Daily or near‐daily cannabis and alcohol use by adults in the United States: A comparison across age groups

Abstract

Background and aims

Patterns of daily or near‐daily (DND) use of alcohol and cannabis among adults in the United States have been changing. The current study measured how these shifts have occurred across developmental periods of adulthood.

Design, Setting, and Participants

U.S. national data from the Monitoring the Future Panel Study include responses from approximately 20 000 adults aged 19 to 65 in 2023. In total, annual data from 1988 to 2023 include 389 649 responses.

Measurements

Self‐report measures of use of cannabis and alcohol DND (i.e., 20 + occasions in the last 30 days) were available for the full age ranges of 19–30 years from 1988 to 2023, of 35–50 years from 2008 to 2023, and of 55–65 years in 2023.

Findings

As of 2023, DND cannabis use was nearly three times as prevalent (10.4%) as DND alcohol use (3.6%) among young adults ages 19 to 30. Early midlife adults have had a convergence but not yet a crossover; there were similar prevalence levels of DND use of cannabis (7.5%) and alcohol (7.8%) among those ages 35 to 50 in 2023. Among late midlife adults ages 55 to 65, DND alcohol use remained more than twice as prevalent (11.4%) as DND cannabis use (5.2%) in 2023.

Conclusions

In the United States, daily or near‐daily (DND) alcohol use remains more prevalent than DND cannabis use among late midlife adults, but the opposite is true for young adults.

INTRODUCTION

In a 2024 article in Addiction published by Caulkins [1], a main finding was that as of 2022 daily cannabis use was more prevalent than daily alcohol use among adults ages 18 and older in the United States (US) National Survey on Drug Use and Health. The purpose of the current data note is to replicate and extend these findings with a particular focus on potential differences across developmental age periods in adulthood. Midlife is commonly considered to be approximately ages 40 to 60 [2], so here we operationalize early midlife as ages 35 to 50 and late midlife as ages 55 to 65. This data note compares the prevalence of daily or near‐daily (DND) use of cannabis with DND use of alcohol and how the pattern of change differs across age. Specifically, we examine (1) the prevalence and trends in DND cannabis use and DND alcohol use among young adults (ages 19–30), early midlife adults (ages 35–50) and late midlife adults (ages 55–65) and (2) whether DND cannabis use is more prevalent than DND alcohol use in 2023 in each of these three stages of adulthood.

METHODS

Data from the Monitoring the Future (MTF) Panel Study are collected from adults ages 19 to 65 [3]. Each year, approximately 20 000 people in the MTF Panel are surveyed across young adulthood (ages 19–30), early midlife (ages 35–50) and late midlife (ages 55–65). Results regarding prevalence levels and trends of a wide variety of substances are released each year in an annual report that also includes details about methods and data collection across the life of the study [3].

Starting in 1976, a new cohort of nationally representative students in 12th grade has completed in‐school surveys across the United States each year [4]. Each year, 2450 of these individuals are selected randomly (with an oversample based on substance use) for participation in the MTF Panel (Patrick et al. [3]). Individuals are randomly assigned to begin their follow‐up surveys at age 19 or at age 20 and they are surveyed every 2 years through age 29/30 and every 5 years after that. In this cohort‐sequential design, data on the full age band of 19 to 30 were first available in 1988, data on the full age band of 35 to 50 were first available in 2008, and the data on the full age band of 55 to 65 were first available in 2023. Analysis weights are used to adjust for sampling and attrition [5]. The total sample sizes for analysis of trends include 266 901 responses from young adults ages 19 to 30 from 1988 to 2023 and 63 076 responses from early midlife adults ages 35 to 50 from 2008 to 2023. Comparisons across age groups in 2023 are based on responses from 4810 for ages 19 to 30, 3328 for ages 35 to 50 and 3011 for ages 55 to 65.

DND use was operationalized as reporting use on 20 or more occasions in the past 30 days (for alcohol and for ‘marijuana’). Prevalence levels and trends in these measures of cannabis use are shown in tables/figures 7–8 in the MTF Panel annual report on 2023 data, and prevalence levels and trends in these measures of alcohol use are shown in tables/figures 19–20 [3]. Significance tests of the linear trends in regression analyses of each use variable are tested over the past 1, 5 and 10 years within each age band (if possible). The analyses were not pre‐registered.

Trends in DND cannabis use

Results show that DND cannabis use has increased over the past 5 and 10 years (from 2018–2023 and from 2013–2023) among young adults and early midlife adults. Specifically, DND cannabis use increased by more than 75% (4.4 percentage points) over 10 years, from a prevalence level of 5.9% in 2013 to 8.6% in 2018 and to 10.4% in 2023 (Figure 1). Among early midlife adults, prevalence increased by approximately 168% (4.8 percentage points) over 10 years, from 2.8% in 2013 to 4.3% in 2018 to 7.5% in 2023 (Figure 2). Trends among late midlife adults are not yet available; however, late midlife adults had a prevalence of 5.2% in 2023 (Figure 3).

