Characteristics and Reasons for Use Associated with Solitary Alcohol and Marijuana Use among U.S. 12^th Grade Students, 2015-2021

Abstract

Background

Little is known regarding what sociodemographic characteristics and reasons for use are associated with adolescent solitary alcohol and marijuana use.

Methods

Data from 7,845 12^th grade students participating in the nationally-representative Monitoring the Future study from 2015-2021 were used to examine cross-sectional associations between sociodemographics, heavy drinking/marijuana use, reasons for use, and past 12-month solitary alcohol or marijuana use among past 12-month users. Historical trends and possible differences related to the COVID-19 pandemic also were examined.

Results

Solitary use prevalence increased from 2015 to 2021 with no evidence of significant COVID-19 deviations. In 2021, solitary alcohol use was reported by 32.1% (SE 3.01) and solitary marijuana use by 55.8% (4.72) of those reporting past 12-month use. Common and substance-specific sociodemographic risk factors were observed. Binge drinking was associated with solitary alcohol use; frequent marijuana use was associated with solitary marijuana use. Reasons for use related to coping with negative affect were associated with solitary use. Compulsive use reasons were more strongly associated with solitary alcohol than marijuana use. Drinking to have a good time with friends was negatively associated with solitary alcohol use but this association was not seen for solitary marijuana use.

Conclusions

The percentage of adolescents who use alcohol or marijuana when they were alone has increased among those who report using each substance. Associations between solitary use and (a) higher levels of consumption and (b) coping with negative affect highlight the importance of solitary use as a risk indicator.

1. Introduction

Adolescent solitary alcohol and marijuana use have been associated with negative outcomes extending into young adulthood (Mason et al., 2020). Solitary alcohol use is associated with high consumption and drinking problems (Skrzynski et al., 2021; Terry-McElrath et al., 2017) and development of problematic alcohol and drug use, violence, and physical health problems (Creswell et al., 2014; Tucker et al., 2006). Solitary marijuana use is associated with more frequent use and later disordered use (Creswell et al., 2015a). Adolescent solitary use has been increasing. Nationally representative data among U.S. 12^th grade students reporting past-year use showed increases in (a) solitary alcohol prevalence from 2014-2019 for males and 2000-2019 for females; and (b) solitary marijuana prevalence from 1992-2019 for both males and females (Terry-McElrath et al., 2020). New information on characteristics and reasons associated with adolescent solitary use may help clinicians and health providers identify subgroups at particularly high risk.

Research examining sociodemographic characteristics associated with adolescent solitary use has focused more on alcohol than marijuana (Mason et al., 2020; Skrzynski et al., 2020). Studies have found mixed results by sex (Creswell et al., 2015a, 2015b; Kask & Markina, 2014; Tucker et al., 2006, 2014), race/ethnicity (Creswell et al., 2015b; Dauber et al., 2009; Stewart & Power, 2003; Tucker et al., 2006), and socioeconomic status (Creswell et al., 2015b; Tucker et al., 2006, 2014). Solitary use appears to be more common among adolescents with single-parent families, lower grades, and less conformity (Tucker et al., 2006). Measures of religiosity are associated with lower overall adolescent alcohol and marijuana use (Ford & Hill, 2012; Miller et al., 2000).

Most studies examining sociodemographic differences have utilized single state/local samples (Creswell et al., 2015a, 2015b; Stewart & Power, 2003; Tucker et al, 2014) or included two states (Tucker et al., 2006). Studies using national or international data (Dauber et al., 2009; Kask & Markina, 2014) have focused on limited characteristics (sex and/or race/ethnicity). The degree to which results generalize to broader populations is unknown. Such samples also limit the ability to examine geographic and population density associations which are associated with overall alcohol and marijuana use (Johnston et al., 2020; Miech et al., 2021b).

