit approximately 50% SGM study participants as this is a vulnerable group potentially more impacted by the COVID-19 pandemic. A total of 247 respondents completed the survey (n=123 SGM adults). Depending on the outcome, complete data were available for 86% (n=212) to 90% (n=222) of the respondents, creating a total analytic sample of 224 adults. There were no statistically significant differences at the α = 0.05 level among those with complete and missing data on any of the predictors used in our models.
This study was approved by the University of Pittsburgh Institutional Review Board. Participants provided informed consent prior to beginning the survey.
Measures
Respondents were asked a series of questions about two time periods: during the month of February 2020 (before the first case of COVID-19 was diagnosed in Allegheny County), and during the month leading up to the day when each respondent took the survey (between April and June 2020). Outcome measures included positive screening for anxiety and depression in the preceding two weeks using the Patient Health Questionnaire-4 (PHQ-4), a four-item scale of two questions each screening for depression and anxiety.39 The PHQ-4 uses a Likert-type scale to score each response from 0 to 3, with a possible range of depression and anxiety scores from 0 to 6. Alcohol use measures reflected the past 4 weeks and included the number of days during which alcohol was consumed (range: 0–28 days), the average number of drinks on days when alcohol was consumed (range: 0–30), and the number of days the respondent was intoxicated (range: 0–28 days). Each outcome measure was assessed for both time periods.
Socio-demographic variables assessed in the survey included age (categorized in our analyses as: 18–24, 25–44, 45–64, and 65–89), race/ethnicity (non-Hispanic Black/African American, non-Hispanic White, other/more than one race/ethnicity), sexual identity (heterosexual vs. sexual minority, which included gay or lesbian, bisexual, queer, or another non-heterosexual identity), gender identity (cisgender woman, cisgender man, or gender minority, which included transgender woman, transgender man, genderqueer, non-binary, or another gender identity), and marital status (married or living with someone in a marriage-like partnership vs. not). Economic measures included past-year household income (categorized as $40,000 or less, $40,001-$80,000, greater than $80,000) and current employment status (working full-time/part-time vs. not currently employed).
Data Analysis Procedures
We first described outcome measures and covariates included in our analyses. Outcome measures are further described by age group (18–24, 25–44, 45–64, and 65–89) and by past-year (2019) household income ($40,000 or less, between $40,001 and $80,000, and greater than $80,000).
We next assessed covariates relating to outcomes both before and during the stay-at-home orders related to the COVID-19 pandemic that were implemented in Allegheny County. We used Bayesian seemingly unrelated regression (SURE) models, run using the R package R-INLA, for our analyses.40–42 As these were Bayesian analyses, we reported the median of each effect’s posterior distribution along with 95% credible intervals, which corresponded to the cutoff values at which 2.5% and 97.5% of each effect’s posterior distribution fell. Effects were considered “well-supported” if both cutoff values fell either above or below 0.
Our models simultaneously estimated outcomes before and during the COVID-19 pandemic (Table 3). We estimated five sets of models using measures of psychological distress (anxiety, depression) and alcohol consumption (number of drinking days, mean number of drinks per drinking days, number of days intoxicated) as dependent variables. For anxiety and depression, whose scores represented continuous measures of each variable, we used Gaussian regression. For alcohol consumption measures, representing discrete numbers of days or drinks, we used Poisson regression. Models for mean number of drinks per day and number of intoxicated days only included respondents who had at least one drinking day during at least one of the assessed time periods (n=150). To account for correlation between the two time points for each participant, we incorporated independent and identically distributed Gaussian random effects at the respondent level as well as ρ, corresponding to correlations between each respondent’s two sets of responses. This allowed for separate assessment of the two time points while accounting for the respondent-level correlations. We adjusted all models for age, race/ethnicity, gender identity, sexual identity, past-year income, current employment status, and education status. Categorical variables corresponding to the Allegheny County stay-at-home orders phase (i.e., “red,” “yellow,” or “green”) at the time of each participant taking the survey were not significantly associated with any of the outcomes and were thus excluded from final models.
Table 3.
