Abstract
Objective |
Driving under the influence (DUI) of psychoactive substances is an important public health and criminal justice issue, impacting the lives of millions of Americans. Although recent research provides up-to-date information regarding DUI among adults, there is a pressing need for research that focuses specifically on younger/underage drivers. We draw from a large, nationally-representative sample to provide up-to-date evidence as to the prevalence and key criminal justice, substance use, and behavioral health correlates of DUI of cannabis and alcohol among drivers ages 16 to 20 in the United States.
Methods |
We used data from the 2020 and 2021 National Survey on Drug Use and Health (young drivers ages 16-20; N=12,863). All analyses—survey adjusted prevalence estimates, logistic regression—were conducted using Stata SE 17.0 and weighted to account for the study’s stratified cluster sampling design.
Results |
The prevalence of DUI-cannabis for the full sample—including those not endorsing past-year use—was 6.3%. Among youth endorsing past-year cannabis use, 24.5% reported DUI of cannabis. In the full sample and among cannabis users, DUI-cannabis risk was elevated among older and male youth. The prevalence of DUI-alcohol was 2.6% among all youth and 6.1% among youth reporting past-year alcohol consumption.
Conclusions |
Estimates indicate that more than one million young drivers each year are placing their lives and those of others at risk by operating motor vehicles after consuming cannabis and/or alcohol. Findings underscore the importance of prevention efforts targeting underaged cannabis and alcohol-impaired driving.
Keywords: Driving under the influence, cannabis, alcohol, adolescence
Driving under the influence (DUI) of psychoactive substances is an important public health and criminal justice issue, impacting the lives of millions of Americans. Data from the National Highway Traffic Safety Administration (NHTSA) indicate that the number of fatalities related to DUI are in the tens of thousands each year and FBI Uniform Crime Reporting Program data routinely place the number of annual DUI arrests at or above one million (NHTSA, n.d., FBI, 2019). Although we see evidence of DUI for an array of psychoactive substances, national data indicate that annual prevalence is especially pronounced with respect to cannabis (4-5% of drivers) and alcohol (7-8% of drivers). This is consistent with these two substances being the most frequently misused psychoactive substances in the United States (US) (Goings et al., 2023, Substance Abuse and Mental Health Services Administration [SAMHSA], 2020).
A number of recent national studies have examined the prevalence and correlates of DUI in the US, including studies focused specifically on cannabis and alcohol-impaired driving. Azofeifa et al. (2019) estimated that 12 million Americans—close to 5% of the population ages 16 and older—reported past year DUI of cannabis with especially elevated rates among male drivers. Salas-Wright et al. (2021) identified noteworthy differences in rates of DUI of cannabis between non-daily versus daily cannabis users and individuals below or above the diagnostic severity threshold for cannabis use disorder (CUD). Individuals reporting DUI of cannabis also reported greater involvement in other illegal behaviors (e.g., theft, drug selling, other DUI) as well as higher odds of criminal justice system involvement and higher levels of psychological distress. With respect to alcohol, Oh et al. (2020) found that rates of DUI of alcohol have decreased modestly over the past two decades and identified elevated rates of DUI among males, non-Hispanic Whites, and individuals reporting higher household income.
The aforementioned studies provide important up-to-date information as to the prevalence and key correlates of DUI of cannabis and alcohol. However, they share important limitations in terms of their consideration of participant age—namely, either respondents of all ages were combined together in the analyses (Azofeifa et., 2019) or younger drivers ages 16 to 17 were simply omitted altogether (Oh et al., 2020; Salas-Wright et al, 2021). This is noteworthy for several reasons. For one, developmentally-focused substance use research strongly suggests that it is critical we pay careful attention to differences across the early-to-late adolescent and young adult years as rates of use/misuse differ markedly across the life course (Salas-Wright et al., 2016; Vaughn et al., 2016). Moreover, specific to DUI, Goings et al. (2023) identified important differences between youth ages 16-17 and 18-25 as compared to adults ages 26 and older. For example, in contrast to adults, only a fraction of teens reporting DUI did so under the influence of alcohol only; the vast majority reporting DUI of alcohol also reported DUI of cannabis. Additionally, there are manifold reasons why epidemiologic and prevention-focused researchers should focus specifically on young drivers, including the elevated rates of car crashes observed among teens and the need to develop/refine specific DUI prevention programs for this demographic group (Centers for Disease Control and Prevention, 2022).
In the present study, we draw from a large nationally representative sample and aim to provide up-to-date evidence as to the prevalence of DUI of cannabis and alcohol among young drivers ages 16 to 20 in the US. Moreover, we specifically test for differences across key demographic subgroups including age (16-17 versus 18-20), gender, race/ethnicity, socioeconomic status, and urbanicity. Additionally, we look at specific correlates across the domains of criminal justice system involvement, cannabis- and alcohol-related factors, and behavioral health. While a growing body of literature examines DUI among young drivers (see Sterzer et al., 2022 for a scoping review), the present study is important and novel inasmuch as it [1] presents up-to-date, national estimates of DUI in the US, [2] focuses specifically on DUI of cannabis, alcohol, and cannabis and alcohol, and [3] looks systematically at differences between youth ages 16 to 17 versus 18 to 20 as well as across other demographic subgroups. This study will provide new evidence to inform policy decisions and prevention efforts by focusing on DUI among the nation’s youngest drivers who we know face elevated risk of both car crashes and substance misuse.
Method
Sample and Procedures
This study drew from public-use data collected in 2020 and 2021 as part of the National Survey on Drug Use and Health (NSDUH). Using multistage area probability sampling methods, the NSDUH offers population estimates for various substance use and health-related behaviors in the US general population. Each independent NSDUH sample is representative of the non-institutionalized population ages 12 and above, including household residents, civilians residing on military bases, and individuals residing in shelters and group homes. To enhance the accuracy of subgroup estimates, adolescents (ages 12-17) and young adults (ages 18-25) were oversampled. The present study’s analytic sample was limited to youth ages 16-20 (N = 12,863) so as to focus on teenage and underage drivers. The most recent NSDUH data from 2020 and 2021 were pooled to improve estimates and model stability. Further details on the procedures followed by NSDUH can be found in other sources (Substance Abuse and Mental Health Services Administration, 2023).