Adolescent solitary alcohol use is associated with drinking in response to negative affect, negative reinforcement, social discomfort (Creswell et al., 2014; Skrzynski et al., 2021), depressive symptoms (Guo et al., 2021; Tomlinson & Brown, 2012), and coping motives (Guo et al., 2021). Coping motives have been associated with increased solitary marijuana use (Guo et al., 2021). One way of obtaining further information on adolescent solitary substance use is to ask individuals why they use a particular substance. Recognized motivations underlying substance use (social, enhancement, coping, and conformity motives; Cooper, 1994; Cooper et al, 1992; Cox & Klinger, 1988) can be differentiated by self-reported reasons for use (Patrick et al., 2011a). Self-reported reasons for alcohol and marijuana use among adolescents and young adults have been shown to be associated with binge drinking, alcohol use disorder, and marijuana use frequency in adulthood (Bray et al., 2021; Patrick et al., 2011a, Patrick et al., 2016, Terry-McElrath et al., 2017). If reasons for use differ significantly between those who do and do not report solitary use, knowledge of which reasons are particularly associated with solitary use may provide a better understanding of the mechanisms driving such behaviors.

Adolescent alcohol and marijuana use change across time (Miech et al, 2021b), as do reasons adolescents give for such use (Johnston & O’Malley, 1986; Patrick et al., 2019; Terry-McElrath et al, 2009). Associations between adolescent sociodemographic characteristics, reasons for use, and solitary use also may vary across time. Research investigating solitary use across years including 2020 and 2021 must address the possible impact of the COVID-19 pandemic but must do so while taking account of previously observed increasing historical trends in such behavior (Jager & Keyes, 2021). Prior research using data from 1976-2019 indicated that, among adolescent users, solitary alcohol and marijuana use increased from 2014/2015 to 2019 among both females and males (Terry-McElrath et al., 2020). Starting in March 2020, the pandemic brought changes that may have affected adolescent solitary substance use via social distancing and limitations on social gatherings and in-person school attendance. While overall adolescent alcohol and marijuana use declined notably across the pandemic compared to prior years (Miech et al., 2021c), some adolescents reported using substances to cope with social distancing and isolation (Patrick et al., 2022). The extent to which adolescent solitary use may have deviated with the onset of the pandemic—or if changes occurred in sociodemographic characteristics or use reason associations with solitary use—is unknown.

The current study expands existing research by investigating possible COVID-19 pandemic-related deviations in historical trends in solitary alcohol and marijuana use, and examining sociodemographic characteristics and reasons for use associated with solitary alcohol and marijuana use among 12^th-grade students during the years 2015-2021. Analyses followed four exploratory research aims: (1) estimate solitary alcohol and marijuana use prevalence across 2015-2021 and trends and/or deviations associated with the pandemic; (2) examine sociodemographic characteristics associated with solitary use; (3) examine associations between reasons for use and solitary use; and (4) model possible pandemic interactions with sociodemographics and reasons for use on solitary use.

2. Methods

2.1. Sample

Data were collected annually in the spring from 2015-2021 through the nationally representative Monitoring the Future (MTF) study; detailed methods are described elsewhere (Miech et al., 2021a, 2021b). In brief, 1976-2018 data collection involved in-school paper surveys (student response rates averaged 83%). Data collection in 2019 involved between-school randomization to either paper or electronic tablets (80% response rate; Miech et al., 2021a). In 2020, data collection involved in-school electronic tablets only and was halted on March 15 because of the COVID-19 pandemic (sample size approximately one-quarter of a typical data collection; 79% response rate).¹ Data collection in 2021 utilized web-based surveys completed remotely/at home or in school (69% response rate). A University of Michigan institutional review board approved the study. Solitary use was asked on 1 of 6 randomly distributed questionnaires. A total of 13,519 12^th grade students responded to the relevant form from 2015-2021. A total of 5,140 students (38.0% of 13,519) reported no past 12-month alcohol use and were thus ineligible because they were not asked reasons of use; of the remaining 8,379 cases, 1,110 (13.3%) had missing data on either overall alcohol use or solitary alcohol use, leaving 7,269 cases for alcohol analyses. A total of 8,194 (60.6% of 13,519) reported no past 12-month marijuana use and were ineligible; of the remaining 5,325 cases, 1,102 (20.7%) had missing data on either overall marijuana use or solitary marijuana use, leaving 4,223 cases for marijuana use analyses.