Posterior median estimates and 95% credible intervals for mental health outcomes (depression, anxiety) using Gaussian-distributed Bayesian seemingly unrelated regression models.
| Depression | Anxiety | |||
|---|---|---|---|---|
| n = 214 | n = 213 | |||
| February 2020 | During the COVID-19 pandemic | February 2020 | During the COVID-19 pandemic | |
| Fixed effects | Median [95% CI] | Median [95% CI] | Median [95% CI] | Median [95% CI] |
| Age (referent: 25–44) | ||||
| 18–24 | 0.19 [−0.40, 0.78] | 0.53 [−0.13, 1.20] | 0.44 [−0.20, 1.08] | 0.24 [−0.47, 0.94] |
| 45–64 | −0.43 [−0.94, 0.08] | −0.53 [−1.10, 0.05] | −0.14 [−0.70, 0.41] | −0.56 [−1.18, 0.06] |
| 65–89 | −0.95 [−1.57, −0.33]a | −0.88 [−1.59, −0.18]a | −1.07 [−1.73, −0.41]a | −1.39 [−2.12, −0.66]a |
| Race/ethnicity (referent: non-Hispanic White) | ||||
| Non-Hispanic Black/African American | 1.12 [0.29, 1.95]a | 0.36 [−0.58, 1.30] | −0.18 [−1.07, 0.72] | −1.14 [−2.13, −0.15]a |
| Other race/ethnicity (including multi-racial, Hispanic) | −0.34 [−1.00, 0.32] | −0.69 [−1.44, 0.05] | 0.34 [−0.36, 1.05] | −0.58 [−1.36, 0.20] |
| Gender (referent: cisgender woman) | ||||
| Cisgender man | −0.09 [−0.55, 0.36] | 0.19 [−0.33, 0.70] | −0.42 [−0.91, 0.07] | −0.19 [−0.73, 0.35] |
| Gender minority | 0.04 [−0.56, 0.63] | 0.79 [0.12, 1.47]a | −0.12 [−0.76, 0.51] | 0.52 [−0.18, 1.23] |
| Sexual minority (referent: heterosexual) | 0.24 [−0.22, 0.71] | 0.31 [−0.22, 0.83] | 0.70 [0.20, 1.21]a | 0.30 [−0.26, 0.86] |
| Past-year income (referent: $0-$40,000) | ||||
| $40,001-$80,000 | −0.25 [−0.69, 0.19] | −0.07 [−0.58, 0.43] | −0.22 [−0.70, 0.25] | −0.34 [−0.87, 0.18] |
| $80,001+ | −0.21 [−0.69, 0.27] | −0.66 [−1.20, −0.11]a | −0.51 [−1.03, 0.01] | −1.15 [−1.73, −0.57]a |
| Working full-time or part-time (referent: not working full-time or part-time) | −0.44 [−0.82, −0.06]a | −0.69 [−1.12, −0.26]a | −0.19 [−0.60, 0.22] | −0.45 [−0.90, −0.00]a |
| Higher education (referent: high school or less) | −0.19 [−0.69, 0.30] | −0.12 [−0.68, 0.44] | −0.21 [−0.75, 0.33] | −0.43 [−1.03, 0.17] |
| Random effects | ||||
| Respondent ID, precision | 0.52 [0.43, 0.63] | 0.41 [0.34, 0.49] | 0.46 [0.38, 0.55] | 0.37 [0.31, 0.45] |
| ρ for respondent ID | 0.49 [0.39, 0.59] | 0.61 [0.52, 0.68] | ||
Well-supported as indicated by a credible interval that excludes 0.00.
CI = credible interval
Results
Respondents with data on at least one outcome measure included 224 Allegheny County residents whose ages ranged from 18 to 89 (mean 45.3, SD=19.5) (Table 1). Approximately 5% (n=12) of respondents were non-Hispanic Black, and 8% (n=18) were Asian, Hispanic, or multi-racial; the rest were non-Hispanic White. The majority (58.9%, n=132) of respondents were cisgender women, 25.9% (n=58) were cisgender men, and 15.2% (n=34) identified as gender minorities. Approximately half of the respondents (49.1%, n=110) were sexual minorities. Most cisgender men (65.5%, n=38) were sexual minorities, while most cisgender women (70.5%, n=93) were heterosexual. A large minority (40.2%, n=90) of respondents reported a past-year household income of $40,000 or less, 34.8% (n=78) earned between $40,001 and $80,000, and 25.0% (n=56) had incomes greater than $80,000. About half of the sample was employed (49.6%, n=111; full-time or part-time) at the beginning of the pandemic, while the other half was not working (e.g., retired, laid off, or unemployed). Most participants (82.1%, n=184) had some college education or more.
Table 1.