Measures
DUI.
We examined three DUI outcomes including past-year DUI of cannabis, DUI of alcohol, and DUI of cannabis and alcohol (1 = yes, 0 = no). The DUI of cannabis and alcohol variable was created by combining the first two variables and 1 = having driven under the influence of cannabis and alcohol at the same time or on separate occasions in the past year.
Criminal justice involvement was measured in terms of past-year general arrest (1= one or more times arrested or booked for breaking the law [not counting minor traffic violations] in the past 12 months, 0= no), DUI-related arrest (1= one or more times arrested and booked for DUI of alcohol or drugs, 0= no), drunkenness-related arrest (1= one or more times arrested and booked for drunkenness or other liquor law violations, 0= no), drug-related arrest (1= one or more times arrested and booked for possession, manufacture, or sale of drugs, 0= no), and probation (1= have been on probation or parole/supervised release at any time in the past 12 months, 0=no).
Cannabis and alcohol-related correlates.
Cannabis-related correlates included a continuous variable (cannabis initiation age) and two dichotomous variables: perceiving great risk, versus no-to-moderate risk, of using cannabis 1-2 times a week (“How much do people risk harming themselves physically and in other ways when they smoke marijuana once or twice a week?”) and fairly/very easy, versus probably-impossible-to-fairly difficult, to obtain cannabis (“How difficult or easy would it be for you to get some marijuana, if you wanted some?”). Alcohol-related correlates included a continuous variable (alcohol initiation age) and a dichotomous variable: perceiving great risk, versus no-to-moderate risk, of having 5+ alcoholic drinks 1-2 times per week (“How much do people risk harming themselves physically and in other ways when they have five or more drinks of an alcoholic beverage once or twice a week?”). Additionally, cannabis use disorder and alcohol use disorder were based on the criteria in the DSM-IV (American Psychiatric Association, 2000) and coded such that 1 = met criteria for the disorder in the past year, 0 = did not meet criteria for the corresponding disorder.
Behavioral health correlates.
Self-reported dichotomous behavioral health correlates were examined, including major depressive episodes, illicit drug use (other than cannabis), and poor overall health.
Past-year major depressive episodes (1 = one or more episodes in the past 12 months, 0 = no episode) was based on the criteria in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) (American Psychiatric Association, 2000). Although there are concerns about the potential for under-reporting (Tam et al., 2020), this measure is widely accepted as valid and is commonly utilized in research studies based on NSDUH data (Hai et al., 2022; Salas-Wright et al., 2021).
Past-year illicit drug use (other than cannabis): 1 = used illicit drug except for marijuana, 0 = never used drugs/only used marijuana in the past year.
Poor overall health was based on participants’ self-rated health and coded 1 = fair/poor health and 0 = excellent/very good/good health.
Demographic factors included age (16-17, 18-20 years old), gender (female, male), race/ethnicity (Non-Hispanic [NH] White, NH Black/African American, Hispanic, Other), Family income (living in poverty, up to 2x federal poverty threshold, >2x federal poverty threshold), and urbanicity (large metro, small metro, non-metro).
Statistical Analysis
All analyses were conducted using Stata SE 17.0 and weighted to account for the NSDUH’s stratified cluster sampling design (Substance Abuse and Mental Health Services Administration, 2023). We first calculated the survey adjusted prevalence estimates with 95% confidence intervals (CIs) for DUI of cannabis, DUI of alcohol, and DUI of cannabis and alcohol in the full sample and demographic subgroups, as well as among youth endorsing/not endorsing past-year cannabis use, alcohol use, and cannabis and alcohol use. In addition, using logistic regression and controlling for demographic confounds and survey year, we examined the association between the three DUI variables and a range of correlates, including criminal justice involvement, cannabis- and alcohol-related factors, and behavioral health.
Results
Prevalence and Correlates among the Full Sample, including Abstainers
As displayed in Table 1, the prevalence of DUI cannabis for the full sample—including those not endorsing past-year cannabis or alcohol use—was 6.27% (95% CI: 5.72-6.88) and 2.63% for DUI alcohol (95% CI: 2.22-3.12). A sizeable percentage of the sample also endorsed both behaviors (1.83%, 95% CI: 1.46, 2.30) (see Table 1 for details). As displayed in Table 2, being an older adolescent (18-20 years old), male, White, and living in a small city (compared to living in a large metropolitan area) were associated with a higher odds ratio of DUI cannabis. For DUI alcohol and DUI cannabis/alcohol, only being an older adolescent and having family income more than 2 times over the federal poverty level (compared to being below the federal poverty level) were associated with increased odds, while the “Other” ethnoracial category (Asian, Native American, Native Hawaiian, multiracial) was associated with lower odds relative to White youth.
Table 1.
