Demographic, substance use, and mental well-being correlates of high-intensity drinking among college students and non-college young adults: implications for intervention

Abstract

Objective

To assess demographic, substance use, and mental wellbeing factors associated with high-intensity drinking (HID; 10+ drinks on one occasion) among college- and non-college young adults, to inform prevention and intervention efforts.

Participants

A total of 1,430 young adults (819 in college and 611 not attending college) in a Midwestern state who reported trying alcohol at least once.

Methods

Participants were recruited via social media between November 2019 and February 2020 to complete a web-based survey assessing demographics, substance use, and mental well-being. Logistic regression was conducted to assess relationships between these measures and HID among (1) college students and (2) non-college young adults.

Results

About 14.0% of participants reported past-month HID. Among both college- and non-college young adults, men, those who perceived slight or no risk of harm from binge drinking, and those who used alcohol and marijuana simultaneously in the past year had greater odds of reporting past-month HID. Among students, past-year prescription drug misuse was also associated with HID.

Conclusions

High intensity-drinking is concerning given potential adverse consequences. Campus programming should address norms that may promote such drinking and other high-risk substance use associated with HID.

Keywords:

• College health
• college students
• high-intensity drinking
• high-risk alcohol use
• high-risk drinking prevention

Introduction

High-intensity drinking (HID), having 10+ drinks on one occasion,¹ may have even more severe consequences than binge drinking. Initial calls for research into HID were based on the recognition that categorizing all drinkers into one risk category, binge drinking, did not cover the full range of problematic drinking behaviors and how these behaviors might differentially impact adverse outcomes.^2,3 Binge drinking definitions do not distinguish between drinking at or just above the binge threshold and far exceeding that threshold.⁴ HID is particularly concerning because of the adverse consequences that have been associated with it. Evidence from the Monitoring the Future Survey indicates HID is associated with acute outcomes such as blackouts and alcohol poisoning.^1,4,5 There is also a greater likelihood of injuries, risky sexual behavior, and academic problems.⁴ In addition, data from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC)-III show that adults who engage in HID have three times greater odds of experiencing alcohol use disorder than those who reported binge-only drinking.⁶ Results for young adults were particularly striking, as 83% of 18-year-olds who reported HID met the criteria for alcohol use disorder.⁶ While research shows that college-attending young adults are particularly likely to engage in HID,^5,7 little is known about correlates of HID among college students. Research on binge drinking among college students and young adults not attending college may provide clues.

Binge drinking, defined as five or more drinks in a two hour period,⁵ is a widespread problem on college campuses. According to the American College Health Association,⁸ 25% of undergraduate students have engaged in binge drinking in the past 2 weeks. Demographic factors correlated with binge drinking include younger age, being male, and identifying as White.^9,10 Additionally, a review found that younger age (<16) when consuming one’s first drink was associated with higher likelihood of binge drinking during college.¹⁰ Population surveys suggest that people who are sexual minorities are at elevated risk of alcohol use disorder and binge drinking.^11–13 Some substance use perceptions and behaviors have also been correlated with greater likelihood of binge drinking. Hanauer et al.⁹ found that lower perceived risk of harm from binge drinking was associated with more days of binge drinking in the past month. Additionally, any past-month marijuana use has been associated with binge drinking,¹⁴ as has more frequent past-month marijuana use.¹⁵ Prescription drug misuse has also been linked with binge drinking^16,17 and more generally, with alcohol misuse among young adults.¹⁸ Finally, mental health challenges have been linked with an increased likelihood of binge drinking¹⁷, which in turn may reduce well-being by increasing psychological distress symptoms.¹⁹

Researchers have identified that several of the above factors are also associated with HID, though it remains unclear how they may be so among college students specifically. Because college students are more likely to engage in HID than non-college young adults, they may be at particular risk of associated consequences.⁷ The odds of engaging in HID are particularly high among four-year students not living with their parents.²⁰ Furthermore, by the mid-20s, there is a higher prevalence of HID among people who have earned a four-year college degree compared to nongraduates.⁵

Various demographic factors have been associated with HID among general populations of young adults. Notably, some of these studies have defined HID as 10+ drinks, while others have used a biological sex-specific definition of 8 drinks for females and 10 drinks for males. Using data from NESARC III, Hingson et al.²¹ found that people 21–25 years old were most likely to engage in HID among all adults. Studies of HID from early adolescence through adulthood, using the sex-specific definition of 8 drinks for females and 10 drinks for males^21,22 or the gender-neutral definition of 10+ drinks,⁵ have shown men are more likely to participate in HID. During their first year of college, men are more likely to engage in HID.³ Subsequently, men aged 19–30 are also more likely to report past-month HID.²³ In addition, White young adults are more likely to participate in HID than young adults of color.^5,21 Additional research is needed to determine if these demographic correlates of HID remain in place among college students.

