Changes in alcohol consumption during the COVID-19 pandemic: a longitudinal cohort study using smart-breathalyzer data

Abstract

Previous studies relying on alcohol sales, alcohol-related injuries, and surveys have suggested that alcohol consumption increased during the COVID-19 pandemic. We sought to leverage over 1 million Breath Alcohol Concentration (BrAC) measurements from Bluetooth-enabled breathalyzers to conduct an objective and longitudinal assessment of alcohol use during the pandemic. Serial BrAC measurements revealed a decrease in drinking between January 1, 2020 and March 30, 2020, an increase between March 30, 2020 and May 25, 2020, a statistically insignificant decrease between May 25, 2020 and January 1, 2021, and an increase again between January 1, 2021 and June 4, 2021. No statistically significant relationships between shelter-in-place orders and alcohol consumption were detected. These findings demonstrate the complex relationship between the pandemic and alcohol consumption patterns, providing insights that may be relevant to the use of this commonly consumed substance with implications relevant to long-term effects from the patterns observed.

Figure 1

Total Number of BrAC observations made, grouped by United States counties. The value within each county on the U.S. map represents the total number of BrAC recordings measured within the county between January 1, 2019, and June 4, 2021. Grey represents counties where no BrAC measurements were made. This map was generated by the studio authors using R (Vienna, Austria) with the urbnmapr package (https://urbaninstitute.github.io/urbnmapr/).

Figure 2

Total number of BrAC recordings made by calendar week. Each blue box represents the total number of BrAC measurements recorded within each calendar week between January 1, 2019, and June 4, 2021. The area shaded red represents the dates between March 15, 2020 and June 15, 2020 where “Mandatory for All Individuals” Shelter-in-Place orders were placed in at least one U.S. county.

Figure 3

Weekly BrAC proportions between January 1, 2020, and June 4, 2021. The triangles in the figure represent the proportion of BrAC recordings detecting any alcohol consumption (> 0 g/dL) to the weekly proportion of BrAC recordings revealing no alcohol (0 g/dL); and the squares represent the weekly proportion of BrAC recordings at or above the legal driving limit (≥ 0.08 g/dL) to the proportion under the legal driving limit (BrAC recordings < 0.08 g/dL). Four distinct time periods emerged (P.1–P.4) following the beginning of the SARS-CoV-2 outbreak were used for our analysis and are represented between dashed lines. The dates for each time period are summarized in Table 1.

Figure 4

Distribution of Shelter-In-Place orders compared to the weekly proportion of BrAC measurements > 0 g/dL. The open triangles represent the proportion of Breath Alcohol Concentration (BrAC) measurements above 0 g/dL to the proportion of measurements equal to 0 g/dL. The histogram represents the number of U.S. counties with the following Shelter-In-Place orders: red denotes “mandatory for all individuals”; blue denotes “mandatory only for at-risk individuals in the jurisdiction”; black denotes “mandatory only for all individuals in certain areas of the jurisdiction”; orange denotes “advisory/recommendation”; green denotes “mandatory only for at-risk individuals in certain areas of the jurisdiction.