Past-year prevalence of driving under the influence of cannabis and alcohol among U.S. youth ages 16-20 (includes past-year abstainers and users)
| DUI Cannabis % (95% CI) |
DUI Alcohol % (95% CI) |
DUI Cannabis & Alcohol % (95% CI) |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
|
||||||||||||
| No | Yes | No | Yes | No | Yes | |||||||
| Full Sample | ||||||||||||
|
| ||||||||||||
| Overall | 93.73 | (93.12, 94.28) | 6.27 | (5.72, 6.88) | 97.37 | (96.88, 97.78) | 2.63 | (2.22, 3.12) | 98.17 | (97.70, 98.54) | 1.83 | (1.46, 2.30) |
| By Subgroup | ||||||||||||
|
| ||||||||||||
| Age | ||||||||||||
|
| ||||||||||||
| 16-17 | 95.96 | (95.22, 96.59) | 4.04 | (3.41, 4.78) | 98.37 | (97.78, 98.81) | 1.63 | (1.19, 2.22) | 98.98 | (98.37, 99.36) | 1.03 | (.64, 1.63) |
| 18-20 | 92.27 | (91.28, 93.15) | 7.73 | (6.85, 8.72) | 96.71 | (95.97, 97.31) | 3.29 | (2.69, 4.03) | 97.64 | (96.93, 98.19) | 2.36 | (1.82, 3.07) |
|
| ||||||||||||
| Gender | ||||||||||||
|
| ||||||||||||
| Female | 94.69 | (93.72, 95.51) | 5.31 | (4.49, 6.28) | 97.73 | (97.04, 98.26) | 2.27 | (1.74, 2.96) | 98.58 | (97.95, 99.03) | 1.42 | (0.97, 2.05) |
| Male | 92.81 | (91.78, 93.72) | 7.19 | (6.28, 8.22) | 97.02 | (96.24, 97.65) | 2.98 | (2.35, 3.77) | 97.77 | (97.00, 98.34) | 2.23 | (1.66, 3.00) |
|
| ||||||||||||
| Race/Ethnicity | ||||||||||||
|
| ||||||||||||
| White | 92.27 | (91.10, 93.29) | 7.73 | (6.71, 8.90) | 96.70 | (95.74, 97.46) | 3.30 | (2.54, 4.26) | 97.76 | (96.88, 98.39) | 2.25 | (1.61, 3.12) |
| Black | 95.34 | (93.68, 96.58) | 4.66 | (3.42, 6.32) | 97.66 | (95.95, 98.66) | 2.34 | (1.34, 4.06) | 98.42 | (96.97, 99.18) | 1.58 | (0.82, 3.03) |
| Hispanic | 94.72 | (93.15, 95.95) | 5.28 | (4.05, 6.85) | 97.64 | (96.41, 98.46) | 2.36 | (1.54, 3.59) | 98.31 | (97.26, 98.96) | 1.69 | (1.04, 2.74) |
| Other | 95.73 | (93.83, 97.06) | 4.27 | (2.94, 6.17) | 99.20 | (98.77, 99.48) | 0.80 | (0.52, 1.23) | 99.29 | (98.86, 99.55) | 0.71 | (0.45, 1.14) |
|
| ||||||||||||
| Family income | ||||||||||||
|
| ||||||||||||
| <Poverty level | 94.51 | (93.21, 95.58) | 5.49 | (4.42, 6.79) | 98.07 | (97.43, 98.55) | 1.93 | (1.45, 2.57) | 98.71 | (98.17, 99.10) | 1.29 | (0.90, 1.84) |
| 1- 2x Pov. level | 93.95 | (92.28, 95.28) | 6.05 | (4.72, 7.72) | 97.47 | (96.40, 98.23) | 2.53 | (1.77, 3.60) | 98.34 | (97.17, 99.03) | 1.66 | (0.97, 2.83) |
| > 2x Pov. level | 93.31 | (92.41, 94.10) | 6.69 | (5.90, 7.59) | 96.97 | (96.20, 97.60) | 3.03 | (2.40, 3.80) | 97.83 | (97.11, 98.37) | 2.17 | (1.62, 2.89) |
|
| ||||||||||||
| Urbanicity | ||||||||||||
|
| ||||||||||||
| Large metro | 94.34 | (93.56, 95.03) | 5.66 | (4.97, 6.44) | 97.72 | (97.00, 98.27) | 2.28 | (1.73, 3.00) | 98.37 | (97.69, 98.85) | 1.64 | (1.16, 2.31) |
| Small metro | 92.45 | (90.80, 93.83) | 7.55 | (6.17, 9.21) | 96.94 | (95.89, 97.73) | 3.06 | (2.27, 4.11) | 97.97 | (97.06, 98.61) | 2.03 | (1.39, 2.94) |
| Non-metro | 94.17 | (92.28, 95.61) | 5.83 | (4.39, 7.72) | 96.96 | (95.41, 98.00) | 3.04 | (2.00, 4.60) | 97.83 | (96.37, 98.71) | 2.17 | (1.29, 3.63) |
Note. Data from 2020-2021 NSDUH were pooled. DUI=driving under the influence.
Table 2.
Demographic correlates of driving under the influence of cannabis and alcohol among U.S. youth ages 16-20
| DUI Cannabis | DUI Alcohol | DUI Cannabis & Alcohol | ||||
|---|---|---|---|---|---|---|
| aOR | 95% CI | aOR | 95% CI | aOR | 95% CI | |
| All Youth | ||||||
|
| ||||||
| Age | ||||||
|
| ||||||
| 16-17 | 1.00 | - | 1.00 | - | 1.00 | - |
| 18-20 | 2.08 | 1.62-2.66 | 2.23 | 1.51-3.28 | 2.54 | 1.45-4.43 |
|
| ||||||
| Gender | ||||||
|
| ||||||
| Female | 1.00 | - | 1.00 | - | 1.00 | - |
| Male | 1.34 | 1.05-1.73 | 1.27 | .88-1.83 | 1.53 | .93-2.51 |
|
| ||||||
| Race/Ethnicity | ||||||
|
| ||||||
| White | 1.00 | - | 1.00 | - | 1.00 | - |
| Black/African American | 0.56 | 0.36-0.88 | 0.78 | 0.38-1.58 | 0.78 | 0.34-1.79 |
| Hispanic | 0.67 | 0.47-0.95 | 0.77 | 0.46-1.29 | 0.83 | 0.44-1.56 |
| Other | 0.53 | 0.32-0.85 | 0.24 | 0.13-0.44 | 0.32 | 0.16-0.62 |
|
| ||||||
| Family income | ||||||
|
| ||||||
| <Poverty level | 1.00 | - | 1.00 | - | 1.00 | - |
| 1- 2x Poverty level | 1.16 | 0.80-1.69 | 1.36 | 0.85-2.19 | 1.34 | 0.70-2.58 |
| > 2x Poverty level | 1.26 | 0.95-1.66 | 1.72 | 1.18-2.51 | 1.87 | 1.19-2.95 |
|
| ||||||
| Urbanicity | ||||||
|
| ||||||
| Large metro | 1.00 | - | 1.00 | - | 1.00 | - |
| Small metro | 1.32 | 1.00-1.74 | 1.32 | 0.87-2.02 | 1.25 | 0.74-2.09 |
| Non-metro | 0.93 | 0.63-1.37 | 1.24 | 0.73-2.09 | 1.29 | 0.65-2.56 |
| Youth Endorsing Use | ||||||
|
| ||||||
| Age | ||||||
|
| ||||||
| 16-17 | 1.00 | - | 1.00 | - | 1.00 | - |
| 18-20 | 1.39 | 1.05-1.84 | 1.38 | 0.92-2.08 | 1.46 | 0.82-2.58 |
