Abstract
Objectives. To examine trends in young adult self-reported driving under the influence of alcohol (DUI-A), cannabis (DUI-C), and simultaneous alcohol and cannabis use (DUI-AC) in a state with legalized nonmedical cannabis use from before to during the COVID-19 pandemic.
Methods. We used logistic regression and annual statewide data from the Washington Young Adult Health Survey to assess DUI behaviors from 2016 to 2021.
Results. Both prepandemic yearly changes in prevalence and deviations from those trends during the pandemic years were small and not statistically significant. However, prevalence estimates were alarming: 12.0% of participants reported DUI-A, 12.5% reported DUI-C, and 2.7% reported DUI-AC. Exploratory moderation analyses indicated a relative increase in DUI-A during 2020 among 4-year college students relative to young adults not attending 4-year colleges.
Conclusions. Young adults in Washington State continued to engage in risky DUI behaviors during the pandemic. College students may have increased their likelihood of DUI-A during COVID-19.
Public Health Implications. Young adults, for whom vehicle crashes remain a leading cause of death, showed little change in DUI behaviors during the COVID-19 pandemic. There is continued need for young adult DUI prevention efforts. (Am J Public Health. 2024;114(S8):S698–S701. https://doi.org/10.2105/AJPH.2024.307767)
Vehicle crashes are a leading cause of death among young adults aged 18 to 25 years, with many crashes caused by alcohol or cannabis impairment.1 Increases in perceived risk and decreases in alcohol use contributed to declines in self-reports of alcohol-impaired driving and related proportions of crash fatalities from the mid-1980s through 2019.2–4 However, self-reports of cannabis-impaired driving have not shown similar declines,3,4 potentially because of legalization of cannabis and declines in the perceived harm of cannabis use.
In 2020, the COVID-19 pandemic resulted in restrictions that reduced the distance people traveled by vehicles.5 In addition, many young adults moved back with their parents, increasing parental oversight and decreasing social opportunities to use alcohol and cannabis.6 Research indicates decreases in young adults’ alcohol use early in the COVID-19 pandemic and mixed results regarding changes in cannabis use.6–9 Despite these changes in driving, living situation, and alcohol use patterns, crash fatalities, including those caused by impaired driving, increased in 2020 and 2021.2
In the current study, we examined driving under the influence of alcohol and cannabis among young adults in Washington State in the 4 years before the pandemic (i.e., 2016–2019) as well as 2020 and 2021. As of November 2023, 24 US states had legalized nonmedical cannabis for adults aged 21 years or older.10 Washington was one of the earliest adopters of legalization and thus constitutes an important context within which to examine patterns of health risk behaviors related to substance use.
We used cross-sectional data from the 2016 to 2021 cohorts of the Washington Young Adult Health Survey (n = 12 585). Participants aged 18 to 25 years were recruited in Washington via direct mailing and online advertising. Participation included completion of a Web-based survey focused on substance use and related risk factors and health behaviors. Full details are provided elsewhere.11
Demographic measures included age, sex assigned at birth, race/ethnicity, employment status (i.e., full time vs not full time), and college status (i.e., enrolled vs not enrolled in a 4-year institution). Participants’ zip codes were used to determine Washington regions (East, Northwest, Southwest) and rural–urban commuting area codes,12 which were recoded as metropolitan (codes 1–3) and nonmetropolitan (codes 4–10). Unweighted and weighted distributions of sample demographics and study variables are provided in Table A (available as a supplement to the online version of this article at https://www.ajph.org).
Participants were asked how many times in the past 30 days they had driven a vehicle (1) after consuming alcohol, (2) within 3 hours after using cannabis, and (3) within 3 hours of using alcohol and cannabis at the same time so that the effects overlapped. Variables were dichotomized to create indicators of any past-month driving under the influence of (1) alcohol (DUI-A), (2) cannabis (DUI-C), and (3) simultaneous use of alcohol and cannabis (DUI-AC).
Participants reported the number of days in the past 30 they drank alcohol, used cannabis, and used alcohol and cannabis at the same time so that their effects overlapped. Participants also reported the number of drinks they consumed on a typical drinking day and how many hours they were high on a typical cannabis use day (1 = “I usually don’t get high,” 10 = “19–24 hours”).
To assess whether there were deviations from prepandemic trends in impaired driving during the COVID-19 pandemic, we specified survey year (2016–2021) as a linear term in our multivariable logistic regression models and included additional indicator variables for 2020 and 2021. Effects for these 2 indicator variables indicated deviations from the prepandemic linear trend observed up to 2019.4 We conducted sensitivity analyses including only data from participants who had used the relevant substance in the past 30 days (e.g., drank alcohol for DUI-A) and controlling for frequency of use of the substance and for typical alcohol quantity or typical hours spent high from cannabis use.4 Exploratory moderations of changes in DUI by gender, age (aged < 21 years vs ≥ 21 years), and college status were examined by adding interaction terms with the linear year term and the 2020 and 2021 indicator variables.
