Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Jul 1.
Published in final edited form as: Alcohol Clin Exp Res. 2017 Jun 1;41(7):1319–1328. doi: 10.1111/acer.13413

Age-Specific Prevalence of Binge and High-Intensity Drinking among US Young Adults: Changes from 2005 to 2015

Megan E Patrick 1, Yvonne M Terry-McElrath 1, Richard A Miech 1, John E Schulenberg 1, Patrick M O'Malley 1, Lloyd D Johnston 1
PMCID: PMC5553703  NIHMSID: NIHMS872824  PMID: 28571107

Abstract

Background

This study examines changes during the past decade, from 2005 to 2015, in binge and high-intensity drinking in seven separate age groups of US 12th graders and young adults.

Methods

National longitudinal data (N = 6,711) from Monitoring the Future were used to examine trends in consuming 5+, 10+, and 15+ drinks on the same occasion in the past two weeks from ages 18 to 29/30 overall and by gender. Results were compared with trends in past 12-month and 30-day alcohol use for the same age groups.

Results

Between 2005 and 2015, binge (5+) and high-intensity drinking (10+, 15+) generally decreased for individuals in their early 20s, remained somewhat stable for individuals in their mid-20s, and increased for individuals at the end of young adulthood (age 29/30). The observed historical trends in binge and high-intensity drinking were similar to those for past 12-month and past 30-day alcohol use for those aged 18 to 20, but diverged for most other age groups in young adulthood. Trends were generally similar for men and women, except that the increase in prevalence began earlier in young adulthood for women than men.

Conclusions

Binge and high-intensity drinking among US 12th graders and young adults are dynamic phenomena. Prevention and intervention efforts aimed at reducing the harms resulting from 5+, 10+, and 15+ drinking should acknowledge and focus on differences in trends in these behaviors by age and gender.

Keywords: binge drinking, high-intensity drinking, historical trend, cohort, young adult

Monitoring and documenting change in the prevalence and volume of alcohol use is important because alcohol-related harm is largely determined by the volume of alcohol consumed and the pattern of drinking (WHO, 2015). In 2010, alcohol was the fifth highest overall risk factor for worldwide mortality and disease burden, and the leading risk factor for individuals between ages 15 and 49 (Lim et al., 2012). More than 88,000 deaths per year in the US were attributable to alcohol use from 2006 to 2010, according to the Alcohol-Related Disease Impact (ARDI) application (CDC, 2013). Individuals aged 20-34 experienced the largest number of alcohol-attributable deaths from acute causes (CDC, 2013). For numerous developmental and cultural reasons (e.g., Schulenberg & Maggs, 2002), young adults tend to drink at high volume, and thus are at particular risk for negative consequences from alcohol use.

Both acute and long-term risks rise exponentially as the amount of alcohol consumed increases (Hingson & White, 2013; Patrick, 2016). In more than half of the US alcohol-attributable deaths from 2006 to 2010, either the decedent or the person responsible for the death had a blood alcohol concentration (BAC) higher than 0.08 g/dL (CDC, 2013), the level used by the National Institute on Alcohol Abuse and alcoholism to define binge drinking (NIAAA, 2016). In regards to economic impact, gender-specific binge drinking (4+ drinks for women and 5+ drinks for men per occasion) was responsible for an estimated 77% of the $249 billion cost of excessive drinking in the US in 2010 (Sacks et al., 2015). The average number of drinks consumed during a binge episode among US young adults aged 18-24 is actually 9.5 (Naimi et al., 2010), about double the 4+/5+ NIAAA-defined gender-specific binge level. There is a need for research focusing on high-intensity drinking, defined as consumption of 10+ or even 15+ drinks on a single occasion (Hingson & White, 2013; Patrick, 2016; Patrick et al., 2013). Documenting change in binge and high-intensity drinking patterns provides the opportunity to estimate future health service and treatment needs (Dawson et al., 2015), as well as evaluate who may be at highest risk for experiencing alcohol-related harm themselves or putting the public at risk from risky drinking behavior.

Alcohol consumption levels are dynamic over time and across age. Both the prevalence and volume of alcohol consumption by adults in the US increased over the first decade of the 21st century, a trend documented in data from the National Alcohol Survey (NAS) (Kerr et al., 2014), the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) (Dawson et al., 2015), the National Health Interview Survey (NCHS, 2013), and the National Study on Drug Use and Health (NSDUH) (Center for Behavioral Health Statistics and Quality, 2015). The World Health Organization (WHO) has projected that per capita alcohol consumption in the Americas overall will increase through at least 2025 (WHO, 2014). Yet, there is evidence that the observed increases have not been consistent across age, with noteworthy differences occurring across young adulthood. For example, increases between 2001/2002 and 2012/2013 in volume (average daily intake), overall drinking frequency, and prevalence of gender-specific binge drinking in NESARC data were observed for individuals aged 25-44, but not for those aged 18-24 (Dawson et al., 2015). Similar age differences were found in NSDUH data examining trends from 2002 through 2014 in having 5+ drinks on the same occasion at least once in the past 30 days: prevalence decreased among individuals aged 18-25, but increased for those aged 26+ (Center for Behavioral Health Statistics and Quality, 2015). Examining historical change in the course of overall binge drinking (5+ drinks in the past two weeks for both men and women), Jager et al. (2015) found that age 18 overall binge drinking decreased for cohorts from 1976 to 2004, but age 26 overall binge drinking for these cohorts (assessed in 1984-2012) remained stable; this resulted in a lower starting point but sharper increase in overall binge drinking across the transition to adulthood for more recent cohorts.

Research has only recently focused on high-intensity drinking and little information is as yet available on trends for 10+ and 15+ drinking. Available studies indicate that 10+ drinking declined significantly from 2005 forward among 18 year-olds (Miech et al., 2016; Patrick et al., 2013) and among 19/20 years-olds (Patrick & Terry-McElrath, 2016), but no trend was observed for individuals aged 25/26 (Terry-McElrath & Patrick, 2016). Prevalence of 15+ drinking did not differ across time for those aged 18 or 19/20 (Patrick et al., 2013; Patrick & Terry-McElrath, 2016).