Overall descriptive statistics (n = 224 survey respondents).
| Variable | n | Mean (SD) |
|---|---|---|
| Depression | ||
| In February 2020 | 215 | 1.5 (1.5) |
| During the COVID-19 pandemic | 217 | 2.2 (1.8) |
| Anxiety | ||
| In February 2020 | 214 | 1.6 (1.6) |
| During the COVID-19 pandemic | 217 | 2.2 (1.9) |
| Number of drinking days in the past 28 days | ||
| In February 2020 | 222 | 3.8 (5.6) |
| During the COVID-19 pandemic | 224 | 4.9 (7.1) |
| Average number of drinks per drinking day in the past 28 days | ||
| In February 2020 | 129 | 2.3 (2.8) |
| During the COVID-19 pandemic | 139 | 2.4 (3.8) |
| Number of days intoxicated in the past 28 days | ||
| In February 2020 | 126 | 1.4 (2.4) |
| During the COVID-19 pandemic | 139 | 1.2 (3.2) |
| Age | 224 | 45.3 (19.5) |
| Variable | n (%) | |
| Race/ethnicity | ||
| Non-Hispanic Black/African American | 12 (5.4) | |
| Non-Hispanic White | 194 (86.6) | |
| Other (including multi-racial, Hispanic) | 18 (8.0) | |
| Gender identity | ||
| Cisgender man | 58 (25.9) | |
| Cisgender woman | 132 (58.9) | |
| Gender minority | 34 (15.2) | |
| Sexual identity | ||
| Heterosexual | 114 (50.9) | |
| Sexual minority | 110 (49.1) | |
| 2019 household income | ||
| $40,000 or less | 90 (40.2) | |
| $40,001-$80,000 | 78 (34.8) | |
| More than $80,000 | 56 (25.0) | |
| Employment status during the COVID-19 pandemic | ||
| Working full-time or part-time | 111 (49.6) | |
| Not working full-time or part-time | 113 (50.5) | |
| Education | ||
| Some college or more | 184 (82.1) | |
| Completed high school or less | 40 (17.9) |
While mean depression and anxiety scores increased from before (February 2020) the COVID-19 pandemic to its early phases (April-June 2020) for all age groups, the youngest participants (ages 18–24) had the highest mean depression and anxiety scores at both time points, while participants in the oldest age group (65–89) had the lowest mean scores (Table 2). The mean number of drinking days in the past 28 days increased for all age groups between February and April-June. However, among those who had at least one drink at either time point, the average number of drinks and the number of days intoxicated decreased for the two younger age groups (e.g., for the 18–24 group, the mean number of drinks decreased from 2.5 to 1.8, and the mean number of intoxicated days from 2.4 to 1.5) but increased for the oldest age group (i.e., from 3.3 to 4.2 mean number of drinks, and from 0.5 to 1.8 mean intoxicated days). Mean mental health scores and number of drinking days likewise increased for all income groups from February to the later time point, while among drinkers, the average number of drinks per drinking day and the number of intoxicated days slightly decreased for those with 2019 annual household incomes of $40,000 or less but increased for those with incomes greater than $80,000 (Table 2).
Table 2.
Descriptive statistics by age group and income.
| Age group | Past-year household income | ||||||
|---|---|---|---|---|---|---|---|
| 18–24 | 25–44 | 45–64 | 65–89 | ≤$40,000 | ≤$40,000 | ≤$40,000 | |
| All respondents | n = 54 | n = 53 | n = 71 | n = 46 | n = 90 | n = 78 | n = 56 |
| Variable | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) |
| Depression | |||||||
| In February 2020 | 2.1 (1.7) | 1.56 (1.7) | 1.24 (1.2) | 0.77 (1.0) | 1.71 (1.4) | 1.27 (1.5) | 1.30 (1.7) |
| During the COVID-19 pandemic | 3.3 (1.9) | 2.15 (1.7) | 1.76 (1.5) | 1.46 (1.4) | 2.49 (1.6) | 2.18 (1.9) | 1.73 (1.7) |
| Anxiety | |||||||
| In February 2020 | 2.6 (1.8) | 1.7 (1.8) | 1.39 (1.2) | 0.59 (1.0) | 1.90 (1.7) | 1.50 (1.6) | 1.33 (1.6) |
| During the COVID-19 pandemic | 3.2 (1.9) | 2.4 (2.0) | 1.77 (1.5) | 1.24 (1.5) | 2.55 (1.9) | 2.21 (2.0) | 1.57 (1.5) |
| Number of drinking days in the past 28 days | |||||||
| In February 2020 | 3.9 (4.6) | 4.2 (5.7) | 4.46 (6.3) | 2.34 (5.3) | 3.43 (5.2) | 3.55 (4.8) | 4.88 (7.0) |
| During the COVID-19 pandemic | 5.5 (6.3) | 5.6 (7.4) | 5.01 (7.6) | 3.07 (6.7) | 4.24 (6.6) | 5.14 (6.9) | 5.50 (8.1) |
| Drinkers only | n = 37 | n = 37 | n = 54 | n = 22 | n = 58 | n = 54 | n = 37 |
| Average number of drinks per drinking day in the past 28 days | |||||||
| In February 2020 | 2.48 (1.3) | 2.03 (1.3) | 2.13 (1.4) | 3.25 (7.3) | 2.76 (4.2) | 2.28 (1.4) | 1.79 (1.2) |