Research on sexual orientation and HID among college students is limited, but evidence from population surveys provides insight about sexual orientation and HID. A large national survey found that sexual minority women were more likely than heterosexual women to engage in HID, and that bisexual men were more likely to consume 15+ drinks than were heterosexual men.²⁴ Evidence from NESARC III also indicated that sexual minority women had greater odds of engaging in HID than heterosexual women.²⁵ Interestingly, sexual minority men were generally less likely to report engaging in HID.²⁵ A similar pattern was found in the 2015 Youth Risk Behavior Survey,²⁶ wherein both lesbian and bisexual girls reported significantly higher rates of HID than heterosexual girls, and gay boys were significantly less likely to partake in HID than were heterosexual boys. Given these findings and the prevalence of HID among college students, further examination of HID and sexual orientation among college students is needed.

In addition to research focused on demographic factors associated with HID, alcohol and other substance use behaviors may also be associated with HID. Patrick et al.⁵ found that individuals who started drinking before high school were more likely to engage in HID in their early to mid-20s. HID participation in 12th grade is associated with a higher risk for simultaneous use of alcohol and marijuana.^27*, Combined use of alcohol and other drugs peaks from 18 to 24,³¹ and this risk may persist into young adulthood. Any marijuana use in the past 30 days and a greater frequency of marijuana use were associated with a greater likelihood of HID compared to moderate drinking.⁵ Regarding prescription drug misuse, McCabe et al.³² found that past-month misuse of opioids, sedatives, stimulants, or tranquilizers was associated with HID, while Bonar et al.²² found that only stimulant misuse was associated with HID. These findings are concerning because of risks associated with simultaneous use of alcohol and prescription drugs, such as overdose, accidents, and injuries.^33–35 Similarly, simultaneous alcohol and marijuana (SAM) use increases the risk of blackouts.³⁶ Prescription drug misuse among college students is concerning; for example, a recent systematic review found that anywhere from 4% to 19.7% of students had misused prescription opioids during their lifetime.³⁷ SAM is also particularly prevalent, as findings from a large college student sample indicate that 72% had ever used marijuana and among them, two-thirds reported past year SAM.³⁶ Students who misuse prescription drugs or engage in SAM may be at greater risk of engaging in HID and of severe consequences, and further examination of how these behaviors are related to HID is warranted. Additionally, because perceived harm of binge drinking is related to actual binge drinking behavior, it is worth examining whether risk perceptions are also related to HID.

Regarding mental health, previous work has examined whether coping (e.g., reducing depression or nervousness) and enhancement (i.e., of positive mood) motives for substance use are associated with HID. Using data from Project INTEGRATE, which pools data from college students participating in multiple studies across several campuses, White et al.³⁸ found that coping motives were more common among participants who engaged in HID than among those who engaged in binge drinking or no drinking. Conversely, using data from the Young Adult Daily Life Study, Patrick et al.⁵ found that coping motives were not stronger among high-intensity drinkers than binge-only drinkers. These authors also found that depression was more common among young adults who reported HID than those who engaged in binge drinking or moderate drinking; there was no association between anxiety and drinking levels.⁵ Moreover, previous work has associated greater enhancement motives with HID.^5,20,39 Conversely, it remains unclear whether in contrast to mental health challenges, mental well-being is protective against HID.

Gaps and purpose of the study

While previous work shows that college students are more likely to engage in HID than non-students, questions remain about correlates of HID among college students. Additional research on demographic correlates would help to inform prevention and intervention efforts. In addition, there is limited knowledge about the potential links between other substance use and HID among college students. Furthermore, while previous work has assessed links between mental health symptomology and HID,^5,38 there is limited knowledge of how mental well-being (i.e., not just the absence of symptoms) may be associated with HID. Additionally, if correlates of HID differ between college students and non-college young adults, prevention programming needs to be developed according to the context (college or non-college). Thus, this study examines how demographics, substance use, and mental well-being are associated with HID among both college students and young adults not attending college to highlight targeted strategies to reduce HID.