It is an open question whether trends in high-intensity drinking follow those of overall alcohol consumption. To the degree that high-intensity drinking consistently follows overall prevalence trends, efforts to monitor and document change in alcohol use could focus primarily on overall prevalence and still capture changes in high-intensity use. In contrast, if high-intensity drinking trends differ from overall prevalence trends to some degree, monitoring of both overall prevalence and high-intensity drinking would be indicated. Data from NESARC indicated that recent gender-specific binge drinking trends were distinct from past 12-month prevalence trends for US young adults aged 18-24, but trends in both behaviors were similar for individuals aged 25-44 (Dawson et al., 2015). Trends from 2005-2014 for alcohol use in the past 12-months, past 30-days, 5+, and 10+ drinking were similar for a national sample of young adults aged 25/26 (Terry-McElrath & Patrick, 2016).

Gender is a key demographic risk factor for binge and high-intensity drinking among young adults. Young adult men have consistently reported higher prevalence of these behaviors (Jager et al., 2015; Johnston et al., 2016b; Kerr et al., 2009, 2014; Keyes & Miech, 2013; Patrick et al., 2013, 2016; Patrick & Terry-McElrath, 2016; Terry-McElrath & Patrick, 2016; White et al., 2006). In recent decades, increases in both gender-specific and overall binge drinking have occurred faster for women than men when combining across age (Dawson et al., 2015; Keyes & Miech, 2013), resulting in an attenuation of gender differences (Johnston et al., 2016a ; Kerr et al., 2009; Keyes et al., 2008, 2011). What is not known is the extent to which recent trends in high-intensity drinking at specific ages across young adulthood may or may not vary by gender.

Based on the studies reviewed above, it appears that levels of participation in binge and high-intensity drinking in the US recently have been decreasing for individuals aged 18-20, but patterns after age 20 are not as clearly documented. To the authors' knowledge, no prior studies have explored these age-specific changes in trends of high-intensity drinking across young adulthood, or the extent to which they may or may not vary by gender. Efforts to reduce or prevent binge and high-intensity drinking among the young adult population will be strengthened by a better understanding of the historical change in ages at which young adults may be most at risk for participating in these behaviors.

The current paper aims to contribute to the literature by examining a decade of change (from 2005 to 2015) in the prevalence of past 2-week binge and high-intensity drinking among a general US 12th grade and young adult sample. Three research aims guided analysis: (1) Examine the extent to which age-specific prevalence of binge and high-intensity drinking levels changed from 2005 to 2015 among individuals in 7 age groups from 18 to 30; (2) Examine whether observed trends parallel overall past 12-month and past 30-day alcohol consumption trends among these age groups; and (3) Test to what degree trends are comparable for men and women.

Method

Sample

The study utilizes data from Monitoring the Future (MTF), a national cohort-sequential study. For detailed description of methodology, see Bachman et al. (2015) and Johnston et al. (2016b). Briefly, nationally representative samples of approximately 15,000 12th graders (modal age 18) from about 130 schools in the contiguous 48 states have been surveyed annually since 1976 yielding sequential cohorts. Students complete self-administered surveys, typically during a normal class period. A sub-sample of about 2,400 12th graders is selected from each annual sample for longitudinal follow-up by mail; substance users are oversampled (analyses include weights accounting for sampling procedures). Respondents are randomly divided with half surveyed one year after graduation (modal age 19) and then every two years after that to age 29, and half surveyed two years after graduation (modal age 20) and then every two years following to age 30. The result is six follow-up surveys, one at each of modal ages 19/20, 21/22, 23/24, 25/26, 27/28, and 29/30. Follow-up questionnaires are mailed in the spring with a modest monetary incentive. The University of Michigan Institutional Review Board approved the study.

High-intensity drinking measures were added to the MTF survey in 2005 on only one of the six different questionnaire forms used in the study (randomly distributed). The current sample was limited to cohorts that had the possibility of responding to the relevant form during calendar years 2005-2015 either at base year (12th grade) or at one or more follow-up surveys between the ages of 19/20 and 29/30 (cohorts from 1993-2015). The average response rate for these cohorts at 12th grade was 82.7% (with almost all non-response due to absenteeism). A total of 4,195 individuals responded to the relevant questionnaire form at age 18 from 2005 to 2015, for an age 18 response rate of 82.3%. Response rates for age-specific follow-ups during calendar years 2005-2015 ranged from 50.0% at modal age 19/20 (2,205 individuals) to 43% at modal ages 27/28 and 29/30 (approximately 1,880 individuals at each age). (Supplemental Table 1 provides a detailed description of eligible cohorts and response rates for the relevant form.) Valid data at one or more modal ages on at least one of the alcohol outcomes of interest to the current analysis were available for 6,711 individuals; 45% of cases (unweighted) were men. Adjustments for attrition are discussed in the Analysis section below.

Measures

At base year and each follow-up survey, respondents were asked, “During the last two weeks, how many times have you had …5 or more drinks in a row; …10 or more drinks in a row; …15 or more drinks in a row?” Responses ranged from none to 10 or more times. Three dichotomous outcomes were coded: 5+ drinking (1=5+ drinks in a row at least once in the past two weeks; 0=never); 10+ drinking (1=10+ drinks in a row at least once in the last two weeks; 0=never); 15+ drinking (1=15+ drinks in a row at least once in the last two weeks; 0=never). As supplemental to the main outcomes listed above, additional dichotomous measures were coded indicating 5-9 drinks and 10-14 drinks, and a dichotomous gender-specific binge drinking measure was coded indicating 5+ drinking for men and 4+ drinking for women (gender-specific high-intensity drinking measures are unavailable).

At base year and each follow-up survey, respondents also were asked, “On how many occasions (if any) have you had any alcoholic beverage to drink—more than just a few sips …during the last 12 months; …during the last 30 days?” Response options ranged from 0 occasions to 40 or more. Two dichotomous measures were coded: past 12-month alcohol use (1=any; 0=none) and past 30-day alcohol use (1=any; 0=none). At age 18, respondents were asked, “What is your sex?” (response options included male, female).