| During the COVID-19 pandemic | 1.81 (0.9) | 1.97 (1.1) | 2.48 (4.2) | 4.22 (7.6) | 2.59 (4.6) | 2.47 (4.0) | 2.03 (1.4) |
| Number of days intoxicated in the past 28 days | |||||||
| In February 2020 | 2.42 (2.3) | 2.00 (3.3) | 0.61 (1.4) | 0.50 (1.5) | 1.61 (2.1) | 1.82 (3.2) | 0.67 (1.3) |
| During the COVID-19 pandemic | 1.49 (1.9) | 1.47 (3.4) | 0.56 (1.5) | 1.78 (6.6) | 1.41 (4.1) | 1.28 (3.0) | 0.71 (1.4) |
Median within-respondent correlations (ρ) ranged from 0.49 to 0.96 for the five dependent measures (Table 3). In adjusted models, compared with those in the 25–44 age group, the 65–89 age group had lower mean scores of both depression and anxiety, effects that were well-supported for both outcomes at both time points (Table 3). Compared with White participants, Black participants reported higher depression scores in February (but not during the COVID-19 pandemic), and lower anxiety scores during the pandemic (but not in February). Gender minorities had higher mean depression scores during the pandemic, but this trend was not well-supported in the anxiety model. Higher past-year income was connected to lower mean depression and anxiety scores during the pandemic (e.g., compared to those with incomes of $40,000 or less in 2019, those with incomes higher than $80,000 had anxiety scores that were lower by 1.15, 95% CI: −1.73, −0.57). While a similar trend was seen during February, this effect was not well-supported. Being employed at the time of the survey was connected to lower depression scores during both time points and lower anxiety scores during the COVID-19 pandemic.
In adjusted alcohol use models, those in the 18–24 age group drank on 0.84 (95% CI: 0.02, 1.66) more days during the pandemic than those in the 25–44 age group (Table 4). The youngest age group also reported substantially more intoxicated days than the 25–44 age group in February, an effect that was not well-supported during the pandemic. Compared to the referent age group, those in the 45–64 group had fewer intoxicated days during both time periods. Those in the oldest age group had fewer intoxicated days in February, but this trend was not well-supported during the pandemic. Being employed and having some college education or more were both associated with drinking on more days during both time periods.
Table 4.
Posterior median estimates and 95% credible intervals for alcohol consumption outcomes using Poisson-distributed Bayesian seemingly unrelated regression models.
| Number of drinking days | Number of drinks per drinking day | Number of days intoxicated | ||||
|---|---|---|---|---|---|---|
| n = 222 | n = 124 | n = 121 | ||||
| February 2020 | During the COVID-19 pandemic | February 2020 | During the COVID-19 pandemic | February 2020 | During the COVID-19 pandemic | |
| Fixed effects | Median [95% CI] | Median [95% CI] | Median [95% CI] | Median [95% CI] | Median [95% CI] | Median [95% CI] |
| Age (referent: 25–44) | ||||||
| 18–24 | 0.72 [−0.08, 1.53] | 0.84 [0.02, 1.66]a | 0.24 [−0.28, 0.77] | 0.26 [−0.30, 0.83] | 1.30 [0.10, 2.51]a | 1.24 [−0.13, 2.63] |
| 45–64 | 0.61 [−0.07, 1.30] | 0.41 [−0.29, 1.11] | 0.20 [−0.24, 0.65] | 0.08 [−0.37, 0.54] | −1.32 [−2.38, −0.26]a | −1.34 [−2.55, −0.13]a |
| 65–89 | −0.40 [−1.24, 0.44] | −0.41 [−1.26, 0.44] | −0.07 [−0.69, 0.54] | 0.33 [−0.28, 0.94] | −2.59 [−4.35, −0.97]a | −0.88 [−2.60, 0.74] |
| Race/ethnicity (referent: non-Hispanic White) | ||||||
| Non-Hispanic Black/African American | −1.63 [−2.96, −0.37]a | −1.82 [−3.17, −0.53]a | 0.38 [−0.78, 1.46] | 0.17 [−1.19, 1.40] | 2.18 [−0.30, 4.64] | 2.76 [−0.03, 5.53] |
| Other race/ethnicity (including multi-racial, Hispanic) | −1.08 [−2.03, −0.14]a | −0.73 [−1.67, 0.21] | −0.49 [−1.12, 0.10] | −0.42 [−1.11, 0.23] | −1.03 [−2.41, 0.30] | −0.68 [−2.27, 0.84] |
| Gender (referent: cisgender woman) | ||||||
| Cisgender man | 0.28 [−0.33, 0.88] | 0.33 [−0.29, 0.94] | −0.01 [−0.41, 0.39] | 0.51 [0.10, 0.92]a | 1.00 [0.08, 1.93]a | 0.93 [−0.13, 1.98] |
| Gender minority | 0.22 [−0.60, 1.03] | 0.41 [−0.41, 1.23] | −0.14 [−0.64, 0.36] | 0.25 [−0.29, 0.79] | −0.31 [−1.45, 0.80] | −0.06 [−1.35, 1.20] |
| Sexual minority (referent: heterosexual) | 0.34 [−0.29, 0.97] | 0.15 [−0.49, 0.79] | 0.08 [−0.33, 0.49] | −0.37 [−0.79, 0.06] | −0.43 [−1.40, 0.54] | −0.29 [−1.39, 0.81] |
| 2019 income (referent: $0-$40,000) | ||||||