|
| ||||||
| Gender | ||||||
|
| ||||||
| Female | 1.00 | - | 1.00 | - | 1.00 | - |
| Male | 1.59 | 1.24-2.03 | 1.43 | 0.97-2.10 | 1.76 | 1.08-2.87 |
|
| ||||||
| Race/Ethnicity | ||||||
|
| ||||||
| White | 1.00 | - | 1.00 | - | 1.00 | - |
| Black/African American | 0.62 | 0.39-0.98 | 1.29 | 0.62-2.70 | 1.15 | 0.49-2.69 |
| Hispanic | 0.86 | 0.60-1.22 | 0.88 | 0.54-1.45 | 1.08 | 0.56-2.09 |
| Other | 0.70 | 0.42-1.16 | 0.37 | 0.20-0.69 | 0.48 | 0.25-0.92 |
|
| ||||||
| Family income | ||||||
|
| ||||||
| <Poverty level | 1.00 | - | 1.00 | - | 1.00 | - |
| 1- 2x Poverty level | 1.28 | 0.85-1.92 | 1.36 | 0.86-2.14 | 1.43 | 0.74-2.76 |
| > 2x Poverty level | 1.56 | 1.12-2.17 | 1.63 | 1.11-2.39 | 2.07 | 1.32-3.24 |
|
| ||||||
| Urbanicity | ||||||
|
| ||||||
| Large metro | 1.00 | - | 1.00 | - | 1.00 | - |
| Small metro | 1.55 | 1.13-2.12 | 1.41 | 0.92-2.15 | 1.42 | 0.81-2.48 |
| Non-metro | 1.15 | 0.73-1.82 | 1.31 | 0.77-2.22 | 1.59 | 0.77-3.27 |
Note. Data from 2020-2021 NSDUH were pooled. DUI=driving under the influence; Bold AOR and 95% CI indicates significant associations at p<.050. Survey year, age, gender, race/ethnicity, family income, and urbanicity were adjusted for in all models.
Prevalence and Correlates among the Youth Endorsing Past-Year Use
Table 3 displays the prevalence estimates of DUI among young drivers endorsing cannabis, alcohol, and cannabis plus alcohol use in the past year, respectively. Among youth endorsing past year cannabis use, nearly one in four (24.46%, 95% CI: 22.2, 26.9) reported DUI of cannabis. Among cannabis users, those who were older and male were more likely to report DUI of cannabis and, compared to White youth, youth of all other ethnoracial backgrounds were less likely to report DUI. Among youth reporting alcohol use, 6.10% (95% CI: 5.2, 7.2) reported DUI of alcohol. Youth residing in households with family incomes two times higher than the poverty level endorsed significantly elevated levels of DUI of alcohol. Among youth endorsing both cannabis and alcohol use, 8.41% (95% CI: 2.7, 10.4) of these youth reported DUI of cannabis and alcohol in the previous 12 months. Male youth and youth residing in households earning two times more than the poverty level reported elevated levels of DUI of cannabis and alcohol.
Table 3.
Past-year prevalence of driving under the influence of cannabis and alcohol among U.S. youth ages 16-20 (includes past-year users only)
| DUI Cannabis % (95% CI) |
DUI Alcohol % (95% CI) |
DUI Cannabis & Alcohol % (95% CI) |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
|
||||||||||||
| No | Yes | No | Yes | No | Yes | |||||||
| Full Sample | ||||||||||||
|
| ||||||||||||
| Overall | 75.54 | (71.1-77.8) | 24.46 | (22.2, 26.9) | 93.90 | (92.8, 94.8) | 6.10 | (5.2, 7.2) | 91.59 | (89.5, 93.3) | 8.41 | (6.7, 10.4) |
| By Subgroup | ||||||||||||
|
| ||||||||||||
| Age | ||||||||||||
|
| ||||||||||||
| 16-17 | 78.80 | (75.1, 82.1) | 21.20 | (17.9, 24.9) | 94.95 | (93.2, 96.3) | 5.05 | (3.7, 6.8) | 93.10 | (89.3, 95.6) | 6.90 | (4.4, 10.6) |
| 18-20 | 74.18 | (70.9, 77.2) | 25.82 | (22.8, 29.1) | 93.45 | (92.0, 94.6) | 6.55 | (5.3, 7.9) | 91.04 | (88.4, 93.1) | 8.96 | (6.9, 11.5) |
|
| ||||||||||||
| Gender | ||||||||||||
|
| ||||||||||||
| Female | 79.82 | (76.5, 82.8) | 20.18 | (17.2, 23.5) | 88.74 | (86.6, 90.6) | 11.26 | (9.4, 13.4) | 93.72 | (91.0, 95.6) | 6.28 | (4.3, 9.0) |
| Male | 71.24 | (67.8, 74.5) | 28.76 | (22.5, 32.2) | 84.19 | (81.9, 86.3) | 15.81 | (13.7, 18.1) | 89.43 | (86.2, 91.9) | 10.57 | (8.0, 13.8) |
|
| ||||||||||||
| Race/Ethnicity | ||||||||||||
|
| ||||||||||||
| White | 72.78 | (68.8, 76.4) | 27.22 | (23.6, 31.1) | 93.37 | (91.5, 94.9) | 6.63 | (5.1, 8.5) | 91.12 | (87.8, 93.6) | 8.88 | (6.4, 12.2) |
| Black | 82.48 | (76.9, 86.9) | 17.52 | (13.0, 23.1) | 92.58 | (87.2, 95.8) | 7.42 | (4.2, 12.8) | 91.62 | (84.7, 95.6) | 8.38 | (4.4, 15.3) |
| Hispanic | 75.85 | (69.4, 81.3) | 24.15 | (18.7, 30.6) | 94.40 | (91.6, 96.3) | 5.60 | (3.7, 8.4) | 91.10 | (85.8, 94.6) | 8.90 | (5.4, 14.2) |
| Other | 79.91 | (72.2, 85.9) | 20.09 | (14.1, 27.7) | 97.49 | (96.2, 98.4) | 2.51 | (1.6, 3.8) | 95.77 | (93.4, 97.3) | 4.23 | (2.7, 6.6) |
|
| ||||||||||||
| Family income | ||||||||||||
|
| ||||||||||||
| <Poverty level | 79.57 | (74.9, 83.6) | 20.43 | (16.4, 25.1) | 95.22 | (93.8, 96.3) | 4.77 | (3.6, 6.2) | 94.21 | (91.9, 95.9) | 5.78 | (4.1, 8.0) |
| 1- 2x Pov. level | 76.38 | (71.1, 80.9) | 23.62 | (19.1, 28.9) | 93.90 | (91.5, 95.7) | 6.09 | (4.3, 8.5) | 92.19 | (87.1, 95.4) | 7.81 | (4.6, 12.9) |