Materials and methods

A sample of young adults in Michigan aged 18–25 were recruited to complete a web-based survey on substance use attitudes and behaviors (N = 1,751). Participants were recruited from November 2019 to February 2020 using paid advertisements on Facebook and Instagram, two widely used social media outlets.⁴⁰ Based on evidence that social media recruitment can result in the over-representation of White participants,⁴¹ we oversampled in zip codes with higher percentages of Black, Hispanic or Latino, Asian American, and American Indian or Alaskan Native residents.^42,43 We also over-sampled in zip codes with colleges and universities to increase our reach in college communities.

Participants provided information on their demographics, substance use, mental well-being, and other topics not included in the present study. This analysis was limited to the subsample of young adults who reported a lifetime history of alcohol use (n = 1,430). The survey took approximately 15–20 min to complete. Due to the susceptibility of incentivized online surveys to produce low-quality and fraudulent data,⁴⁴ we did not provide individual incentives to encourage participation. Instead, respondents could enter a drawing for one of five $100 gift cards. To ensure adequate data quality, respondents who had unrealistic completion times (<5 min) and nonsensical or duplicate responses to open-ended items were removed from the sample.⁴⁴ The Wayne State University Institutional Review Board approved all study procedures.

Measures

Demographic variables

Participants were asked to report their age (in years), gender identity (man, woman, transgender, non-binary, or another gender identity, with an open response option), race/ethnicity (White, African American/Black, Hispanic/Latino, Asian American, American Indian/Alaska Native, Native Hawaiian/Other Pacific Islander, or Multiracial), and sexual orientation (straight/heterosexual, gay/homosexual, bisexual, pansexual, queer, or asexual, with an open-text response option for other sexual orientations). Age was dichotomized (<21 = 0 and 21+ = 1) to compare students of legal drinking age to those who were not. Because White respondents comprised 85% of the sample, race was trichotomized as non-Hispanic White (=0), non-Hispanic Black (=1), or other people of color (=2) to ensure the statistical power needed to detect group differences. Similarly, due to relatively small numbers of participants who identified as transgender, non-binary, or another gender identity, we trichotomized gender as man (=1), woman (=2), or transgender, non-binary, or another gender identity (=3). Finally, because previous research has suggested an increased risk of problematic drinking among sexual minority adults, sexual orientation was dichotomized as straight/heterosexual orientation (=0) or sexual minority orientation (=1).

Age of first alcohol use

Participants were asked to indicate the age of their first use of alcohol. Previous work examining the association of onset of alcohol or other substance use with later substance use has compared onset before age 14 to onset at 14 or older.^45,46 Thus, we compared participants who reported any alcohol use before age 14 (=1) to those who did not drink until age 14 or older (=0).

Perceived risk of binge drinking

Participants were asked about their perceived risk of harm from having five or more drinks once or twice a week, using an item adapted from Monitoring the Future.⁴⁷ Response options were “no risk,” “slight risk,” “moderate risk,” and “great risk.” For the current analysis, responses were dichotomized as moderate/great risk (=0) or no/slight risk (=1).

Other substance use

Marijuana and prescription drug use were assessed with items adapted from the National Survey of Drug Use and Health.⁴⁸ Participants were asked to report how many days in the past 30 days they used marijuana. Participants were also asked to indicate if they had ever used four classes of prescription drugs nonmedically: opioids, sedatives, stimulants, and sleeping medications. Nonmedical use of prescription drugs was defined as “use in any way not directed by a doctor, including use without a prescription of one’s own; use in greater amounts, more often, or longer than told to take a drug; or use in any other way not directed by a doctor.”⁴⁸ Participants were also asked if they had used alcohol and marijuana simultaneously. Response options were “within the past 30 days”, “within the past year, but not within the past 30 days”, “in my lifetime, but not in the past year”, or “never”. For the current analysis, each variable was dichotomized as no past-year use (=0) versus past-year use (=1).