| $40,001-$80,000 | −0.03 [−0.64, 0.57] | 0.23 [−0.38, 0.83] | −0.24 [−0.64, 0.17] | 0.07 [−0.36, 0.50] | −0.27 [−1.23, 0.69] | 0.31 [−0.79, 1.41] |
| $80,001+ | −0.30 [−0.96, 0.35] | −0.24 [−0.91, 0.43] | −0.52 [−0.95, −0.10]a | −0.24 [−0.69, 0.20] | −1.06 [−2.12, −0.02]a | −0.23 [−1.41, 0.92] |
| Working full-time or part-time (referent: not working full-time or part-time) | 0.93 [0.42, 1.43]a | 0.98 [0.47, 1.50]a | −0.02 [−0.34, 0.31] | 0.08 [−0.26, 0.42] | 0.16 [−0.63, 0.95] | −0.07 [−0.96, 0.83] |
| Higher education (referent: high school or less) | 1.44 [0.75, 2.14]a | 1.15 [0.47, 1.85]a | 0.27 [−0.23, 0.78] | 0.24 [−0.28, 0.78] | 0.34 [−0.84, 1.58] | 0.58 [−0.78, 2.03] |
| Random effects | ||||||
| Respondent ID, precision | 0.34 [0.25, 0.47] | 0.33 [0.24, 0.44] | 2.93 [1.89, 4.52] | 2.47 [1.64, 3.71] | 0.37 [0.21, 0.64] | 0.29 [0.16, 0.52] |
| ρ for respondent ID | 0.96 [0.94, 0.98] | 0.76 [0.56, 0.87] | 0.93 [0.84, 0.97] | |||
Well-supported as indicated by a credible interval that excludes 0.00.
CI = credible interval
Discussion
In this study, mental health and alcohol use outcomes were simultaneously assessed at two time points: during the initial phase of the COVID-19 pandemic (i.e., April to June 2020) and before (i.e., in February 2020) the COVID-19 pandemic arrived in Allegheny County, Pennsylvania. Consistent with other studies, our survey participants generally reported increases in psychological distress upon the onset of the pandemic.1–3,10–12 Alcohol consumption trends in our survey sample were less straightforward to interpret than the mental health trends. While all participants generally drank on more days per month, prior to adjusting for other variables, older drinkers reported more intoxicated days and mean number of drinks per day during the pandemic compared to during February 2020 (Table 2); the opposite was true for younger drinkers. After adjusting for other characteristics, those in the youngest age group drank and were intoxicated on more days than those in older age groups in February 2020 but not during the early phases of the pandemic. In both adjusted and unadjusted analyses, those in the oldest age group were, on average, less depressed and less anxious than respondents younger than 45 during both time points.
As other early studies have reported, economic factors were important predictors of mental health and drinking outcomes. Individuals employed at the time of taking the survey were less psychologically distressed during the COVID-19 pandemic than those who were not employed. While a similar association was apparent in February 2020, it was smaller than after the arrival of the pandemic. Higher past-year income was also associated with lower depression and anxiety scores during the pandemic, placing lower-income individuals at greater mental health risk. The relationship between economic factors and drinking measures was less clear-cut. Prior to the pandemic, incomes of over $80,000 per year were associated with drinking fewer mean drinks per drinking day and fewer intoxicated days compared to incomes of $40,000 or less. However, this association was not well-supported during the pandemic. Being employed and more highly educated were both associated with drinking on more days during both time periods. Other literature has similarly found that while those with higher incomes tended to drink on more days, they did not experience more drinking-related problems.38
There are several trends that our study did not support. For example, in contrast to other early studies, cisgender women in our sample did not fare substantially worse on any of the outcomes.7,9,10,6,21,22,25 This may be, in part, because other studies did not oversample SGM adults, whose needs and experiences during the early phases of the COVID-19 pandemic have been underexplored. While most cisgender women in our sample were heterosexual, most cisgender men were sexual minorities, which may confound direct comparisons based on gender. Compared to their heterosexual and cisgender peers, sexual and gender minority adults commonly experience worse mental health outcomes and more alcohol misuse.18 Interactive factors from discrimination to economic stressors may further exacerbate these outcomes early in the pandemic.44,45 For example, our findings indicate that gender minorities had higher depression scores during the first months of the pandemic than cisgender individuals.