2.2. Measures

2.2.1. Substance use

Respondents were asked, “On how many occasions (if any) have you (1) had any alcoholic beverage to drink—more than just a few sips…; (2) used hashish…; (3) used marijuana…during the last 12 months?” If respondents reported any past 12-month use of alcohol or marijuana/hashish, they were asked, “When you used [alcohol, marijuana or hashish] during the last year, how often did you use it in each of the following situations? …when you were alone”. Dichotomous indicators of any solitary use versus none were coded separately for alcohol and marijuana/hashish (hereafter referred to as marijuana). Binge drinking indicated having 5+ drinks in a row over the past two weeks at least once (vs. not). Frequent marijuana use indicated use on 20+ occasions in the past 30 days (vs. not).

2.2.2. Sociodemographic covariates

Sex was coded as male, female, or other/prefer not to answer/missing (before 2021, response categories included male or female; in 2021, “other or prefer not to answer” was added²). Race/ethnicity was coded as Black, Hispanic, White, or other (including multiracial). Number of parents in the household was coded as two (vs. none/one). Parental education (used as a proxy for family of origin socioeconomic status; Patrick et al., 2012) indicated if at least one parent had a college degree (vs. not). Average high school grades were coded as B- or above (vs. lower). Truancy indicated any whole days of school skipped during the past four weeks (vs. none). Religious commitment combined service attendance and importance of religion and was coded as low, medium, or high (O’Malley & Johnston, 2013). Population density and region were coded based on the location of the participant’s school. Population density was coded as large metropolitan statistical area (MSA), other MSA, or non-MSA (Miech et al., 2021b). Region was coded according to the US Census as Northeast, Midwest, South, or West.

2.2.3. Reasons for use

If respondents reported any past 12-month use of alcohol or marijuana, for each substance, they were asked, “What have been the most important reasons for your [drinking alcoholic beverages/using marijuana]?” A series of reasons were listed; respondents were instructed to mark all that applied. For both substances, the following 13 reasons were listed: “to experiment–to see what it’s like”; “to relax or relieve tension”; “to feel good or get high”; “to seek deeper insights and understanding”; “to have a good time with my friends”; “to fit in with a group I like”; “to get away from my problems or troubles”; “because of boredom, nothing else to do”; “because of anger or frustration”; “to get through the day”; “to increase the effects of some other drug(s)”; “to decrease (offset) the effects of some other drug(s)”; “because I am “hooked”–I feel I [have to drink (alcohol)/have to have it (marijuana)]”. Two additional reasons were listed for alcohol: “to get to sleep”; “because it tastes good”.

MTF reasons for use measures were designed to capture a broad range of reasons for a large range of substances within a national survey (Johnston & O’Malley, 1986; Patrick et al., 2011a). The measures have been asked since 1976, prior to the development of the specific motive subscales identified by later research (social, enhancement, coping, and conformity motives; Cooper, 1994; Cox & Klinger, 1988). MTF reasons for use can be grouped meaningfully (as has been done in prior publications Johnston & O’Malley, 1986; Patrick et al., 2011a, 2011b, 2017, 2019; Terry-McElrath et al., 2009, 2017) as follows: social/recreational reasons (good time; experiment; feel good/get high; boredom; fit in); coping with negative affect reasons (relax; problems; anger/frustration); compulsive use reasons (get through the day; hooked); drug effect reasons (increase effects; decrease effects); and miscellaneous reasons (taste; sleep; insights/understanding).