| > 2x Pov. level | 73.04 | (69.5, 76.3) | 26.96 | (23.7, 30.5) | 93.23 | (91.6, 94.6) | 6.77 | (5.4, 8.4) | 89.93 | (86.8, 92.4) | 10.07 | (7.6, 13.2) |
|
| ||||||||||||
| Urbanicity | ||||||||||||
|
| ||||||||||||
| Large metro | 78.05 | (74.7, 81.1) | 21.95 | (18.9, 25.3) | 94.62 | (92.9, 95.9) | 5.38 | (4.1, 7.0) | 92.50 | (89.3, 94.8) | 7.50 | (5.2, 10.7) |
| Small metro | 70.85 | (65.9, 75.3) | 29.15 | (24.7, 34.0) | 92.93 | (90.7, 94.6) | 7.07 | (5.3, 9.3) | 90.77 | (87.2, 93.4) | 9.23 | (6.6, 12.8) |
| Non-metro | 76.26 | (69.3, 82.0) | 23.74 | (17.9, 30.6) | 93.32 | (89.9, 95.6) | 6.68 | (4.4, 10.1) | 89.93 | (83.9, 93.9) | 10.07 | (6.1, 16.1) |
Note. Data from 2020-2021 NSDUH were pooled. DUI=driving under the influence
Criminal Justice, Substance-Specific and Behavioral Health Correlates
The criminal justice factors (arrest, DUI-related arrest, drunkenness-related arrest, drug-related arrest, and probation) were all significantly associated with higher odds of all three outcomes. DUI-related arrest had the strongest association with all three (aOR = 54.48 for DUI-cannabis, aOR = 51.52 for DUI-alcohol, and aOR = 71.61 for DUI-cannabis and alcohol).
For cannabis-related correlates, perceiving a great risk with using cannabis 1-2 times a week, and age of cannabis initiation were negatively associated with all three outcomes. Finding it easy to obtain cannabis was strongly associated with higher risk of all three outcomes. For alcohol-related correlates, perceiving great risk with having 5+ alcoholic drinks 1-2 times per week, and age of initiation of alcohol use were negatively associated with all three outcomes (see Table 4 for full results). Cannabis and alcohol use disorder were associated with much greater odds of driving under the influence of each respective substance. This is consistent with our finding that youth who reported DUI of alcohol and cannabis reported more days of use of the respective substance per month (see Figure 1).
Table 4.
Social/behavioral health correlates of driving under the influence of cannabis and alcohol among U.S. youth ages 16-20
| DUI Cannabis | DUI Alcohol | DUI Cannabis & Alcohol | ||||
|---|---|---|---|---|---|---|
| aOR | 95% CI | aOR | 95% CI | aOR | 95% CI | |
| Criminal Justice System | ||||||
|
| ||||||
| Arrest, any | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 9.71 | 5.89-16.02 | 7.68 | 3.76-15.66 | 10.98 | 5.03-23.96 |
| DUI-related arrest | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 54.48 | 11.00-269.82 | 51.52 | 12.30-215.82 | 71.61 | 15.40-333.04 |
| Drunkenness-related arrest | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 15.13 | 2.53-90.37 | 11.45 | 2.63-49.79 | 15.45 | 3.57-66.92 |
| Drug-related arrest | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 17.03 | 5.34-54.32 | 11.29 | 3.52-36.24 | 15.52 | 4.71-51.09 |
| Probation | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 5.15 | 2.89-9.20 | 11.69 | 5.01-27.23 | 9.18 | 3.73-22.55 |
|
| ||||||
| Cannabis-Related | ||||||
|
| ||||||
| Cannabis use disorder | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 43.05 | 31.79-58.31 | 11.23 | 7.88-16.00 | 34.87 | 20.10-60.52 |
| Great risk, using cannabis weekly | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 0.05 | 0.02-0.12 | 0.14 | 0.07-0.28 | 0.04 | 0.01-0.27 |
| Easy to obtain cannabis | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 22.27 | 14.11-35.15 | 37.05 | 11.57-118.59 | 29.84 | 7.17-124.28 |
| Cannabis initiation age | .86 | .80-.93 | .85 | .79-.91 | .85 | .79-.92 |
|
| ||||||
| Alcohol-Related | ||||||
|
| ||||||
| Alcohol use disorder | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 8.31 | 6.27-11.01 | 23.00 | 15.26-34.68 | 25.19 | 15.31-41.47 |
| Great risk, weekly binge drinking | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 0.39 | 0.30-0.51 | 0.45 | 0.27-0.74 | 0.27 | 0.16-0.47 |
| Alcohol initiation age | .84 | .79-.89 | .83 | .78-.89 | .81 | .74-.88 |
|
| ||||||
| Health / Behavioral Health | ||||||
|
| ||||||
| Major depressive episodes | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 1.91 | 1.22-3.01 | 1.21 | 0.59-2.46 | 2.19 | 0.89-5.37 |
| Illicit drug use (besides cannabis) | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 12.45 | 9.26-16.75 | 9.60 | 6.79-13.59 | 14.89 | 9.64-22.98 |
| Poor overall health | ||||||
| No | 1.00 | - | 1.00 | - | 1.00 | - |
| Yes | 1.46 | 0.89-2.39 | 1.37 | 0.68-2.79 | 1.79 | 0.83-3.86 |
Note. Data from 2020-2021 NSDUH were pooled. DUI=driving under the influence; Bold AOR and 95% CI indicates significant associations at p<.050. Survey year, age, gender, race/ethnicity, family income, and urbanicity were adjusted for in all models.