There are several limitations to our analyses. First, our sample size was fairly small, which may be connected to the lack of support for some of the trends noted in our results. The proportion of racial and ethnic minorities in our sample was also small; Black respondents (5.4%) were underrepresented relative to the proportion of Black residents in Allegheny County (12.5%). Second, as we collected self-report measures, responses were subject to recall bias, particularly in retrospective measures referencing February 2020. Outcome measures were not verified using clinical interviews and were also subject to reporting bias. Third, we did not assess pharmacotherapy use, a potentially important confounder, among participants. Finally, participants were recruited using Facebook and Instagram, which introduced self-selection bias and undercoverage, thus limiting the generalizability of results. Our study was designed to survey the early trends of the evolving repercussions of the COVID-19 pandemic. Despite several limitations, our results indicate notable preliminary trends, which will likely continue to shift as the pandemic evolves. Importantly, our sample oversampled SGM adults, whose experiences during the COVID-19 pandemic have not been widely studied.
Conclusions
As changes in alcohol use patterns and mental health outcomes varied by subgroup, interventions should consider risks specific to economic and age-related factors. Interventions must also heed worsening mental health outcomes associated with the pandemic, with particular sensitivity to the needs of vulnerable groups, such as SGM individuals, those experiencing poverty, and younger individuals. Furthermore, as long-lasting effects of quarantine measures have been noted, these findings are important to consider as the COVID-19 pandemic and its implications continue to evolve. Future research should more closely explore mental health and alcohol use across all phases of the pandemic, particularly among vulnerable populations.
Funding information:
This project was supported by the National Institute on Alcohol Abuse and Alcoholism (NIAAA) grant K01AA027564, National Institutes of Health (NIH) grant P60AA006282, and the National Institute on Drug Abuse (NIDA) grant R25DA050687.
Footnotes
Competing interests: None to declare
Data availability:
Due to the senstive nature of the content of this research, participants of this study did not agree for their data to be shared publicly, so supporting data are not available.
References
- 1.Czeisler MÉ, Lane RI, Petrosky E, et al. Mental Health, Substance Use, and Suicidal Ideation During the COVID-19 Pandemic — United States, June 24–30, 2020. MMWR Morb Mortal Wkly Rep. 2020;69(32):1049–1057. doi: 10.15585/mmwr.mm6932a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Marroquín B, Vine V, Morgan R. Mental Health During the COVID-19 Pandemic: Effects of Stay-at-Home Policies, Social Distancing Behavior, and Social Resources. Psychiatry Research. Published online August 20, 2020:113419. doi: 10.1016/j.psychres.2020.113419 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Twenge JM, Joiner TE. U.S. Census Bureau-assessed prevalence of anxiety and depressive symptoms in 2019 and during the 2020 COVID-19 pandemic. Depression and Anxiety. doi: 10.1002/da.23077 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Biddle PN, Sollis K. Alcohol consumption during the COVID-19 period: May 2020. Published online 2020:19. [Google Scholar]
- 5.Brooks SK, Webster RK, Smith LE, et al. The psychological impact of quarantine and how to reduce it: rapid review of the evidence. The Lancet. 2020;395(10227):912–920. doi: 10.1016/S0140-6736(20)30460-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lee SA. Coronavirus Anxiety Scale: A brief mental health screener for COVID-19 related anxiety. Death Studies. 2020;44(7):393–401. doi: 10.1080/07481187.2020.1748481 [DOI] [PubMed] [Google Scholar]
- 7.Holingue C, Badillo-Goicoechea E, Riehm KE, et al. Mental distress during the COVID-19 pandemic among US adults without a pre-existing mental health condition: Findings from American trend panel survey. Preventive Medicine. 2020;139:106231. doi: 10.1016/j.ypmed.2020.106231 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Liu CH, Stevens C, Conrad RC, Hahm HC. Evidence for elevated psychiatric distress, poor sleep, and quality of life concerns during the COVID-19 pandemic among U.S. young adults with suspected and reported psychiatric diagnoses. Psychiatry Research. 2020;292:113345. doi: 10.1016/j.psychres.2020.113345 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Xiong J, Lipsitz O, Nasri F. Impact of COVID-19 pandemic on mental health in the general population: A systematic review. J Affect Disord. 2020;277:55–64. doi: 10.1016/j.jad.2020.08.