2.3. Analysis

Analyses were conducted using SAS v.9.4 (SAS Institute Inc., Cary, NC) and accounted for the MTF complex sample design by using SURVEY commands. Multiple imputation (PROC MI; 10 multiply imputed data sets) was conducted using all sample data to address missing data; analyses were restricted to cases with no missing on relevant solitary use measures. PROC MIANALYZE was used to synthesize results (Berglund & Heeringa, 2014). Due to the large sample, discussion will be limited to results significant at p<.01.

All analyses were conducted separately for solitary alcohol use and solitary marijuana use. For descriptive statistics, SURVEYFREQ and SURVEYMEANS procedures were used. For research aim 1 (RA1), testing for significant deviations in solitary use trends involved three SURVEYLOGISTIC models regressing solitary use on: (1) a continuous 1 to 7 linear year measure (for 2015-2021), (2) continuous linear measure and a dichotomous 2021 indicator, and (3) centered linear year measure (−3 to 3) and quadratic term of the centered year measure. For RA2 (sociodemographic characteristic and solitary use associations), SURVEYLOGISTIC was used to regress solitary use on the relevant substance use covariate (binge drinking for the solitary alcohol use model, and frequent marijuana use for solitary marijuana use model) and sociodemographic covariates in bivariate models, and then in multivariable models simultaneously including substance use, sociodemographic covariates, and time (continuous measure). For RA3 (reasons for use and solitary use associations), SURVEYLOGISTIC was used to regress solitary use on each use reason bivariately, and then in multivariable models simultaneously including only one substance-specific reason per model, substance use, covariates, and time. For RA4, multivariable models for each solitary use measure were re-run testing time interactions with one covariate or reason for use per model (e.g., year x binge drinking; 2021 indicator x binge drinking).

3. Results

Supplement Table 1 provides descriptive statistics.

3.1. Solitary use prevalence and change (Aim 1)

Figures 1 and 2 provide solitary alcohol and marijuana use trends among those who use (for context, trends also are provided for overall 12-month use prevalence and solitary use among all students³). Among past 12-month alcohol users, 26.5% (SE 0.70) reported solitary alcohol use. Among past 12-month marijuana users, 48.0% (SE 1.08) reported solitary marijuana use. As shown in Table 1, solitary alcohol and marijuana use prevalence among users increased linearly across 2015-2021 with no significant quadratic associations or 2021 deviations. Solitary alcohol use increased from 25.1% (SE 1.56) in 2015 to 32.1% (3.01) in 2021; solitary marijuana use increased from 41.9% (2.41) to 55.8% (4.72).

Figure 1.

Trends in 12-month alcohol prevalence overall, and in solitary alcohol use among all students and past 12-month users, U.S. 12^th grade students, 2015-2021

Figure 2.

Trends in 12-month marijuana prevalence overall, and in solitary marijuana use among all students and past 12-month users, U.S. 12^th grade students, 2015-2021

Table 1.

Associations between time and any past 12-month solitary alcohol use or marijuana use among past 12-month users, U.S. 12^th grade students, 2015-2021

	Solitary alcohol use OR/AOR (95% CI) p	Solitary marijuana use OR/AOR (95% CI) p
Model 1
Year (continuous 1-7)	1.07 (1.03, 1.12) 0.002	1.09 (1.03, 1.15) 0.001
Model 2
Year (continuous 1-7)	1.06 (1.01, 1.11) 0.016	1.09 (1.03, 1.14) 0.001
Year 2021 indicator	1.08 (0.77, 1.50) 0.658	1.03 (0.69, 1.54) 0.887
Model 3
Year (continuous centered)	1.07 (1.03, 1.12) 0.002	1.09 (1.03, 1.15) 0.003
Quadratic term	1.01 (0.99, 1.03) 0.549	1.00 (0.98, 1.03) 0.842

Notes: n(unweighted)=7,269 for solitary alcohol use; 4,223 for solitary marijuana use. OR = bivariate odds ratio. AOR = adjusted odds ratio from models simultaneously including 2 terms. CI = confidence interval. Bold font indicates associations with p<.01.