Figure 1.
Distribution of days of cannabis and alcohol use per month among young drivers who (A, C) did not report DUI in the past year or (B, D) reported DUI in the past year. DUI, driving under the influence.
For behavioral health and health correlates, illicit drug use (other than cannabis), was strongly associated with all three outcomes. Past-year major depressive episode was positively associated with DUI-cannabis (aOR: 1.91) but not DUI-alcohol or DUI-cannabis and alcohol. Self-reported poor overall health was not significantly associated with any of the three outcomes.
Discussion
Drawing from a large national sample, findings from the present study shed new light on our understanding of the prevalence and correlates of DUI among young drivers in the US. Below we highlight several key findings.
The first key finding is simply to highlight that a disconcerting proportion of young drivers ages 16-20 in the US are operating motor vehicles under the influence of both cannabis (annual prevalence: 6.3%) and alcohol (annual prevalence: 2.6%). These estimates translate to more than a million young people each year who are taking part in illegal behaviors that risk putting their lives and the lives of others in serious danger. Within this first key finding, we can also highlight three important sub-findings. First, we should note that the prevalence of DUI of cannabis is more than twice that of DUI of alcohol, which points to the importance of prevention efforts targeting this behavior, particularly amid a liberalizing cannabis use landscape in the US (Pacula et al., 2022; Salas-Wright & Vaughn, 2016). This finding also raises questions as to why this is the case—one possibility is that youth perceive DUI of cannabis to be safer than DUI of alcohol (i.e., less likely to result in a crash) and another possibility is that youth may believe they are less likely to be caught for DUI of cannabis given that law enforcement cannot as easily identify cannabis intoxication (see Donnan et al., 2022 for youth perspectives; Goodman et al., 2019). Further research into this topic, including qualitative research, is warranted. Second, we should not overlook the rather remarkable finding that many new drivers—ages 16 to 17—report DUI of cannabis (4%) and alcohol (1.6%) at a time when they are at the highest risk of crashes, even under the best of circumstances. Third, we see that, among youth endorsing underage cannabis and alcohol use, DUI is remarkably common—particularly when it comes to cannabis. Indeed, nearly one in every four young cannabis users (24.5%) reported DUI of cannabis in the past year while just over one in twenty underage drinkers (6.1%) report DUI of alcohol.
A second key finding is that a sizable percentage of the sample, but certainly not all those reporting DUI, endorsed both DUI of cannabis and alcohol. This does seem to suggest that prevention efforts should not be limited to a single substance but should address both cannabis and alcohol. This is of particular importance in light of research indicating that (a) cannabis use rates are particularly disconcerting—and increasing—among binge alcohol users (Salas-Wright et al., 2021) and (b) that simultaneous alcohol and marijuana use rates are increasing among young adults (Hai et al., 2022). Prevention scientists may also consider targeting/including other psychoactive drugs like opioids, stimulants, and hallucinogens, which we know are widely used (Cano et al., 2020; Killion et al., 2021; Park et al., 2022) and are related to DUI, albeit at substantially lower rates (Goings et al., 2022, 2023; Salas-Wright et al., 2021).
The third key finding is that racial/ethnic differences were observed for DUI of cannabis and alcohol in the general population of youth. Specifically, we see that non-Hispanic White youth reported a much higher prevalence of DUI of cannabis (7.7%) compared to the other groups examined (e.g., Black/African American: 4.7%; Hispanic: 5.3%). While this pattern is consistent with recent research examining DUI of cannabis among adults ages 18 and older (Salas-Wright et al., 2021), it is not immediately apparent as to what may be driving this phenomenon. Epidemiologic surveillance data indicates that rates of past-year cannabis use are only slightly higher among non-Hispanic White adolescents (11%) than among Black (10%) and Hispanic (9%) adolescents ages 12 to 17, such that it is likely that differences in rates of use are not having a major impact in the prevalence of DUI. Although the NSDUH data do not allow us to examine the following hypothesis, it is possible that non-Hispanic White youth may be more likely to reside in geographic areas where driving—or driving longer distances—is more common, and that this contextual factor may have influenced the results. Interestingly, we see a different pattern in terms of DUI of alcohol in which no differences are observed between non-Hispanic White, Black, and Hispanic youth, but we did see a significant difference between Non-Hispanic White (3.3%) and “other” (0.8%) young drivers. This finding coheres with recent research pointing to a convergence of binge drinking prevalence among White, Black, and Hispanic youth (Goings et al., 2019).
A fourth key finding is that there is a strong relationship between both DUI of cannabis and DUI of alcohol and criminal justice system involvement. This was observed for an array of outcomes, including any arrest, DUI-related arrest, drunkenness-related arrest, drug-related arrest, and probation status. While this is not a surprising finding per se (indeed the cooccurrence of substance use, lawbreaking, and arrest among adolescents and adults is well-established [see DeLisi et al., 2015; Vaughn et al., 2014, 2020]), it is noteworthy that the magnitude of this association—particularly for cannabis—is far more robust among adolescents than what has been previously observed among adults (Salas-Wright et al., 2021). This result is in keeping with the notion that cannabis use (and certainly cannabis-impaired driving) represents a more severe violation of social norms and law during adolescence than during adulthood. It is plausible that teenagers who take part in such behaviors are, on average, more aberrant or deviant than their adult, especially young adult, counterparts and are therefore more likely to be arrested. It may also be the case that DUI among this population is more likely to be detected—due to less experienced drivers or less familiarity/experience with substance use—thereby attracting the attention of law enforcement officers. Additionally, there is evidence that younger drivers are more likely to be stopped and to receive a ticket versus comparatively older drivers (Anderson, 2021; Factor, 2018).