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Santabarbara J, Lasheras I,Lipnicki DM, Bueno-Notivol J, Perez-Moreno M, Lopez-Anton R, et al. Prevalence of anxiety in the COVID-19 pandemic: An updated meta-analysis of community-based studies. Progress in Neuro-pharmocology and Biological Psychiatry 2021;109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wu T, Jia X, Shi H, Niu J, Yin X, Xie J, et al. Prevalence of mental health problems during the COVID-19 pandemic: A systematic review and meta-analysis. Journal of Affective Disorders 2021; 281:91–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cenat JM, Blais-Rochette C, Kokou-Kpolou CK, Noorishad PG, Mukunzi JN, McIntee SE, et al. Prevalence of symptoms of depression, anxiety, insomnia, posttraumatic stress disorder, and psychological distress among populations affected by the COVID-19 pandemic: A systematic review and meta-analysis. Psychiatry Research 2021; 295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Prati G, Mancini AD. The psychological impact of COVID-19 pandemic lockdowns: a review and meta-analysis of longitudinal studies and natural experiments. Psychological Medicine 2021:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Asmundson GG, Paluszek MM, Landry CA, Rachor GS, McKay D, Taylor S. Do pre-existing anxiety-related and mood disorders differentially impact COVID-19 stress responses and coping? Journal of Anxiety Disorders. 2020;74:102271. doi: 10.1016/j.janxdis.2020.102271 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Taylor S, Landry CA, Paluszek MM, Fergus TA, McKay D, Asmundson GJG. COVID stress syndrome: Concept, structure, and correlates. Depression and Anxiety. 2020;37(8):706–714. doi: 10.1002/da.23071 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.González-Sanguino C, Ausín B, Castellanos MÁ, et al. Mental health consequences during the initial stage of the 2020 Coronavirus pandemic (COVID-19) in Spain. Brain, Behavior, and Immunity. 2020;87:172–176. doi: 10.1016/j.bbi.2020.05.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hao F, Tan W, Jiang L, et al. Do psychiatric patients experience more psychiatric symptoms during COVID-19 pandemic and lockdown? A Case-Control Study with Service and Research Implications for Immunopsychiatry. Brain Behav Immun. 2020;87:100–106. doi: 10.1016/j.bbi.2020.04.069 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Institute of Medicine (2011). The health of lesbian, gay, bisexual, and transgender people: Building a foundation for better understanding. Washington, DC: Institute of Medicine, 10, 13128. [PubMed] [Google Scholar]
- 19.van Dorn A, Cooney RE, Sabin ML. COVID-19 exacerbating inequalities in the US. Lancet 2020;395:1243–1244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Qiu J, Shen B, Zhao M, Wang Z, Xie B, Xu Y. A nationwide survey of psychological distress among Chinese people in the COVID-19 epidemic: implications and policy recommendations. Gen Psych. 2020;33(2):e100213. doi: 10.1136/gpsych-2020-100213 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mazza C, Ricci E, Biondi S, et al. A Nationwide Survey of Psychological Distress among Italian People during the COVID-19 Pandemic: Immediate Psychological Responses and Associated Factors. International Journal of Environmental Research and Public Health. 2020;17(9):3165. doi: 10.3390/ijerph17093165 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lai J, Ma S, Wang Y, et al. Factors Associated With Mental Health Outcomes Among Health Care Workers Exposed to Coronavirus Disease 2019. JAMA Netw Open. 2020;3(3):e203976–e203976. doi: 10.1001/jamanetworkopen.2020.3976 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Cameron EE, Joyce KM, Delaquis CP, Reynolds K, Protudjer JLP, Roos LE. Maternal psychological distress & mental health service use during the COVID-19 pandemic. Journal of Affective Disorders. 2020;276:765–774. doi: 10.1016/j.jad.2020.07.081 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Davenport MH, Meyer S, Meah VL, Strynadka MC, Khurana R. Moms Are Not OK: COVID-19 and Maternal Mental Health. Front Glob Womens Health. 2020;1. doi: 10.3389/fgwh.2020.00001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wang C, Pan R, Wan X, et al. Immediate Psychological Responses and Associated Factors during the Initial Stage of the 2019 Coronavirus Disease (COVID-19) Epidemic among the General Population in China. International Journal of Environmental Research and Public Health. 2020;17(5):1729. doi: 10.3390/ijerph17051729 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Martinotti G, Alessi MC, et al. Psychopathological Burden and Quality of Life in Substance Users During the COVID-19 Lockdown Period in Italy. Front Psychiatry. 