3.1. Characteristic associations with solitary use (Aim 2)

Table 2 provides associations between binge drinking, sociodemographic covariates, time, and solitary alcohol use; Table 3 provides respective results for marijuana. Bivariate models indicated (1) reporting male sex or other/prefer not to answer/missing, lower grades, truancy, and low religious commitment were associated with greater odds of both solitary alcohol and marijuana use; (2) living in a non-MSA (vs. large MSA) was associated with greater odds of solitary alcohol use; and (3) identification as Black or Hispanic (vs. White), living with fewer than 2 parents, and not having a parent with a college degree were associated with greater odds of solitary marijuana use. In multivariable models, the following were associated with greater odds of solitary alcohol use: binge drinking, other sex/prefer not to answer/missing, lower grades, and non-MSA population density. The continuous year term remained significant at p<.01 until controlling for binge drinking. Multivariable associations with greater odds of solitary marijuana use included frequent marijuana use, being male, identification as Black, lower grades, low religious commitment, and year. Over 90% (92.1%) of those who frequently used marijuana sometimes used while alone.

Table 2.

Associations between covariates and any past 12-month solitary alcohol use among past 12-month alcohol users, U.S. 12^th grade students, 2015-2021

	%a (SE)	OR (95% CI) p	AOR (95% CI) p
Binge drinking
None	23.2 (0.7)	(ref)	(ref)
Any	39.8 (1.7)	2.19 (1.88, 2.55) <.001	2.10 (1.78, 2.47) <.001
Sex
Male	28.6 (1.1)	1.33 (1.14, 1.56) <.001	1.17 (1.00, 1.38) 0.055
Female	23.2 (1.0)	(ref)	(ref)
Other/prefer not to answer/missingb	33.9 (2.1)	1.70 (1.38, 2.10) <.001	1.37 (1.11, 1.70) 0.004
Race
Black	29.1 (2.1)	1.19 (0.95, 1.49) 0.134	1.25 (0.98, 1.59) 0.075
Hispanic	27.5 (1.6)	1.10 (0.91, 1.31) 0.323	1.14 (0.92, 1.40) 0.231
White	25.7 (0.9)	(ref)	(ref)
Other	26.8 (1.7)	1.06 (0.87, 1.29) 0.557	1.10 (0.89, 1.35) 0.386
Number of parents in household
0 or 1	28.7 (1.1)	(ref)	(ref)
Two	25.3 (0.9)	0.84 (0.73, 0.97) 0.015	0.95 (0.82, 1.11) 0.539
Parental education
No college degree	27.4 (1.0)	(ref)	(ref)
College degree	25.8 (1.0)	0.92 (0.80, 1.07) 0.285	1.02 (0.87, 1.20) 0.775
Average high school grades
C+ or below	36.8 (2.1)	(ref)	(ref)
B− or above	24.6 (0.8)	0.56 (0.46, 0.69) <.001	0.63 (0.51, 0.78) <.001
Truancy
None	24.7 (0.9)	(ref)	(ref)
Any	29.4 (1.1)	1.27 (1.09, 1.47) 0.002	1.09 (0.93, 1.28) 0.275
Religious commitment
Low	28.2 (1.0)	1.32 (1.08, 1.61) 0.006	1.26 (1.01, 1.55) 0.037
Medium	26.1 (1.5)	1.19 (0.93, 1.51) 0.160	1.12 (0.88, 1.44) 0.346
High	23.0 (1.5)	(ref)	(ref)
Population density
Large MSA	24.0 (1.0)	(ref)	(ref)
Other MSA	26.5 (1.1)	1.14 (0.98, 1.34) 0.097	1.19 (1.01, 1.39) 0.035
Non-MSA	30.8 (1.6)	1.41 (1.17, 1.70) <.001	1.41 (1.16, 1.73) 0.001
Region
Northeast	24.6 (1.7)	0.90 (0.71, 1.16) 0.419	0.91 (0.70, 1.18) 0.477
Midwest	24.4 (1.2)	0.89 (0.73, 1.10) 0.276	0.90 (0.73, 1.11) 0.323
South	28.7 (1.2)	1.12 (0.92, 1.36) 0.265	1.09 (0.90, 1.33) 0.365
West	26.5 (1.6)	(ref)	(ref)
Yearc		1.07 (1.03, 1.12) 0.002	1.06 (1.01, 1.11) 0.012