Mental well-being

Mental well-being was assessed with the Warwick-Edinburgh Mental Well-being Scale (WEMBS).⁴⁹ The scale presents 14 feelings or thoughts representing a state of mental well-being, such as “I’ve been feeling relaxed” and “I’ve been feeling close to other people.” Participants indicated how often they experienced each feeling or thought over the past two weeks on a scale ranging from 1 = none of the time to 5 = all of the time. Items were summed to create total scores ranging from 14 to 70. Higher scores indicated better mental well-being (Cronbach’s ɑ = 0.937).

High-intensity drinking

HID was assessed by asking respondents how many days in the past 30 (if any) they consumed 10 or more drinks in a row.^1,7 Response options were provided on a six-point scale: none, once or twice, three to five times, six to 10 times, 11–19 times, and 20 or more times. Since most participants (86.0%) did not report HID in the past 30 days, this variable was dichotomized to compare participants who reported HID (=1) to those who did not (=0).

Data analysis

Data analyses were conducted in IBM SPSS Statistics version 29 (IBM SPSS Statistics, Armonk, NY) using multiple imputation with 20 imputations for missing data.⁵⁰ Data were screened for outliers and multicollinearity among independent variables. Cook’s Distance was used to test for extreme data points that may distort results, with no cases greater than 0.50.⁵¹ Correlational analyses showed no evidence of problems with multicollinearity. A chi-square test was conducted to assess whether students and non-college young adults differed on likelihood of engaging in past-month HID. Results indicated that there was no significant difference. To assess correlates of HID among each group, analyses were conducted for students (model 1) and other young adults (model 2). For each model, descriptive statistics were calculated for all independent and dependent variables. Chi-square tests were conducted to assess the relationship between each categorical predictor and HID independently. Bivariate logistic regression was used to examine the pair-wise relationships between each continuous predictor and HID. Finally, multivariable logistic regression was used to determine which independent variables were associated with the odds of engaging in past-month HID. The significance threshold was set at p < .05.

Results

Model 1: College students

Descriptive and bivariate statistics are presented in Table 1. About 60.3% of the sample (n = 494) was 21 or older. The sample was largely made up of women (71.5%, n = 586), but also 23.7% men (n = 194) and 4.8% individuals who reported a transgender, non-binary, or another gender identity (n = 39). Most participants identified as White (85.0%, n = 696), though 2.1% (n = 17) identified as Black and 12.9% (n = 106) identified as another race or ethnicity. About two-thirds of the students (67.9%, n = 556) identified as straight/heterosexual, while 32.1% (n = 263) identified with a sexual minority orientation (i.e., gay/homosexual, bisexual, pansexual, queer, asexual, or another orientation). About 15.8% of students (n = 129) reported initiation of alcohol use under age 14. A third (33.2%, n = 272) believed there was slight or no risk of harm from having five-plus drinks once or twice a week. The mean number of days of past-month marijuana use was 7.14 (SD = 8.66), including the 65.8% of the sample who reported no past-month use. Nearly a third of students (30.8%, n = 252) reported past-year nonmedical prescription drug use and 42.7% (n = 350) reported using alcohol and marijuana simultaneously in the past year. Over half (54.6%, n = 447) reported any marijuana use in the past year, suggesting that SAM is prevalent among this sample. Wellness scores ranged from 14 to 70 (M = 45.93, SD = 10.42). Finally, 13.6% (n = 111) reported past-month HID, while 86.4% (n = 708) did not. At the bivariate level, male gender (p < .001), simultaneous use of alcohol and marijuana (p < .001), nonmedical prescription drug use (p < .001), perceived risk of binge drinking (p = .005), and days of past-month marijuana use (p < .001) were associated with HID in the past month.

Table 1. Descriptive and bivariate statistics – college students.