Analysis

The proportions of respondents reporting each outcome by calendar year within modal age group were estimated using the SURVEYMEANS procedure in SAS 13.2, which also provided standard error estimates and 95% confidence intervals. Gender-specific prevalence estimates for each outcome by calendar year within modal age group showed that prevalence estimates for 15+ drinking among women for several modal ages were near 0%. Thus, for women, the SURVEYFREQ procedure was used to obtain asymmetric confidence intervals, as well as prevalence and standard error estimates. Statistical comparisons of recent prevalence by gender were obtained using the Rao-Scott chi square in SURVEYFREQ models.

Age 18 analyses were weighted by the sampling weight correcting for oversampling of age 18 substance users. Analyses for all follow-up data were weighted using follow-up specific attrition weights, calculated as the inverse of the probability of responding at each modal age based on covariates measured at age 18 (gender, race/ethnicity, college plans, high school grades, number of parents in the home, religiosity, parental education, alcohol use, cigarette use, marijuana use, region of country, cohort, sampling weight correcting for oversampling of age 18 substance users). Trends in prevalence of each alcohol use behavior within modal age group were estimated using Joinpoint 4.3.1.0 (Kim et al., 2000; NCI, 2016a), wherein linear trend lines are connected together at “joinpoints” at which a significant change in slope occurs (modeled using a Monte Carlo Permutation method). The most parsimonious model is preferred; that is, the model begins with a straight line and tests whether additional joinpoints are statistically significant and must be added to the model. Given that a total of 11 data points were modeled for each age group (one for each year of data), the maximum number of joinpoints was set at 1 (as recommended by NCI, 2016b). As gender-specific estimates were obtained from the same population (thus errors were correlated), gender comparisons were made by computing a new time series (the difference in yearly estimates between men and women), and then fitting a regression model and testing whether the slope coefficient for time was zero (both linear and quadratic time functions were examined). A non-significant slope indicated that the two series were parallel, because the difference time series had a constant mean (NCI, 2016c).

Results

Mean two-week prevalence levels (2005-2015 combined) for 5+, 10+, and 15+ drinking by age group are presented in Table 1, overall and by gender. Similar data for 5-9 drinking, 10-14 drinking, and gender-specific binge drinking are provided in Supplemental Tables 2-3.

Table 1. Mean Prevalence of Binge and High-Intensity Drinking by Age Group among US Young Adults Overall and by Gender, 2005-2015.

Age All Respondents Men Women Men vs. Women
na %b (95% CI) n % (95% CI) n % (95% CI) pc
5+ drinks in the past 2 weeks
18 3,990 17.7 (16.5, 18.9) 1,886 22.4 (20.5, 24.3) 2,104 13.7 (12.3, 15.2) <.0001
19/20 2,132 23.5 (21.5, 25.4) 876 28.1 (24.9, 31.3) 1,254 19.5 (17.2, 21.8) <.0001
21/22 2,067 32.5 (30.0, 34.9) 826 40.0 (35.9, 44.2) 1,239 26.0 (23.2, 28.8) <.0001
23/24 2,038 32.1 (29.7, 34.5) 834 41.0 (36.9, 45.0) 1,204 24.1 (21.5, 26.7) <.0001
25/26 1,879 32.5 (29.9, 35.0) 750 43.9 (39.6, 48.2) 1,129 22.8 (20.0, 25.6) <.0001
27/28 1,799 26.0 (23.6, 28.4) 710 36.8 (32.6, 41.1) 1,089 16.9 (14.4, 19.3) <.0001
29/30 1,806 25.0 (22.4, 27.6) 710 34.5 (29.9, 39.2) 1,096 16.6 (14.1, 19.1) <.0001
10+ drinks in the past 2 weeks
18 3,990 8.7 (7.9, 9.6) 1,886 13.0 (11.5, 14.5) 2,104 5.1 (4.2, 5.9) <.0001
19/20 2,132 10.0 (8.7, 11.4) 876 15.5 (13.0, 18.0) 1,254 5.2 (4.0, 6.4) <.0001
21/22 2,067 13.5 (11.7, 15.4) 826 21.9 (18.4, 25.4) 1,239 6.4 (5.1, 7.8) <.0001
23/24 2,038 12.3 (10.6, 14.1) 834 20.3 (17.0, 23.6) 1,204 5.1 (3.9, 6.4) <.0001
25/26 1,879 11.7 (10.1, 13.3) 750 18.7 (15.7, 21.7) 1,129 5.7 (4.2, 7.3) <.0001
27/28 1,799 9.1 (7.6, 10.7) 710 16.7 (13.7, 19.7) 1,089 2.7 (1.8, 3.7) <.0001
29/30 1,806 8.9 (7.3, 10.5) 710 15.1 (12.0, 18.2) 1,096 3.5 (2.1, 4.8) <.0001
15+ drinks in the past 2 weeks
18 3,990 4.6 (4.0, 5.2) 1,886 7.4 (6.3, 8.6) 2,104 2.2 (1.6, 2.8) <.0001
19/20 2,132 4.2 (3.3, 5.1) 876 7.4 (5.6, 9.3) 1,254 1.3 (0.7, 2.0) <.0001
21/22 2,067 5.2 (4.2, 6.3) 826 9.1 (7.1, 11.2) 1,239 1.9 (1.3, 2.6) <.0001
23/24 2,038 5.5 (4.2, 6.8) 834 10.0 (7.4, 12.6) 1,204 1.4 (0.8, 2.1) <.0001
25/26 1,879 3.4 (2.6, 4.3) 750 6.3 (4.6, 8.0) 1,129 1.0 (0.4, 1.6) <.0001
27/28 1,799 2.9 (2.1, 3.7) 710 5.6 (3.9, 7.3) 1,089 0.7 (0.2, 1.1) <.0001
29/30 1,806 3.1 (1.9, 4.2) 710 5.5 (3.4, 7.7) 1,096 0.9 (0.0, 1.8) 0.0002
Open in a new tab
a

Unweighted n for each specified group.

b

Weighted percentage of each age group reporting the specified drinking behavior.

c

p-values from Rao-Scott chi square tests of differences in prevalence by gender.