Notes: n(unweighted)=7,269. OR = bivariate odds ratio. AOR = adjusted odds ratio from models simultaneously including all covariates noted. CI = confidence interval. Bold font indicates associations with p<.01.

^aPercentage of each covariate subgroup reporting any past 12-month solitary alcohol use.

^bFrom 2015-2020, response options included female and male. In 2021, the response “other/prefer not to answer” was added. For all years, data were coded as a 3-category nominal measure of female, male, and missing. In 2021, those who responded “other/prefer not to answer” were combined with missing cases for the pupose of analysis.

^cContinuous values of 1-7 (for 2015-2021).

Table 3.

Associations between covariates and any past 12-month solitary marijuana use among past 12-month marijuana users, U.S. 12^th grade students, 2015-2021

	%a (SE)	OR (95% CI) p	AOR (95% CI) p
Frequent marijuana use
None	37.1 (1.0)	(ref)	(ref)
Any	92.1 (1.4)	19.69 (13.42, 28.89) <.001	17.08 (11.64, 25.05) <.001
Sex
Male	52.9 (1.4)	1.65 (1.41, 1.93) <.001	1.32 (1.11, 1.57) 0.002
Female	40.5 (1.5)	(ref)	(ref)
Other/prefer not to answer/missingb	59.6 (2.7)	2.17 (1.70, 2.75) <.001	1.30 (0.97, 1.74) 0.077
Race
Black	55.8 (2.7)	1.57 (1.23, 2.01) <.001	1.54 (1.15, 2.06) 0.004
Hispanic	52.5 (2.5)	1.37 (1.10, 1.72) 0.006	1.32 (1.02, 1.70) 0.034
White	44.5 (1.4)	(ref)	(ref)
Other	48.3 (2.7)	1.16 (0.93, 1.46) 0.193	1.17 (0.88, 1.56) 0.290
Number of parents in household
0 or 1	53.6 (1.6)	(ref)	(ref)
Two	44.0 (1.3)	0.68 (0.58, 0.80) <.001	0.90 (0.75, 1.08) 0.255
Average parental education
No college degree	51.8 (1.5)	(ref)	(ref)
College degree	44.4 (1.3)	0.74 (0.64, 0.87) <.001	0.89 (0.72, 1.09) 0.244
Average high school grades
C+ or below	63.3 (2.3)	(ref)	(ref)
B− or above	44.1 (1.2)	0.46 (0.37, 0.56) <.001	0.60 (0.47, 0.76) <.001
Truancy
None	44.8 (1.5)	(ref)	(ref)
Any	52.0 (1.7)	1.34 (1.12, 1.59) 0.001	1.15 (0.96, 1.38) 0.135
Religious commitment
Low	53.3 (1.4)	1.97 (1.52, 2.56) <.001	1.58 (1.21, 2.08) 0.001
Medium	43.3 (2.2)	1.32 (0.98, 1.79) 0.071	1.22 (0.91, 1.64) 0.190
High	36.7 (2.7)	(ref)	(ref)
Population density
Large MSA	49.1 (1.7)	(ref)	(ref)
Other MSA	46.7 (1.8)	0.91 (0.75, 1.10) 0.314	0.90 (0.74, 1.09) 0.278
Non-MSA	49.4 (1.8)	1.01 (0.83, 1.23) 0.918	1.00 (0.81, 1.23) 0.983
Region
Northeast	46.3 (2.0)	0.82 (0.62, 1.08) 0.161	1.00 (0.77, 1.31) 0.978
Midwest	44.5 (2.0)	0.76 (0.58, 1.01) 0.055	0.89 (0.68, 1.17) 0.404
South	49.0 (1.6)	0.91 (0.70, 1.19) 0.491	1.07 (0.82, 1.40) 0.604
West	51.3 (2.9)	(ref)	(ref)
Yearc		1.09 (1.03, 1.14) 0.001	1.10 (1.04, 1.16) <.001