Variable, n (%)	Full sample (N = 819)	No HID (n = 708)	HID (n = 111)
Age
Under 21	325 (39.7)	279 (39.4)	46 (41.4)
21 or over	494 (60.3)	429 (60.6)	65 (58.6)
Race
White	696 (85.0)	597 (84.3)	99 (89.2)
Black	17 (2.1)	16 (2.3)	1 (0.9)
People of color, non-Black	106 (12.9)	95 (13.4)	11 (10.4)
Gender
Woman	586 (71.5)	522 (73.7)***	64 (57.7)***
Man	194 (23.7)	152 (21.5)***	42 (37.8)***
Transgender, non-binary, or another identity	39 (4.8)	34 (4.8)***	5 (4.5)***
Sexual orientation
Straight/heterosexual	556 (67.9)	484 (68.4)	72 (64.9)
Sexual minority	263 (32.1)	224 (31.6)	39 (35.1)
Age of first alcohol use
Under 14	129 (15.8)	109 (15.4)	20 (18.0)
14 or older	690 (84.2)	599 (84.6)	91 (82.0)
Frequency of marijuana use, past 30 days, M (SD)^a	7.14 (8.66)	6.56 (8.44)***	10.87 (9.13)***
Mixing marijuana and alcohol, past year
Yes	350 (42.7)	272 (38.4)***	78 (70.3)***
No	469 (57.3)	436 (61.6)***	33 (29.7)***
Prescription drug misuse, past year
Yes	252 (30.8)	199 (28.1)***	53 (47.8)***
No	567 (69.2)	509 (71.9)***	58 (52.2)***
Perceived binge drinking risk
No/slight risk	272 (33.2)	222 (31.4)**	50 (45.0)**
Moderate/great risk	547 (66.8)	486 (68.6)**	61 (55.0)**
Well-being, M (SD)^a	45.93 (10.42)	45.79 (10.34)	46.83 (10.93)

*p < .05; **p < .01; ***p < .001.

^aFrequency of marijuana use and well-being are continuous measures; all other variables are categorical.

Results of the multivariable logistic regression are summarized in Table 2. The overall model was significant (X² = 73.42, p < .001). Men had greater odds than women of reporting past-month HID (OR = 2.01, p = .004), but there was no difference between women and people who reported a transgender, nonbinary, or another gender identity. Past-year simultaneous use of alcohol and marijuana (OR = 2.90, p < .001), past-year nonmedical prescription drug use (OR = 2.01, p < .001), and low perceived risk of harm from binge drinking (OR = 1.62, p = .04) were also associated with greater odds of past-month HID. Age, race, sexual orientation, age of first alcohol use, past 30-day frequency of marijuana use, and well-being were not significantly associated with the odds of HID.

Table 2. Results of multiple logistic regression predicting high-intensity drinking among college students.

	B	SE	OR	95% CI	p
Aged 21 and older	−0.035	0.231	0.966	0.614 − 1.518	.880
Race/ethnicity
Black^a	−1.564	1.060	0.209	0.026 − 1.672	.140
Other person of color^a	−0.422	0.356	0.656	0.327 − 1.317	.236
Gender identity
Man^b	0.697	0.242	2.007	1.249 − 3.227	.004
Transgender, non-binary, or another identity^b	−0.033	0.549	0.967	0.330 − 2.835	.952
Sexual orientation	0.061	0.247	1.062	0.655 − 1.724	.806
Age of first alcohol use	0.134	0.303	1.143	0.632 − 2.070	.658
Frequency of marijuana use (past 30 days)^c	0.017	0.014	1.017	0.990 − 1.045	.209
Mixing marijuana and alcohol (past year)	1.064	0.263	2.899	1.730 − 4.859	<.001
Prescription drug misuse (past year)	0.698	0.231	2.010	1.279 − 3.161	.002
Perceived risk of binge drinking	0.480	0.231	1.615	1.028 − 2.539	.038
Well-being^c	0.014	0.011	1.014	0.992 − 1.036	.209

^aWhite is the reference group for race/ethnicity.

^bWoman is the reference group for gender identity.

^cFrequency of marijuana use and well-being are continuous measures.

Model 2: Non-student young adults

Descriptive and bivariate statistics are presented in Table 3. The vast majority (89.4%) of the sample of non-student young adults (n = 546) was 21 or older. The sample was largely comprised of women (64.8%, n = 396), but also included 28.0% men (n = 171) and 7.2% people who reported a transgender, non-binary, or another gender identity (n = 44). Similar to the sample of college students, most participants identified as White (86.1%, n = 526), while 3.4% (n = 21) identified as Black and 10.5% (n = 64) identified as another race or ethnicity. About two-thirds (65.1%, n = 398) identified as straight/heterosexual, while 34.9% (n = 213) identified with a sexual minority orientation (i.e., gay/homosexual, bisexual, pansexual, queer, asexual, or another orientation). About 18.2% (n = 111) reported initiation of alcohol use under age 14, and slightly less than a third (31.3%, n = 191) believed there was slight or no risk of harm from having five-plus drinks once or twice a week. The mean number of days of past-month marijuana use was 9.75 (SD = 10.34), including the 62.2% of participants who reported no past-month use. Nearly a third (30.8%, n = 188) reported past-year nonmedical prescription drug use and 39.6% (n = 242) reported using alcohol and marijuana simultaneously in the past year. Similar to college students, over half (56.3%, n = 344) of non-college young adults reported past-year marijuana use, suggesting that SAM is somewhat common among this sample. Wellness scores ranged from 14 to 70 (M = 45.86, SD = 10.42). Finally, 14.6% (n = 89) reported past-month HID, while 85.4% (n = 522) did not. At the bivariate level, correlates were the same as those for college students. Specifically, male gender (p = .007) simultaneous use of alcohol and marijuana (p < .001), nonmedical prescription drug use (p = .038), perceived risk of binge drinking (p < .001), and days of past-month marijuana use (p = .032) were associated with HID in the past month.