Models controlled for all variables used in weighting, recruitment type (direct mailing vs social media), age, employment status, and college status. Poststratification weights based on birth sex, Washington geographic region, and race/ethnicity were used to improve generalizability to Washington State young adults as a whole.
From 2016 to 2021, 12.0% of the sample reported DUI-A, 12.5% reported DUI-C, and 2.7% reported DUI-AC (Table A). The prevalence for each of the 3 DUI outcomes was stable during the study years, and no significant linear changes were observed (Table 1). Similarly, the indicators for 2020 and 2021 were not statistically significant, suggesting that DUI behavior in those years did not significantly deviate from the linear trend before 2020 (i.e., before the COVID-19 pandemic).
TABLE 1— Annual Changes in the Prevalence of Driving Under the Influence of Alcohol, Cannabis, and Simultaneous Alcohol and Cannabis Use and Possible Deviations During 2020 and 2021: Washington State, 2016–2021
| DUI-A (n = 11 866), OR (95% CI) | DUI-C (n = 11 876), OR (95% CI) | DUI-AC (n = 11 873), OR (95% CI) | |
| Intercept | 0.13 (0.11, 0.14) | 0.12 (0.11, 0.14) | 0.02 (0.02, 0.03) |
| Time | |||
| Linear | 0.94 (0.88, 1.01) | 1.02 (0.95, 1.10) | 0.95 (0.79, 1.13) |
| 2020 indicator | 0.80 (0.60, 1.08) | 0.91 (0.69, 1.19) | 1.09 (0.54, 2.19) |
| 2021 indicator | 1.08 (0.78, 1.50) | 0.91 (0.66, 1.24) | 1.34 (0.62, 2.90) |
Note. CI = confidence interval; DUI-A = driving under the influence of alcohol; DUI-AC = driving under the influence of simultaneous alcohol and cannabis use; DUI-C = driving under the influence of cannabis; OR = odds ratio. All logistic regression models included birth sex, age, race/ethnicity, geographic region, recruitment type, college student status, and employment status as covariates.
Sensitivity analyses among participants who reported substance use were consistent with the initial findings (Table B, available as a supplement to the online version of this article at https://www.ajph.org). Models revealed no significant moderation by gender or age for the linear trends or the 2020 and 2021 indicators (Tables C and D, available as supplements to the online version of this article at https://www.ajph.org). For DUI-A only, the magnitude of the interaction between college status and the linear trend (odds ratio [OR] = 0.86; 95% confidence interval [CI] = 0.72, 1.01) and the interaction with the 2020 indicator were suggestive of differential trends and approached statistical significance (OR = 1.92; 95% CI = 0.99, 3.73; Table E, available as a supplement to the online version of this article at https://www.ajph.org). These estimates may reflect a decrease in DUI-A between 2016 and 2019 for college students but no important change for young adults not in college. In 2020, however, the prevalence of DUI-A among college students increased relative to the prevalence among young adults not in college.
In this study, we found little change in DUI-A, DUI-C, or DUI-AC in Washington State during the first year of the pandemic and after the relaxation of restrictions in 2021, with nonmedical cannabis use being legalized for adults aged 21 years or older. This suggests that many young adults continued to make risky driving decisions, regardless of societal changes during the pandemic. DUI may have been particularly dangerous during the pandemic. Research has indicated that reduced vehicle congestion resulted in increased speeding,5 which may have contributed to increased crash fatalities. Exploratory moderation analyses indicated little difference in DUI trends by gender or age; however, there was an increase (although not statistically significant) in DUI-A during the pandemic among 4-year college students relative to young adults not attending a 4-year college.
Limitations of this study include that the data were self-reported and our inferences rest on the assumption that the same population is represented in each assessment year. Strengths include the use of 6 years of DUI assessments and the use of poststratification weights to increase the generalizability of our findings to Washington State young adults. Future research should consider bidirectional associations between DUI and substance use and whether individuals who engaged in DUI before COVID-19 exhibited fewer reductions in substance use during the pandemic than those who did not.
Young adults, for whom vehicle crashes remain a leading cause of death, showed little change in DUI behaviors during the COVID-19 pandemic. With approximately 1 in 8 young adults reporting DUI-A and DUI-C and almost 3% reporting DUI-AC, there is a continued need for young adult DUI prevention efforts.
ACKNOWLEDGMENTS
This research was funded by National Institute on Drug Abuse (NIDA) grant R01DA057705 (Katarina Guttmannova, principal investigator), and data were collected with support from the Division of Behavioral Health and Recovery, Washington State Health Care Authority (contract 1265–62496; Jason R. Kilmer, principal investigator).
Note. The content is solely the responsibility of the authors and does not necessarily represent the official views of NIDA, the National Institutes of Health, or the Washington State Department of Health.
CONFLICTS OF INTEREST
The authors report no potential or actual conflicts of interest from funding or affiliation-related activities.
HUMAN PARTICIPANT PROTECTION
Approval was granted by the institutional review board of the University of Washington. Informed consent was obtained from all individual participants included in the study.