For 5+ and 10+ drinking, mean prevalence increased from ages 18 through 21/22, remained relatively stable from ages 21/22 through 25/26 (averaging 32% for 5+ drinking and 13% for 10+ drinking), and then decreased somewhat through age 29/30. Mean prevalence of 15+ drinking remained relatively stable from ages 18 through 23/24 (averaging 5%), and decreased through age 29/30. Men reported significantly higher mean prevalence of 5+, 10+, and 15+ drinking than women at every age.

Mean prevalence of 5-9 drinking and gender-specific binge drinking increased from ages 18 through 25/26, and then decreased through the remainder of young adulthood. Mean prevalence of 10-14 drinking across age was similar to that for 10+ drinking. Men and women had similar mean prevalence of 5-9 drinking and gender-specific binge drinking from ages 18 through 23/24; thereafter, mean prevalence was higher for men. Mean prevalence of 10-14 drinking was higher for men than women at every age.

Young adult binge and high-intensity drinking across time (Aim 1)

Trends for 5+, 10+, and 15+ drinking from 2005 to 2015 by young adult age group are presented graphically in Figure 1; slope estimates are provided in Table 2. Results for 5-9 drinking, 10-14 drinking, and gender-specific binge drinking are provided in Supplemental Figures 1-2 and Supplemental Tables 4-5.

Figure 1. Trends in Prevalence of Binge and High-Intensity Drinking among US Young Adults by Age Group, 2005-2015.

Figure 1

Open in a new tab

Notes: Slope estimates for all trends are provided in Table 2.

Table 2. Slope Estimates for Trends in Prevalence of Binge and High-Intensity Drinking among US Young Adults by Age Group, 2005-2015.

Age Na Slope 1 (SE) p Joinpointb Slope 2 (SE) p
5+ drinks in the past 2 weeks
18 3,990 -0.001 (0.003) 0.772 2012 -0.025 (0.009) 0.031
19/20 2,132 -0.014 (0.003) 0.003
21/22 2,067 -0.004 (0.003) 0.301
23/24 2,038 -0.012 (0.002) <0.001
25/26 1,879 -0.002 (0.004) 0.642
27/28 1,799 0.001 (0.004) 0.734
29/30 1,806 0.010 (0.003) 0.012
10+ drinks in the past 2 weeks
18 3,990 -0.002 (0.001) 0.106
19/20 2,132 -0.007 (0.001) 0.001
21/22 2,067 -0.004 (0.002) 0.132
23/24 2,038 -0.002 (0.002) 0.443
25/26 1,879 -0.002 (0.002) 0.455
27/28 1,799 -0.001 (0.001) 0.585
29/30 1,806 0.005 (0.002) 0.098
15+ drinks in the past 2 weeks
18 3,990 -0.001 (0.001) 0.126
19/20 2,132 -0.002 (0.002) 0.384
21/22 2,067 -0.004 (0.001) 0.009
23/24 2,038 -0.002 (0.002) 0.225
25/26 1,879 -0.002 (0.001) 0.032
27/28 1,799 0.000 (0.002) 0.979
29/30 1,806 0.004 (0.001) 0.003
Open in a new tab
a

Unweighted sample n for each noted age group.

b

Joinpoint = year in which significant change in slope occurred. If no joinpoint is noted for a specific age group, no statistically significant change in slope estimate was observed over time.

5+ drinking

Among age 18 respondents, 5+ drinking prevalence did not change significantly from 2005 to 2012; thereafter, a significant decline was observed from 2012 to 2015. Significant declines in 5+ drinking from 2005 to 2015 were also observed among individuals aged 19/20 and 23/24. Slope estimates for age groups 21/22, 25/26, and 27/28 were not statistically significant. A significant increase in 5+ drinking from 2005 to 2015 was observed among individuals aged 29/30. Results for 5-9 drinking and gender-specific binge drinking were generally similar to those for 5+ drinking, but the positive slope for 5-9 drinking at age 29/30 did not reach statistical significance.

For age groups with significant historical trends in 5+ drinking, modeled prevalence levels from 2005 and 2015 were compared. The modeled prevalence of 5+ drinking decreased 6 percentage points among individuals aged 18 (from 18% in 2005 to 12% in 2015); 20 percentage points among individuals aged 19/20 (from 32% to 12%); and 12 percentage points among individuals aged 23/24 (from 39% to 27%). In contrast, the prevalence of 5+ drinking increased 8 percentage points among individuals aged 29/30 (from 19% to 27%).

10+ drinking

Significant declines in 10+ drinking prevalence across time were observed among individuals aged 19/20. Trends for age groups 18 and 21/22 to 29/30 were not statistically significant. Comparison of 10+ prevalence from 2005 to 2015 for individuals aged 19/20 showed that the modeled prevalence of 10+ drinking decreased 9 percentage points (from 14% to 5%). Results for 10-14 drinking were generally similar to those for 10+ drinking, with a significant decline across time for individuals aged 19/20. However, prevalence of 10-14 drinking significantly increased among individuals aged 27/28 from 2005 to 2011. Slope estimates for 10-14 drinking for other age groups were not significantly significant.

15+ drinking

Significant declines in 15+ drinking two-week prevalence from 2005 to 2015 were observed among individuals aged 21/22 and 25/26. Slope estimates for those aged 18, 19/20, 23/24, and 27/28 were not statistically significant. A significant increase in 15+ drinking from 2005 to 2015 was observed among individuals aged 29/30. Comparison of modeled 15+ prevalence from 2005 and 2015 for age groups with significant trends showed that the prevalence of 15+ drinking decreased 3 percentage points among individuals aged 21/22 (from 5% to 2%) and 5 percentage points among individuals aged 25/26 (from 6% to 1%). In contrast, the prevalence of 15+ drinking increased 5 percentage points among individuals aged 29/30 (from 0% to 5%).