Table 3. Descriptive and bivariate Statistics – non-student young adults.

Variable, n (%)	Full sample (N = 611)	No HID (n = 522)	HID (n = 89)
Age
Under 21	65 (10.6)	58 (11.1)	7 (7.9)
21 or over	546 (89.4)	464 (88.9)	82 (92.1)
Race
White	526 (86.1)	454 (87.0)	72 (80.9)
Black	21 (3.4)	19 (3.6)	2 (0.2)
People of color, non-Black	64 (10.5)	49 (9.4)	15(18.9)
Gender
Woman	396 (64.8)	350 (67.0)**	46 (51.7)
Man	171 (28.0)	134 (25.7)**	37 (41.6)
Transgender, non-binary, or another identity	44 (7.2)	38 (7.3)**	6 (6.7)
Sexual orientation
Straight/heterosexual	398 (65.1)	335 (64.2)	63 (70.8)
Sexual minority	213 (34.9)	187 (35.8)	26 (29.2)
Age of first alcohol use
Under 14	111 (18.2)	93 (17.8)	18 (20.2)
14 or older	500 (81.8)	429 (82.2)	71 (79.8)
Frequency of marijuana use, past 30 days, M (SD)^a	9.75 (10.34)	9.34 (10.23)***	12.12 (10.76)***
Mixing marijuana and alcohol, past year
Yes	242 (39.6)	187 (35.8)***	55 (61.8)***
No	369 (60.4)	335 (64.2)***	34 (38.2)***
Prescription drug misuse, past year
Yes	188 (30.8)	153 (29.3)*	35 (39.3)*
No	423 (69.2)	369 (70.7)*	54 (60.7)*
Perceived binge drinking risk
No/slight risk	191 (31.3)	148 (28.4)**	43 (48.3)**
Moderate/great risk	420 (68.7)	374 (71.6)**	46 (51.7)**
Well-being, M (SD)^a	45.86 (10.42)	46.23 (10.12)	43.73 (11.84)

*p < .05; **p < .01; ***p < .001.

^aFrequency of marijuana use and well-being are continuous measures; all other variables are categorical.

Results of the multivariable logistic regression are summarized in Table 4. The overall model was significant (X² = 50.70, p < .001). Correlates of HID were similar to those for college students, with the exception of past-year prescription drug misuse (p = .315). Among non-student young adults, men had greater odds than women of reporting past-month HID (OR = 2.03, p = .009), but there was no difference between women and people who reported a transgender, nonbinary, or another gender identity. Past-year simultaneous use of alcohol and marijuana (OR= 2.64, p < .001), and low perceived risk of harm from binge drinking (OR = 2.08, p = .004) were also associated with greater odds of past-month HID. In addition to past-year prescription drug misuse, age, race, sexual orientation, age of first alcohol use, past 30-day frequency of marijuana use, and well-being were not significantly associated with the odds of HID.

Table 4. Results of multiple logistic regression predicting high-intensity drinking among non-student young adults.