Notes: n(unweighted)=4,223. OR = bivariate odds ratio. AOR = adjusted odds ratio from models simultaneously including all covariates noted. CI = confidence interval. Bold font indicates associations with p<.01.

^aPercentage of each covariate subgroup reporting any past 12-month solitary marijuana use.

^bFrom 2015-2020, response options included female and male. In 2021, the response “other/prefer not to answer” was added. For all years, data were coded as a 3-category nominal measure of female, male, and missing. In 2021, those who responded “other/prefer not to answer” were combined with missing cases for the purpose of analysis.

^cContinuous values of 1-7 (for 2015-2021).

3.2. Reason associations with solitary use (Aim 3)

3.2.1. Solitary alcohol use

Table 4 provides prevalence levels of reasons for alcohol use among all users and past 12-month users, as well as bivariate and multivariable associations.

Table 4.

Prevalence of reasons for alcohol use overall and associations between reasons and any past 12-month solitary alcohol use: U.S. 12^th grade students reporting past 12-month alcohol use, 2015-2021

	Prevalence of reason for alcohol use		Associations between alcohol use reasons and solitary alcohol use
	All drinkers	Solitary drinkers
	%a (SE)	%b (SE)	OR (95% CI) p	AOR (95% CI) p
Social/recreational reasons
To have a good time with my friends	69.4 (0.81)	67.2 (1.33)	0.87 (0.76, 0.99) 0.029	0.75 (0.65, 0.87) <.001
Experiment	58.7 (0.78)	50.1 (1.52)	0.62 (0.54, 0.71) <.001	0.69 (0.60, 0.80) <.001
To feel good or get high	32.1 (0.72)	41.2 (1.47)	1.73 (1.49, 2.01) <.001	1.52 (1.29, 1.78) <.001
Because of boredom, nothing else to do	18.9 (0.70)	29.9 (1.37)	2.43 (2.09, 2.83) <.001	2.14 (1.84, 2.50) <.001
To fit in with a group I like	6.6 (0.38)	6.7 (0.68)	1.03 (0.80, 1.32) 0.830	1.03 (0.80, 1.32) 0.818
Coping with negative affect reasons
To relax or relieve tension	41.2 (0.77)	63.0 (1.42)	3.40 (2.97, 3.90) <.001	3.18 (2.76, 3.66) <.001
To get away from my problems or troubles	23.1 (0.74)	39.9 (1.61)	3.23 (2.77, 3.77) <.001	2.96 (2.52, 3.47) <.001
Because of anger or frustration	10.3 (0.46)	21.2 (1.19)	3.97 (3.22, 4.90) <.001	3.46 (2.78, 4.32) <.001
Compulsive use reasons
To get through the day	4.8 (0.32)	11.1 (0.85)	4.95 (3.82, 6.40) <.001	4.20 (3.22, 5.47) <.001
Because I am “hooked”—I feel I have to drink	1.6 (0.19)	3.6 (0.52)	3.95 (2.50, 6.22) <.001	2.76 (1.78, 4.28) <.001
Drug effect reasons
To increase the effects of some other drug(s)	4.4 (0.32)	8.5 (0.75)	3.08 (2.27, 4.17) <.001	2.43 (1.76, 3.36) <.001
To decrease (offset) the effect of some other drug(s)	1.3 (0.16)	2.8 (0.43)	3.84 (2.29, 6.43) <.001	2.72 (1.61, 4.59) <.001
Miscellaneous reasons
Because it tastes good	30.0 (0.75)	39.8 (1.48)	1.84 (1.60, 2.12) <.001	1.86 (1.61, 2.15) <.001
To seek deeper insights and understanding	7.6 (0.40)	11.7 (0.87)	2.03 (1.64, 2.51) <.001	1.67 (1.33, 2.09) <.001
To get to sleep	7.1 (0.40)	14.8 (0.97)	3.88 (3.05, 4.92) <.001	3.38 (2.63, 4.35) <.001