Comparing binge and high-intensity drinking trends with trends in any alcohol use (Aim 2)

Trend estimates for past 12-month and past 30-day alcohol use prevalence from 2005 to 2015 by young adult age group are reported in Table 3. Among age 18 respondents, past 12-month prevalence significantly declined from 2005 to 2015. Past 30-day alcohol use at age 18 did not show significant change over time, but the negative slope (p=.050) from 2012 through 2015 is consistent with previously published findings showing a significant linear decrease in 30-day alcohol use at this age (Miech et al., 2016). Among age 19/20 respondents, significant decreases in both 12-month and 30-day alcohol use were observed from 2012 to 2015. Thus, there appears to be some degree of similarity between decreasing trends in overall alcohol use (past 12-month and/or past 30-day use) and decreasing trends in 5+ and 10+ drinking at ages 18 and 19/20. Slope estimates for trends in past 12-month and past 30-day alcohol use for all other young adult age groups were not significant, in contrast to significant negative trends in 5+ and/or 15+ drinking at ages 21/22, 23/24, and 25/26, and significant positive trends at age 29/30.

Table 3. Slope Estimates for Trends in Prevalence of Past 12-month and Past 30-day Alcohol Use among US Young Adults by Age Group, 2005-2015.

Age Na Slope 1 (SE) p Joinpointb Slope 2 (SE) p
Any past 12-month alcohol use
18 4,060 -0.008 (0.003) 0.011
19/20 2,143 -0.014 (0.003) 0.001
21/22 2,068 -0.005 (0.005) 0.299
23/24 2,054 -0.003 (0.005) 0.586
25/26 1,910 0.004 (0.004) 0.342
27/28 1,862 0.001 (0.003) 0.777
29/30 1,864 0.002 (0.002) 0.534
Any past 30-day alcohol use
18 4,071 -0.004 (0.004) 0.305 2012 -0.031 (0.013) 0.050
19/20 2,144 -0.014 (0.002) <0.001
21/22 2,073 -0.003 (0.005) 0.535
23/24 2,058 -0.005 (0.004) 0.158
25/26 1,912 0.004 (0.005) 0.455
27/28 1,865 0.005 (0.005) 0.321
29/30 1,865 0.004 (0.004) 0.281
Open in a new tab
a

Unweighted sample n for each noted age group.

b

Joinpoint = year in which significant change in slope occurred. If no joinpoint is noted for a specific age group, no statistically significant change in slope estimate was observed over time.

Gender differences in trends (Aim 3)

Limited gender differences in trends for binge and high-intensity drinking outcomes were observed. For 5+ drinking, significant regression estimates for gender-difference time series data (calculated as the difference between prevalence for men and women) were found at ages 23/24 and 25/26. (Results not tabled: For 5+ drinking at age 23/24, the linear term (LT) p=0.017, and the quadratic term (QT) p=0.042). Gender-specific Joinpoint models indicated that 5+ drinking prevalence at age 23/24 significantly decreased across time for both men and women, but with a steeper slope for men (women -0.007 [SE 0.003, p=0.034]; men -0.018 [SE 0.002, p<0.001]). For 5+ drinking at age 25/26, both LT and QT regression estimates were significant (p=0.024 and 0.043, respectively), and gender-specific models indicated slope estimates were 0.008 (SE 0.005; p=0.138) for women and -0.015 (SE 0.007; p=0.057) for men. Thus, trends in 5+ drinking were not significant for either men or women at age 25/26, but an indication of an increasing slope was observed for women, while an indication of a decreasing slope was observed for men. For 5-9 drinking, gender comparisons were different only at age 23/24 (LT p=0.348, QT p=0.034). Slope estimates for 5-9 drinking at age 23/24 were negative for both women and men, but higher variance for men resulted in a non-significant slope (women -0.008 [SE 0.003, p=.020]; men -0.011 [SE 0.005, p=.067]). Trends for gender-specific binge drinking moved in similar directions for men and women for most age groups. However, women began to experience the observed increase in prevalence earlier in young adulthood than men (age 25/26 for women; age 29/30 for men).

Significant regression estimates for 10+ and 15+ drinking gender-difference time series data were found only at age 21/22 (10+ drinking LT p=0.685, QT p=0.047; 15+ LT p=0.028 with no significant quadratic association). No gender differences in trends were observed for 10-14 drinking. Gender-specific Joinpoint models for 10+ drinking at age 21/22 showed no significant change over time for men (est. -0.008 [SE 0.005; p=0.199]). In contrast, 10+ estimates for women indicated no significant change from 2005 to 2009 (est. 0.016 [SE 0.015; p=0.307]), followed by a significant decrease from 2009 to 2015 (est. -0.019 [SE 0.006; p=0.020]). Joinpoint models for 15+ drinking at age 21/22 for men indicated a linear decrease from 2005 to 2015 (est. -0.008 [SE 0.002; p=0.012]). Estimates for 15+ drinking among women indicated no significant change over time (est. 0.004 [SE 0.002; p=0.119] from 2005 to 2011, and est. -0.008 [SE 0.004; p=0.075] from 2011 to 2015).

Discussion

Among national samples of US 12th graders and young adults, overall (2005-2015) two-week prevalence for 5+ and 10+ drinking was highest for young adults aged 21/22 to 25/26; among these age groups, approximately 3 in 10 reported 5+ drinking, and 1 in 10 reported 10+ drinking. Recent prevalence of 15+ drinking was relatively constant from ages 18 through 23/24, averaging 5%. Between 2005 and 2015, trends showed that prevalence of binge and high-intensity drinking generally decreased for individuals in the early 20s, remained somewhat stable for individuals in mid-20s, and generally increased for individuals at age 29/30. Furthermore, in terms of specific levels of drinking, prevalence of 10-14 drinking significantly increased among individuals aged 27/28 from 2005 to 2011 and remained constant thereafter. Prior studies have speculated that increases in high-intensity drinking help explain the increases in alcohol-related emergency department visits (Hingson & White, 2013; Patrick et al., 2013), although the current study suggests these increases are evident only in certain age groups. Additional research is needed to document the extent to which these trends of increasing high-intensity drinking among older young adults have actually contributed to increases in alcohol-related ED visits. The observed trends in binge and high-intensity drinking were similar to those for past 12-month and past 30-day alcohol use at ages 18 and 19/20 but diverged for other young adult age groups, especially age 29/30. While trends have moved in similar directions for men and women in early and late young adulthood, women have experienced the observed increase in prevalence earlier in young adulthood (age 25/26) than men (age 29/30).