FIGURE 1.

Cannabis and alcohol: Trends in 30‐day prevalence of daily/near‐daily use among respondents of modal ages 19 through 30. Note: N = 266 901 for ages 19 to 30 from 1998 to 2023.

FIGURE 2.

Cannabis and alcohol: Trends in 30‐day prevalence of daily/near‐daily use among respondents of modal ages 35 through 50. Note: N = 63 076 for ages 19 to 30 from 1998 to 2023.

FIGURE 3.

Cannabis and alcohol: 30‐day prevalence of daily/near‐daily use among respondents of modal ages 18 through 65 by age group, 2023. *Difference significant at p < 001 level. Note: N = 4810 for ages 19–30, N = 3328 for ages 35–50 and N = 3011 for ages 55–65 in 2023.

Trends in DND alcohol use

DND alcohol use has decreased over the past 5 and 10 years among young adults, but only over the past 5 years among early midlife adults. DND alcohol use has decreased by approximately 35% (1.9 percentage points) among young adults over the past decade, from 5.5% in 2013 to 4.8% in 2018 and to 3.6% in 2023 (Figure 1). Among early midlife adults, there was a significant decrease by approximately 20% (1.6 percentage points) over the past 5 years, from 9.5% in 2018 to 7.6% in 2023 (with a very noticeable 1‐year increase in 2020 during the COVID pandemic) (Figure 2). Late midlife adults had a prevalence of 11.4% in 2023 (Figure 3).

Comparing DND cannabis use and DND alcohol use by age group

Comparing across substances, in 2013, DND cannabis use prevalence was not significantly different from DND alcohol use among young adults ages 19 to 30 (5.5% vs. 5.9%, respectively) (see Figure 1). In 2023, DND cannabis use prevalence was significantly higher than DND alcohol use prevalence among young adults (10.4% vs. 3.6%, respectively) (Figure 3), because of the opposing trends among young adults.

For early midlife adults ages 35 to 50, in 2013, DND cannabis use prevalence was significantly lower (P < 0.01) than DND alcohol use use prevalence (2.8% vs. 4.8%, respectively) (Figure 2). In 2023, DND cannabis use prevalence was not significantly different than DND alcohol use prevalence among early midlife adults (7.5% vs. 7.8%, respectively) (Figure 3).

Among late midlife adults ages 55 to 65, in 2023, DND cannabis use prevalence was significantly lower than DND alcohol use prevalence (5.2% vs. 11.4%, respectively) (Figure 3).

LIMITATIONS AND CONCLUSIONS

Limitations of these data include that they are based on samples of high school students followed longitudinally; adolescents who dropped out of high school are not included. Analysis weights [5] enable weighting back to the original nationally representative samples (of 12th grade students) and mitigate the impact of attrition on analyses. However, it is important to acknowledge that people who engage in more frequent substance use are more likely to become non‐respondents over time [5].

Overall, there has been a crossover of DND use of cannabis and DND use of alcohol among young adults in the past decade. As of 2023, DND cannabis use was more prevalent than DND alcohol use among those ages 19 to 30. Early midlife adults have had a convergence, but not yet a crossover. There were similar prevalence levels of DND use of cannabis and alcohol among those ages 35 to 50 in 2023. Among late midlife adults ages 55 to 65, DND alcohol use remained more prevalent than DND cannabis use by a considerable margin in 2023, with more than double the prevalence of DND alcohol use compared to DND cannabis use. It is important to examine these trends by age to reveal nuanced patterns of change in cannabis and alcohol use in the population. In fact, DND alcohol use remains more prevalent than DND cannabis use among adults ages 55 and older, although if trends continue this may shift in the coming years.

AUTHOR CONTRIBUTIONS

Megan E. Patrick: Conceptualization (lead); formal analysis (equal); funding acquisition (lead); project administration (lead); writing — original draft (lead).

DECLARATION OF INTERESTS

None.

Patrick ME. Daily or near‐daily cannabis and alcohol use by adults in the United States: A comparison across age groups. Addiction. 2025;120(4):779–782. 10.1111/add.16748

DATA AVAILABILITY STATEMENT

MTF Panel data are available through the National Addiction and HIV Data Archive Program https://www.icpsr.umich.edu/web/pages/NAHDAP/index.html. The MTF Panel data dashboard is available https://monitoringthefuture.org/data/panel/.