	B	SE	OR	95% CI	P
Aged 21 and older	0.667	0.494	1.948	0.739–5.135	.177
Race/ethnicity
Black^a	−0.193	0.842	0.824	0.158–4.307	.818
Other person of color^a	0.509	0.344	1.664	0.848–3.265	.139
Gender identity
Man^b	0.709	0.272	2.033	1.192–3.467	.009
Transgender, non-binary, or another identity^b	0.269	0.546	1.309	0.449–3.816	.622
Sexual orientation	−0.549	0.302	0.578	0.319–1.045	.070
Age of first alcohol use	0.008	0.331	1.008	0.526–1.930	.981
Frequency of marijuana use (past 30 days)^c	−0.008	0.015	0.992	0.963–1.023	.618
Mixing marijuana and alcohol (past year)	0.971	0.298	2.640	1.471–4.739	.001
Prescription drug misuse (past year)	0.279	0.278	1.322	0.767–2.279	.315
Perceived risk of binge drinking	0.732	0.254	2.079	1.264–3.420	.004
Well-being^c	−0.024	0.013	0.976	0.951–1.002	.069

^aWhite is the reference group for race/ethnicity.

^bWoman is the reference group for gender identity.

^cFrequency of marijuana use and well-being are continuous measures.

Discussion

This study examines demographic, substance use, and wellness variables in relation to HID among college students and non-student young adults, thereby extending previous work that has examined HID among adolescents and adults. Additionally, demographic and policy differences (e.g., legality of various substances, access to substance use treatment) across states may impact rates of HID. Thus, this study also extends previous work on HID by utilizing data from one state that has legalized marijuana. Although most young adults in this sample (86.0%) did not engage in HID, a sizable minority engaged in HID. This is a concern given the potential adverse consequences associated with HID. The relationship between prescription drug misuse and HID among college students is unsettling given the increased risk of overdose, accidents and injuries when combining alcohol and prescription opioids or benzodiazepines.^32–34 By the time people experiment with stimulants, opioids, or sedatives, they often are using alcohol heavily,⁵² which may further exacerbate risks related to combined use.

Consistent with some studies of adolescents and young adults,^3,5 we found that both college and non-college men were more likely than women to engage in HID. This is in line with studies showing men are generally more likely than women to engage in risky alcohol use during college.^10,53,54 We did not find an increased risk of HID in women or individuals who reported a transgender, non-binary, or another gender identity. Previous research indicates that substance use is prevalent among young adults who are transgender or gender-diverse,⁵⁵ with most studies reporting rates that are comparable to or higher than those found among cisgender young adults.³⁵ The current findings bring attention to the importance of using non-binary measures of gender in substance use research.

Additionally, this study builds on previous work that found both binge drinking behaviors and beliefs were associated with HID among national samples of college students and non-college young adults.^3,5,7 Perceived binge drinking norms have been associated with one’s own drinking⁵⁶ and our results highlight how beliefs about binge drinking harm are similarly linked with HID. This study also shows that the relationship between SAM use and HID in adolescence²⁷ persists among college students and non-college young adults. Interestingly, our finding that frequency of marijuana use was not associated with HID contrasts with previous work.⁵ This may be due to differences between samples, as all participants in the present study were residents of a state that has legalized marijuana. On the other hand, Patrick et al.⁵ utilized data from the national Young Adult Daily Life Study.

Implications

Young adulthood is an important time to prevent HID because both short-term consequences and long-term problems can ensue. As one ages, heavy alcohol use is associated with changes in brain structure, and increased risk of cognitive impairment or dementia.⁵⁷ HID is associated with alcohol use disorder across age groups⁶ and with other consequences, such as missing work and relational problems.⁵ Missing work may damage one’s career, and relational problems may increase the odds of marriages or other important relationships ending, which may then exacerbate HID. College students may be especially vulnerable, as there is a higher prevalence of HID in the mid-20s among people who have earned a four-year college degree, compared to nongraduates.⁵ So far, prevention and harm reduction efforts have primarily focused on binge drinking.⁵ Correlates of HID among college students and non-college young adults revealed by this study offer a good starting point for both prevention and harm reduction initiatives.

Findings related to gender differences have implications for prevention and intervention efforts. Men were more likely to engage in HID, and addressing masculine norms linked with risky alcohol use may help to reduce HID. Other studies have found that conformity to norms such as being a “playboy”, taking risks, and endorsing homophobia and misogyny (e.g., ‘heterosexual presentation’), predicts alcohol use.^53,54,58 These norms could be combatted by showing how excessive drinking is linked with poor physical performance (e.g., in sports), as competitiveness is another norm to which many college men ascribe.⁵⁴ Programming could also be implemented that challenges these norms altogether. Moreover, college men who conform to these norms are less likely to seek help for their problems.^58,59 These men may be at greater risk of HID and may also be less likely to seek help for it or even consider it a problem.