2020;11:572245. doi: 10.3389/fpsyt.2020.572245 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Chao M, Xue D, Liu T, Yang H, Hall BJ. Media use and acute psychological outcomes during COVID-19 outbreak in China. Journal of Anxiety Disorders 2020;74:102248. doi: 10.1016/j.janxdis.2020.102248 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Schmidt RA, Genois R, Jin J, Vigo D, Rehm J, Rush b. The early impact of COVID-19 on the incidence, prevalence, and severity of alcohol use and other drugs: A systematic review. Drug and Alcohol Dependence 2021;228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.National Institute on Alcohol Abuse and Alcoholism (NIAAA). Digest of State Alcohol-Related Laws During the COVID-19 Emergency for On-Premise and Off-Premise Establishments as of December 15, 2020 (PDF 1.33 MB). Alcohol Policy Information System (APIS). [Google Scholar]
- 30.Satre DD, Hirschtritt ME, Silverberg MJ, Sterling SA. Addressing Problems With Alcohol and Other Substances Among Older Adults During the COVID-19 Pandemic. The American Journal of Geriatric Psychiatry. 2020;28(7):780–783. doi: 10.1016/j.jagp.2020.04.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Sanchez TH, Zlotorzynska M, Rai M, Baral SD. Characterizing the Impact of COVID-19 on Men Who Have Sex with Men Across the United States in April, 2020. AIDS Behav. 2020;24(7):2024–2032. doi: 10.1007/s10461-020-02894-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Da BL, Im GY, Schiano TD. COVID‐19 Hangover: A Rising Tide of Alcohol Use Disorder and Alcohol‐Associated Liver Disease. Hepatology. Published online May 5, 2020:hep.31307. doi: 10.1002/hep.31307 [DOI] [PubMed] [Google Scholar]
- 33.Dubey MJ, Ghosh R, Chatterjee S, Biswas P, Chatterjee S, Dubey S. COVID-19 and addiction. Diabetes & Metabolic Syndrome: Clinical Research & Reviews. 2020;14(5):817–823. doi: 10.1016/j.dsx.2020.06.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Testino G Are Patients With Alcohol Use Disorders at Increased Risk for Covid-19 Infection? Alcohol and Alcoholism. 2020;55(4):344–346. doi: 10.1093/alcalc/agaa037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Brody DJ, Pratt LA, Hughes J. Prevalence of depression among adults aged 20 and over: United States, 2013–2016. NCHS Data Brief, no 303. Hyattsville, MD: National Center for Health Statistics. 2018. [PubMed] [Google Scholar]
- 36.Burt VK, Stein K. Epidemiology of depression throughout the female life cycle. J Clin Psychiatry. 2002;63 Suppl 7:9–15. [PubMed] [Google Scholar]
- 37.Terlizzi EP, Villarroel MA. Symptoms of generalized anxiety disorder among adults: United States, 2019. NCHS Data Brief, no 378. Hyattsville, MD: National Center for Health Statistics. 2020. [PubMed] [Google Scholar]
- 38.Key substance use and mental health indicators in the United States: Results from the 2018 National Survey on Drug Use and Health. Rockville, MD: Center for Behavioral Health Statistics and Quality. Substance Abuse and Mental Health Services Administration (SAMHSA). 2019. [Google Scholar]
- 39.Kroenke K, Spitzer RL, Williams JBW, Löwe B. An ultra-brief screening scale for anxiety and depression: the PHQ-4. Psychosomatics. 2009;50(6):613–621. doi: 10.1176/appi.psy.50.6.613 [DOI] [PubMed] [Google Scholar]
- 40.Keshavarzi S, Ayatollahi SMT, Zare N, Pajfetrat M. Application of Seemingly Unrelated Regression in Medical Data with Intermittently Observed Time-Dependent Covariates. Computational and Mathematical Methods in Medicine. 2012. doi: 10.1155/2012/821643 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.R Core Team (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/. [Google Scholar]
- 42.Rue H, Martino S, Chopin N. 2009. Approximate Bayesian inference for latent Gaussian models using integrated nested Laplace approximations. J R Stat Soc Series B Stat Methodol, 71(2):319–392. [Google Scholar]
- 43.Mair C, Sumetsky N, Gruenewald PJ, Lee JP. Microecological Relationships Between Area Income, Off‐Premise Alcohol Outlet Density, Drinking Patterns, and Alcohol Use Disorders: The East Bay Neighborhoods Study. Alcohol Clin Exp Res. 2020;44(8):1636–1645. doi: 10.1111/acer.14387 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.McGarrity LA. (2014). Socioeconomic status as context for minority stress and health disparities among lesbian, gay, and bisexual individuals. Psychology of Sexual Orientation and Gender Diversity, 1(4), 383. [Google Scholar]
- 45.Meyer IH. (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: conceptual issues and research evidence. Psychological bulletin, 129(5), 674. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Due to the senstive nature of the content of this research, participants of this study did not agree for their data to be shared publicly, so supporting data are not available.