Binge drinking is associated with a wide range of negative health outcomes, including unintentional and intentional injuries, alcohol poisoning, sexually transmitted diseases, unintended pregnancy, fetal alcohol spectrum disorders, cardiovascular and liver disease, neurological damage, sexual dysfunction, and poor diabetes control (CDC, 2015). These risks rise as the quantity of alcohol consumed increases. Adults who reported consuming 5+ drinks at least once in the past 30 days were found to be 14 times more likely to drive while impaired compared to individuals who drank but not at such high levels (Naimi et al., 2003). Young adults who consumed 15+ drinks were even more likely to drive after drinking than those consuming fewer maximum drinks and were also more likely to drink after experiencing negative consequences from drinking (Hingson & White, 2013). In the current study, between ages 21/22 and 25/26, roughly 30% of US young adults reported 5+ drinking, and about 11% reported 10+ drinking in the past two weeks during the years 2005-2015. While rates decreased somewhat in later young adulthood, more than one quarter continued to report 5+ drinking and just under 10% reported 10+ drinking by age 29/30. These results indicate that, on average in the US in the past decade, more than a quarter of legal age young adults are engaging in drinking behavior associated with high individual and public health risks. Such results help explain that the burden of alcohol-attributable deaths from acute causes is highest among individuals aged 20-34 (CDC, 2013).

Of special importance was the finding that the recent prevalence of 15+ drinking remained relatively constant at about 5% from ages 18 to 23/24, spanning both underage drinkers and those of legal drinking age. In contrast, prevalence of both 5+ and 10+ drinking were noticeably lower for individuals below the legal drinking age (ages 18 and 19/20). Thus, unlike 5+ and 10+ drinkers, individuals who engage in 15+ drinking may reflect a unique risk population that may not be as sensitive to normative drinking policy approaches. Consumption of 15+ drinks on an empty stomach has been estimated to result in BAC levels above 0.30% for men and 0.45% for women (Hingson & White, 2013), levels associated with severe and/or life-threatening impairment (NIAAA, 2015). Thus, the sub-population of 15+ drinkers appears to be a uniquely high-risk group, both in regards to a lack of normative policy response and degree of impairment.

Age-specific changes in prevalence of binge and high-intensity drinking

The current study indicated that significant decreases in binge and/or high-intensity drinking were observed for ages 18 through 25/26: 5+ drinking at age 18; 5+ and 10+ drinking at age 19/20; 15+ drinking at age 21/22; 5+ drinking at age 23/24, and 15+ drinking at age 25/26. In contrast, significant increases were observed at age 29/30 for both 5+ and 15+ drinking. These results support and expand on previous findings from both NESARC and NSDUH. In NESARC, stability over time was found for past-year-drinker average daily ethanol intake, overall drinking frequency, and prevalence of past 30-day gender-specific binge drinking among those aged 18-24, while trends showed significant increases in these outcomes for those aged 25-44 (Dawson et al., 2015). In NSDUH, significant declines were observed for those aged 18-25 in any past 30-day 5+ drinking as well as high-frequency 5+ drinking (engaging in 5+ drinking 5 or more times in the past 30 days) (Center for Behavioral Health Statistics and Quality, 2015). In contrast, for those aged 26 and older, trends showed significant increases in any 5+ drinking and stability for high-frequency 5+ drinking (Center for Behavioral Health Statistics and Quality, 2015).

Trends in high-risk drinking were similar to those for overall 12-month and 30-day alcohol use among young adults under the legal drinking age (18 and 19/20), but not for most other age groups in this key developmental period. There is a need for research examining possible co-occurring changes over time in factors such as the perceived risks of, reasons for, or meanings/functions of high-risk drinking across different ages of young adulthood that may help explain the observed differences in behavior trends across age, and the lack of similarity with overall drinking prevalence for most age groups above the legal drinking age. The results of the current study support efforts to monitor a range of alcohol consumption measures in order to accurately capture changes in drinking behaviors and co-occurring risks.

Gender differences in prevalence and trends

The current study found that from 2005 to 2015, mean prevalence for 5+, 10+, and 15+ drinking was significantly higher for men than women across young adulthood. These results are in line with prior binge and high-intensity drinking research (Jager et al., 2015; Johnston et al., 2016a; Kerr et al., 2009, 2014; Keyes & Miech, 2013; Patrick et al., 2013; Patrick & Terry-McElrath, 2016; Terry-McElrath & Patrick, 2016). Yet, the current study found that mean prevalence for 5-9 drinking, as well as gender-specific binge drinking, showed no significant gender differences from ages 18 through 23/24. This indicates that across young adulthood, men are consistently more likely than women to engage in high-intensity drinking. However, men and women appear to be equally likely to participate in drinking at binge levels below the 10+ high-intensity threshold during early young adulthood.

Few significant differences by gender were observed. In general, for both men and women, participation in binge and high-intensity drinking showed similar rates of increase, decrease, or stability within age group. These results are similar to other studies that have documented positive period effects of increasing alcohol use for both men and women (Dawson et al., 2015; Keyes & Miech, 2013). Some important gender differences were observed; these generally indicated that for some age groups, gender differences were narrowing, either from prevalence among men decreasing faster than among women, or from prevalence among women increasing faster than among men. At age 21/22, 15+ drinking significantly decreased across time for men but did not change significantly for women. At age 23/24, negative slopes were observed for 5+ drinking prevalence for both men and women, but with steeper slopes for men. At age 25/26, men had a non-significant decrease in 5+ drinking, while women had a non-significant increase in the behavior over time. For gender-specific binge drinking, women aged 25-30 have increased prevalence in the past decade, while only men aged 29-30 have increased similarly. These results are supported by prior work that has found increases in high-risk drinking have occurred faster for women than men when combining across age (Dawson et al., 2015; Johnston et al., 2016a; Keyes & Miech, 2013). While a significant decrease in 10+ drinking is evident at age 21/22 among women (and not men), the other gender differences in the current study indicate that at ages 21/22 and 23/24, men are experiencing faster rates of decline in 5+ and 15+ drinking than women, and women may be experiencing the observed increase in use earlier in young adulthood than men.