Study Limitations
Study findings should be interpreted in light of several limitations. First, all data were based on respondent self-report. As such, some respondents may have under or over reported their behavior related to DUI. Second, the NSDUH data did not allow us to measure the number of times an individual took part in cannabis or alcohol impaired driving nor do they allow us to account for the severity of impairment. Third, all data are cross-sectional and therefore we cannot interpret these relationships causally. Fourth, these data were collected during some of the most intense moments in the COVID-19 pandemic. As such, some caution is warranted as to the degree to which these findings may be uniquely shaped by their historical context. We can note that the prevalence of DUI of cannabis was roughly equivalent to that observed among young adults in 2016-2018 (Salas-Wright et al., 2021) and substantially lower than that of young adults in 2016-2017 (Salas-Wright et al., 2020). Finally, the NSDUH does not allow us to look at jurisdiction level factors that may influence cannabis or alcohol use risk or risk of young drivers DUI (e.g., access to public transportation or ride shares like Uber/Lift, or differences in access to cannabis or alcohol dispensaries/stores). On a related point, the NSDUH data cannot be geocoded so as to allow for an analysis of how distances between locations may relate to the likelihood of DUI
Conclusion
The present study provides new evidence as to the prevalence of DUI of cannabis and alcohol among young drivers ages 16 to 20 in the US. Our findings indicate that more than one million young drivers each year are placing their lives and those of others at risk by operating a motor vehicle after consuming powerful psychoactive drugs. Notably, we observed that DUI was more likely under the influence of cannabis than alcohol; however, a sizable proportion of those reporting DUI endorsed this behavior both for cannabis and for alcohol, suggesting the importance of prevention efforts the target more than one psychoactive substance. Racial/ethnic differences were observed in which White youth stood out as more likely than Black and Hispanic youth to endorse DUI of cannabis, but no differences were observed between these groups in terms of DUI of alcohol. We also observed a powerful association between DUI of cannabis and alcohol and the likelihood of criminal justice system involvement. In all, findings make clear that DUI is by no means limited to adults and underscore the importance of prevention efforts targeting teen and underaged cannabis and alcohol misuse and impaired driving.
Highlights.
Over one million youth drive annually under the influence of cannabis/alcohol.
More youth drive under the influence (DUI) of cannabis (6.3%) than alcohol (2.6%).
Among past-year cannabis users, 24.5% report past-year DUI of cannabis.
Among past-year alcohol users, only 6.1% report past year DUI of alcohol.
Driving under the influence is highly associated with criminal justice involvement.
Acknowledgments
This work was supported by the National Institute on Alcohol Abuse and Alcoholism [NIAAA] of the National Institutes of Health [Award Number K01AA026645]. The content is solely the responsibility of the authors and does not necessarily represent the official views of NIAAA or the NIH.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
CRediT
- Salas-Wright lead the conceptualization and writing of the manuscript.
- Hai implemented the statistical analyses and lead the writing of the method section and the construction of tables. Hai contributed to the selection of variables as well.
- Vaughn contributed to study conceptualization, variable selection, and the overall writing of the manuscript.
- Hodges lead the writing of the method section and assisted in statistical analysis. Hodges also contributed to the refinement of the writing.
- Goings worked with Salas-Wright to refine the study conceptualization, study implications, and overall writing of the manuscript.
Author Agreement
The authors understand that, upon acceptance, we will need to sign an author agreement giving Addictive Behaviors the right to publish and distribute the manuscript in part or in its entirety.
Conflict of Interest
The authors have no conflicts to disclose.
Declaration of Interest
None.
References
- American Psychiatric Association, 2000. Diagnostic and statistical manual of mental disorders, 4th ed. Te. ed. Amer Psychiatric Pub Inc, Washington, DC. [Google Scholar]
- Azofeifa A, Rexach-Guzmán BD, Hagemeyer AN, Rudd RA, Sauber-Schatz EK. Driving Under the Influence of Marijuana and Illicit Drugs Among Persons Aged ≥16 Years — United States, 2018. MMWR Morb Mortal Wkly Rep 2019;68:1153–1157. DOI: 10.15585/mmwr.mm6850a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cano M, Oh S, Salas-Wright CP, & Vaughn MG (2020). Cocaine use and overdose mortality in the United States: evidence from two national data sources, 2002–2018. Drug and alcohol dependence, 214, 108148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Centers for Disease Control and Prevention (2022). Teen driver and passenger safety. Retrieved on May 10, 2023 from: https://www.cdc.gov/transportationsafety/teen_drivers/index.html
- DeLisi M, Vaughn MG, Salas-Wright CP, & Jennings WG (2015). Drugged and dangerous: Prevalence and variants of substance use comorbidity among seriously violent offenders in the United States. Journal of Drug Issues, 45(3), 232–248. [Google Scholar]
- Donnan JR, Drakes DH, Rowe EC, Najafizada M, & Bishop LD (2022). Driving under the influence of cannabis: perceptions from Canadian youth. BMC public health, 22(1), 2384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Factor R. (2018). An empirical analysis of the characteristics of drivers who are ticketed for traffic offences. Transportation research part F: traffic psychology and behaviour, 53, 1–13. [Google Scholar]