Additionally, a recent meta-analysis found that binge drinking among college men was linked with sexual violence perpetration.⁶⁰ Therefore, college men who participate in HID may also be more likely to commit acts of sexual violence. The present study found that men were more likely to engage in HID, which has implications for sexual violence prevention on college campuses. Namely, HID should be addressed in such prevention programming. Spencer et al.⁶⁰ also found conformity to various masculine norms to be associated with sexual violence perpetration. Therefore, addressing these norms may reduce the prevalence of both HID and sexual violence perpetration among college men.

Our findings that HID may be related to other substance use variables can also inform interventions. Previous work has shown that college students generally overestimate other students’ drinking and estimates positively correlate to one’s own alcohol consumption.⁵⁶ Thus, prevention programming should show that HID is non-normative and encourage alternative choices and harm reduction approaches. Education may also be needed on the increased risk of overdose when alcohol is used in combination with other drugs.³² This education may be particularly important for students and other young adults engaged in HID. To reduce other high-risk substance use associated with HID, campuses should offer substance use treatment or referrals to services in the community. Service providers should assess for and treat polysubstance use, as it is associated with poorer prognosis and worse health outcomes than is single substance use.⁵²

Limitations and directions for future research

While this study adds to the knowledge base pertaining to HID among college students and non-college young adults, its limitations should also be considered. For one, this study utilized cross-sectional data and participants were self-selected, thus we cannot infer directionality of relationships between HID and other study variables. Additionally, there may be more nuances to HID by race among college students than we had the statistical power to detect. Using social media for recruitment can result in overrepresentation of White women.⁴¹ Future work should thus seek more racially diverse samples. Moreover, because systemic racism drives various health disparities,⁶¹ future work should also explicitly address how systemic racism may contribute to HID among college students of color. Similarly, while 71% of young adults ages 18–29 use Instagram and 70% use Facebook,⁴⁰ we were unable to reach young adults who did not use these services. Future work should consider other approaches for recruiting participants.

There are several other directions worth pursuing in future work. Future studies should address whether HID prevalence and frequency vary by campus-level covariates such as institution type (e.g., 2-year versus 4-year colleges), Greek membership, dorm residence, and distance education enrollment. Future studies should consider whether correlates of HID vary based on other campus characteristics, such as urban or rural campuses, or commuter or residential campuses. Uncovering such nuances may help to tailor prevention programing to specific campuses and student needs. Another direction to pursue is whether other types of substance use (e.g., tobacco use) not examined in this study are associated with HID, which would further augment prevention approaches. Future researchers might also consider strictly defining SAM, by delineating a time period within which use of both substances is considered simultaneous. Future work should also address whether there are lower rates of HID on campuses with treatment and/or recovery services for students. If the presence of these services is associated with reduced HID, that may prompt campuses lacking such services to adopt them. This would benefit students in need and could also shift norms related to heavy drinking on campus. Finally, future work should evaluate the effectiveness of interventions suggested in this paper.

Conclusion

A sizable minority of young adults (14.0%) in this statewide sample reported HID in the past 30 days. This is concerning given the potential adverse consequences, including injuries and higher rates of alcohol use disorder. Factors associated with HID among both college students and non-college young adults included being a man, mixing alcohol and marijuana in the past year, and perceived harm of binge drinking. Among college students in particular, past-year misuse of prescription drugs was also associated with HID. These findings can inform tailored prevention programming. Future work can also uncover other ways to prevent this high-risk drinking behavior.

Conflict of interest disclosure

The authors have no conflicts of interest to report. The authors confirm that the research presented in this article met the ethical guidelines, including adherence to the legal requirements, of the United States of America and received approval from the Institutional Review Board of Wayne State University.

Additional information

Funding

This work was supported by the Michigan Department of Health and Human Services under the Partnership for Success grant number SP020797.

Notes

* Throughout this paper we use the term marijuana. While the term ‘cannabis’ has become more popular, ‘marijuana’ (or marihuana’) is the term used more often in state and local policies in the USA.²⁸ We recognize the problematic history of the term marijuana. Historians, for example, have shown that the term has been used as an anti-immigrant pejorative.²⁹ However, other scholars suggest the term marijuana came about as an act of open resistance in Mexico, and that discontinuing use of the term could obscure its racist history in the USA.³⁰