Limitations

These findings must be considered within the study's limitations. The sample was based on 12th grade students and did not include the high school dropout population. Those who drop out of high school report higher prevalence of 5+ drinking in the past 30 days (Substance Abuse and Mental Health Services Administration, 2013). Further, attrition raises the possibility of bias in prevalence estimates of high-risk alcohol use, although the use of attrition weights results in recapturing baseline sample distributions on relevant variables. Data were based on self-reports; further research on the validity of self-reported high-intensity drinking is needed (Northcote & Livingston, 2011; Patrick, 2016). A further limitation is that 10+ and 15+ drinking was assessed with the same items for both genders, rather than gender-specific cut-offs that help to account for differences in average body size and alcohol metabolism. Finally, the measures of binge and high-intensity drinking assessed prevalence in the past two weeks and did not ask respondents to record their maximum number of drinks. Use of a two-week time frame limits the ability to examine variation in the frequency of high-intensity drinking. Future research should consider assessing patterns in the frequency of high-intensity drinking over longer time spans and using maximum number of drinks consumed. These limitations notwithstanding, the current analyses provide needed data on age-specific trends in binge and high-intensity drinking in a national sample of young adults.

Conclusions

Trends from 2005 to 2015 in binge and high-intensity drinking among US 12th graders and young adults did not simply follow overall alcohol use prevalence trends; instead, high-intensity drinking tended to decline among 18-22 year-olds, to be stable among 23-28 year-olds, and to increase among 29-30 year-olds. These trends were generally similar for men and women, but the historical gender gap in high-risk drinking appeared to be decreasing among some age groups due to differing gender trends. These findings illustrate that high-intensity drinking among US young adults is a dynamic phenomenon. Prevention and intervention efforts aimed at reducing the harms resulting from high-intensity drinking should acknowledge and focus on differences in trends in these behaviors by age and gender.

Supplementary Material

Supp info

Acknowledgments

Development of this manuscript was supported by research grant R01AA023504 (to M. Patrick) from the National Institute on Alcohol Abuse and Alcoholism. Data collection and manuscript preparation were supported by research grants R01DA001411 and R01DA016575 (to L. Johnston) from the National Institute on Drug Abuse. The study sponsors had no role in the study design, collection, analysis or interpretation of the data, writing of the manuscript, or the decision to submit the paper for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the study sponsor.