- Federal Bureau of Investigation (2019). Uniform Crime Reporting: Crim in the United States. Retrieved on May 10, 2023 from https://ucr.fbi.gov/crime-in-the-u.s/2019/crime-in-the-u.s.-2019/topic-pages/persons-arrested
- Goings TC, Cano M, Salas-Wright CP, Campos BM, & Vaughn MG (2022). Prevalence and correlates of driving under the influence of stimulants: evidence from a national sample. Addictive behaviors, 132, 107364. [DOI] [PubMed] [Google Scholar]
- Goings TC, Salas-Wright CP, Belgrave FZ, Nelson EJ, Harezlak J, & Vaughn MG (2019). Trends in binge drinking and alcohol abstention among adolescents in the US, 2002-2016. Drug and alcohol dependence, 200, 115–123. [DOI] [PubMed] [Google Scholar]
- Goings TC, Salas-Wright C, & Vaughn M (2023). Toward a typology of driving under the influence of alcohol and drugs. Social psychiatry and psychiatric epidemiology, 58(2), 227–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Erin Goodman S, Leos-Toro C, & Hammond D (2020). Risk perceptions of cannabis-vs. alcohol-impaired driving among Canadian young people. Drugs: Education, Prevention and Policy, 27(3), 205–212. [Google Scholar]
- Hai AH, Carey KB, Vaughn MG, Lee CS, Franklin C, & Salas-Wright CP (2022). Simultaneous alcohol and marijuana use among college students in the United States, 2006–2019. Addictive behaviors reports, 16, 100452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hai AH, Oh S, Lee CS, Kelly JF, Vaughn MG, Salas-Wright CP, 2022. Mutual-help group participation for substance use problems in the US: Correlates and trends from 2002 to 2018. Addictive Behaviors 128. 10.1016/J.ADDBEH.2021.107232 [DOI] [PubMed] [Google Scholar]
- Killion B, Hai AH, Alsolami A, Vaughn MG, Oh PS, & Salas-Wright CP (2021). LSD use in the United States: Trends, correlates, and a typology of us. Drug and alcohol dependence, 223, 108715. [DOI] [PubMed] [Google Scholar]
- National Highway Traffic Safety Administration (no date). Drunk driving. Retrieved on May 10, 2023 from https://www.nhtsa.gov/risky-driving/drunk-drivingonMay10
- Oh S, Vaughn MG, Salas-Wright CP, AbiNader MA, & Sanchez M (2020). Driving under the influence of alcohol: findings from the NSDUH, 2002–2017. Addictive behaviors, 108, 106439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pacula RL, Smart R, Lira MC, Pessar SC, Blanchette JG, & Naimi TS (2022). Relationships of cannabis policy liberalization with alcohol use and co-use with cannabis: A narrative review. Alcohol research: current reviews, 42(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- Park D, Oh S, Cano M, Salas-Wright CP, & Vaughn MG (2022). Trends and distinct profiles of persons who inject drugs in the United States, 2015–2019. Preventive Medicine, 164, 107289. [DOI] [PubMed] [Google Scholar]
- Salas-Wright CP, Cano M, Hai AH, Oh S, & Vaughn MG (2021). Prevalence and Correlates of Driving Under the Influence of Cannabis in the US. American journal of preventive medicine, 60(6), e251–e260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Salas-Wright CP, Cano M, Hodges J, Oh S, Hai AH, & Vaughn MG (2021). Driving while under the influence of hallucinogens: Prevalence, correlates, and risk profiles. Drug and alcohol dependence, 228, 109055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Salas-Wright CP, Hai AH, Oh S, Alsolami A, & Vaughn MG (2021). Trends in cannabis views and use among American adults: intersections with alcohol consumption, 2002–2018. Addictive behaviors, 116, 106818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Salas-Wright CP, Hodges JC, Hai AH, Alsolami A, Vaughn MG, 2021. Toward a typology of hallucinogen users in the United States. Drug Alcohol Depend 229, 109139. 10.1016/J.DRUGALCDEP.2021.109139 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Salas-Wright CP, & Vaughn MG (2016). The changing landscape of adolescent marijuana use risk. Journal of Adolescent Health, 59(3), 246–247. [DOI] [PubMed] [Google Scholar]
- Salas-Wright CP, Vaughn MG, & González JMR (201). Drug abuse and antisocial behavior: A biosocial life course approach. Springer. [Google Scholar]
- Sterzer FR, Caird JK, Simmons S, & Bourdage JS (2022). A scoping review of predictors of driving under the influence of cannabis (DUIC) in young drivers. Transportation research part F: traffic psychology and behaviour, 88, 168–183. [Google Scholar]
- Substance Abuse and Mental Health Services Administration. (2020). Key substance use and mental health indicators in the United States: Results from the 2019 National Survey on Drug Use and Health (HHS Publication No. PEP20-07-01-001, NSDUH Series H-55). Rockville, MD: Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration. Retrieved from https://www.samhsa.gov/data/ [Google Scholar]
- Substance Abuse and Mental Health Services Administration, 2023. What is NSDUH? [WWW Document], URL https://nsduhweb.rti.org/respweb/homepage.cfm (accessed 2.21.23).
- Tam J, Mezuk B, Zivin K, Meza R, 2020. U.S. Simulation of Lifetime Major Depressive Episode Prevalence and Recall Error. Am J Prev Med 59, e39–e47. 10.1016/J.AMEPRE.2020.03.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vaughn MG, AbiNader M, Salas-Wright CP, Holzer K, Oh S, & Chang Y (2020). Trends in cannabis use among justice-involved youth in the United States, 2002–2017. The American Journal of Drug and Alcohol Abuse, 46(4), 462–471. [DOI] [PubMed] [Google Scholar]
- Vaughn MG, Salas-Wright CP, Alsolami AS, Oh S, & Goings TC (2021). Margin for error: examining racial and ethnic trends in adolescent risk propensity. Social Psychiatry and Psychiatric Epidemiology, 56, 993–1002. [DOI] [PubMed] [Google Scholar]
- Vaughn MG, Salas-Wright CP, DeLisi M, & Maynard BR (2014). Violence and externalizing behavior among youth in the United States: Is there a severe 5%?. Youth violence and juvenile justice, 12(1), 3–21. [Google Scholar]
- Vaughn MG, Salas-Wright CP, & Reingle-Gonzalez JM (2016). Addiction and crime: The importance of asymmetry in offending and the life-course. Journal of addictive diseases, 35(4), 213–217. [DOI] [PubMed] [Google Scholar]