References

  1. Bachman JG, Johnston LD, O'Malley PM, Schulenberg JE, Miech RA. The Monitoring the Future project after four decades: design and procedures (Monitoring the Future Occasional Paper No. 82) Ann Arbor, MI: Institute for Social Research, University of Michigan; 2015. [Google Scholar]
  2. Center for Behavioral Health Statistics and Quality. Behavioral health trends in the United States: Results from the 2014 National Survey on Drug Use and Health (HHS Publication No. SMA 15-4927, NSDUH Series H-50) 2015 [Google Scholar]
  3. Centers for Disease Control and Prevention. [Accessed September 22 2016];Alcohol Related Disease Impact (ARDI) application. 2013 Available at www.cdc.gov/ARDI.
  4. Centers for Disease Control and Prevention. [Accessed September 22 2016];Fact sheets – Binge drinking. 2015 Available at http://www.cdc.gov/alcohol/fact-sheets/binge-drinking.htm.
  5. Dawson DA, Goldstein RB, Saha TD, Grant BF. Changes in alcohol consumption: United States, 2001-2002 to 2012-2013. Drug Alc Depend. 2015;148:56–61. doi: 10.1016/j.drugalcdep.2014.12.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Hingson RW, White A. Trends in extreme binge drinking among US high school seniors. JAMA Pediatr. 2013;167(11):996–998. doi: 10.1001/jamapediatrics.2013.3083. [DOI] [PubMed] [Google Scholar]
  7. Jager J, Keyes KM, Schulenberg JE. Historical variation in young adult binge drinking trajectories and its link to historical variation in social roles and minimum legal drinking age. Dev Psych. 2015;51(7):962–974. doi: 10.1037/dev0000022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Johnston LD, O'Malley PM, Bachman JG, Schulenberg JE, Miech RA. Demographic subgroup trends among young adults in the use of various licit and illicit drugs, 1989-2015 (Monitoring the Future Occasional Paper 87) Ann Arbor, MI: Institute for Social Research; 2016a. [Accessed September 22 2016]. Available at monitoringthefuture.org/pubs/occpapers/mtf-occ87.pdf. [Google Scholar]
  9. Johnston LD, O'Malley PM, Bachman JG, Schulenberg JE, Miech RA. Monitoring the Future national survey results on drug use, 1975-2015: Volume II, college students and adults ages 19-55. Ann Arbor, MI: Institute for Social Research, The University of Michigan; 2016b. [Accessed September 22 2016]. Available at http://monitoringthefuture.org/pubs/monographs/mtf-vol2_2015.pdf. [Google Scholar]
  10. Kerr WC, Mulia N, Zemore SE. U.S. trends in light, moderate, and heavy drinking episodes from 2000 to 2010. Alcohol Clin Exp Res. 2014;38(9):2496–2501. doi: 10.1111/acer.12521. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Kerr WC, Greenfield TK, Bond J, Ye Y, Rehm J. Age-period-cohort modelling of alcohol volume and heavy drinking days in the US National Alcohol surveys: divergence in younger and older adult trends. Addiction. 2009;104:27–37. doi: 10.1111/j.1360-0443.2008.02391.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Keyes KM, Grant BF, Hasin DS. Evidence for a closing gender gap in alcohol use, abuse, and dependence in the United States population. Drug Alcohol Depend. 2008;93:21–29. doi: 10.1016/j.drugalcdep.2007.08.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Keyes KM, Li G, Hasin DS. Birth cohort effects and gender differences in alcohol epidemiology: a review and synthesis. Alcohol Clin Exp Res. 2011;35:2101–2112. doi: 10.1111/j.1530-0277.2011.01562.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Keyes KM, Miech RA. Age, period, and cohort effects in heavy episodic drinking in the US from 1985-2009. Drug Alcohol Depend. 2013;132:140–148. doi: 10.1016/j.drugalcdep.2013.01.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kim HJ, Fay MP, Feuer EJ, Midthune DN. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med. 2000;19:335–51. doi: 10.1002/(sici)1097-0258(20000215)19:3<335::aid-sim336>3.0.co;2-z. (correction: 2001; 20:655) [DOI] [PubMed] [Google Scholar]
  16. Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, Adair-Rohani H. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2224–60. doi: 10.1016/S0140-6736(12)61766-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Miech RA, Johnston LD, O'Malley PM, Bachman JG, Schulenberg JE. Monitoring the Future national survey results on drug use, 1975-2015: Volume I, secondary school students. Ann Arbor, MI: Institute for Social Research, The University of Michigan; 2016. [Accessed January 31 2017]. Available at http://monitoringthefuture.org/pubs/monographs/mtf-vol1_2015.pdf. [Google Scholar]
  18. Naimi TS, Brewer RD, Mokdad A, Denny C, Serdula MK, Marks JS. Binge drinking among US adults. JAMA. 2003;289(1):70–75. doi: 10.1001/jama.289.1.70. [DOI] [PubMed] [Google Scholar]
  19. Naimi TS, Nelson DE, Brewer RD. The intensity of binge alcohol consumption among US adults. Am J Prev Med. 2010;38(2):201–207. doi: 10.1016/j.amepre.2009.09.039. [DOI] [PubMed] [Google Scholar]
  20. National Cancer Institute. Joinpoint Regression Program, Version 4.3.1.0. Bethesda, MD: Statistical Research and Applications Branch, National Cancer Institute; 2016a. [Google Scholar]
  21. National Cancer Institute. [Accessed October 3 2016];Number of joinpoints. 2016b Available at https://surveillance.cancer.gov/help/joinpoint/setting-parameters/advanced-tab/number-of-joinpoints.
  22. National Cancer Institute. [Accessed September 9 2016];Jointpoint frequently asked questions: correlation in two series. 2016c Available at https://surveillance.cancer.gov/joinpoint/faq/correlation.html.
  23. National Center for Health Statistics. Health, United States, 2012 with special feature on emergency care. Hyattsville, MD: 2013. [PubMed] [Google Scholar]
  24. National Institute on Alcohol Abuse and Alcoholism. [Accessed September 20 2016];Drinking levels defined. 2016 Available at https://www.niaaa.nih.gov/alcohol-health/overview-alcohol-consumption/moderate-binge-drinking.
  25. National Institute on Alcohol Abuse and Alcoholism. [Accessed September 22 2016];Alcohol overdose: the dangers of drinking too much. 2015 Available at http://pubs.niaaa.nih.gov/publications/AlcoholOverdoseFactsheet/Overdosefact.htm.
  26. Naimi TS, Brewer RD, Mokdad A, Denny C, Serdula MK, Marks JS. Binge drinking among US adults. JAMA. 2003;289:70–75. doi: 10.1001/jama.289.1.70. [DOI] [PubMed] [Google Scholar]
  27. Northcote J, Livingston M. Accuracy of self-reported drinking: observational verification of ‘last occasion ’ drink estimates of young adults. Alcohol Alcohol. 2011;46(6):709–713. doi: 10.1093/alcalc/agr138. [DOI] [PubMed] [Google Scholar]
  28. Patrick ME. A call for research on high-intensity alcohol use. Alcohol Clin Exp Res. 2016;40(2):256–259. doi: 10.1111/acer.12945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Patrick ME, Schulenberg JE, Martz ME, Maggs JL, O'Malley PM, Johnston LD. Extreme binge drinking among 12-grade students in the United States: prevalence and predictors. JAMA Pediatr. 2013;167(11):1019–1025. doi: 10.1001/jamapediatrics.2013.2392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Patrick ME, Terry-McElrath YM. High-intensity drinking by underage young adults in the United States. Addiction. 2016 doi: 10.1111/add.13556. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Patrick ME, Terry-McElrath YM, Kloska DD, Schulenberg JE. High-intensity drinking among young adults in the United States: prevalence, frequency, and developmental change. Alcohol Clin Exp Res. 2016 doi: 10.1111/acer.13164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sacks JJ, Gonzales KR, Bouchery EE, Tomedi LE, Brewer RD. National and state costs of excessive alcohol consumption. Am J Prev Med. 2010;49(5):e73–e79. doi: 10.1016/j.amepre.2015.05.031. [DOI] [PubMed] [Google Scholar]
  33. Schulenberg JE, Maggs JL. A developmental perspective on alcohol use and heavy drinking during adolescence and the transition to young adulthood. J Stud Alc Suppl. 2002;14:54–70. doi: 10.15288/jsas.2002.s14.54. [DOI] [PubMed] [Google Scholar]
  34. Substance Abuse and Mental Health Services Administration. The NSDUH Report: Substance use among 12th grade aged youths by dropout status. Rockville, MD: Substance Abuse and Mental Health Services Administration, Center for Behavioral Health Statistics and Quality; Feb 2, 2013. [Accessed January 31, 2017]. Available at https://www.samhsa.gov/data/sites/default/files/NSDUH036/NSDUH036/SR036SubstanceUseDropouts.htm. [Google Scholar]
  35. Terry-McElrath YM, Patrick ME. Intoxication and binge and high-intensity drinking among US young adult sin their mid-twenties. Sub Abuse. 2016 doi: 10.1080/08897077.2016.1178681. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. White AM, Kraus CL, Swartzwelder H. Many college freshmen drink at levels far beyond the binge threshold. Alcohol Clin Exp Res. 2006;30(6):1006–1010. doi: 10.1111/j.1530-0277.2006.00122.x. [DOI] [PubMed] [Google Scholar]
  37. World Health Organization. Global Status Report on Alcohol and Health, 2014 ed. Geneva, Switzerland: Management of Substance Abuse, Department of Mental Health and Substance Abuse, World Health Organization; 2014. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp info
NIHMS872824-supplement-Supp_info.pdf (99.7KB, pdf)

RESOURCES