Notes: Overall n(unweighted)=7,269. SE = standard error. OR = bivariate odds ratio. AOR = adjusted odds ratio from models simultaneously included the noted reason as well as all covariates shown in Table 2. CI = confidence interval. Bold font indicates associations with p<.01.

^aPercentage of all past 12-month alcohol users endorsing each particular reason.

^bPercentage of past 12-month alcohol users endorsing a particular reason who reported any past 12-month solitary alcohol use.

Among all alcohol users, the top five reasons for use were (in descending order): good time (69.4%), experiment (58.7%), relax (41.2%), feel good/get high (32.1%), and taste (30.0%). Results among solitary alcohol users were: good time (67.2%), relax (63.0%), experiment (50.1%), feel good/get high (41.2%), and problems (39.9%). In bivariate models, neither good time nor fit in were significantly associated with solitary alcohol use. Experiment was negatively associated with solitary alcohol use; all other reasons were positively associated. The reason with the strongest bivariate association was get through the day (OR 4.95). In multivariable models, both good time and experiment were associated with lower odds of solitary alcohol use, and fit in was not associated with solitary use; all other reasons were associated with greater odds of solitary alcohol use. Four reasons were associated with increased odds of solitary alcohol use by at least a factor of 3: get through the day (AOR 4.20), anger/frustration (AOR 3.46), sleep (AOR 3.38), and relax (AOR 3.18).

3.2.2. Solitary marijuana use

Table 5 presents associations between reasons for marijuana use and solitary marijuana use using the same format as Table 3. Among all past 12-month marijuana users, the top five reasons were relax (69.4%), feel good/get high (67.0%), good time (64.5%), experiment (54.7%), and problems (35.2%). Results among solitary marijuana users were: relax (84.2%), feel good/get high (76.8%), good time (67.9%), problems (48.3%), and bored (41.2%). Bivariate models showed that experiment was negatively associated with solitary marijuana use; all other reasons (excluding fit in) were positively associated with solitary use. The reason with the strongest bivariate association was hooked (OR 6.36). In multivariable models, experiment continued to be associated with lower odds of solitary marijuana use; no significant associations were observed for good time, fit in, hooked, increase or decrease effects. Remaining reasons were associated with greater odds of solitary marijuana use. Three reasons increased the multivariable odds of solitary marijuana use by at least a factor of 3: relax (AOR 3.44), get through the day (AOR 3.42), and anger/frustration (AOR 3.12).

Table 5.

Prevalence of reasons for marijuana use overall and associations between reasons and any past 12-month solitary marijuana use: U.S. 12^th grade students reporting past 12-month marijuana use, 2015-2021

	Prevalence of reason for marijuana use		Associations between marijuana use reasons and solitary marijuana use
	All users	Solitary users
	%a (SE)	%b (SE)	OR (95% CI) p	AOR (95% CI) p
Social/recreational reasons
To have a good time with my friends	64.5 (1.09)	67.9 (1.40)	1.33 (1.14, 1.56) <.001	1.20 (1.01, 1.43) 0.043
Experiment	54.7 (1.06)	39.4 (1.59)	0.29 (0.25, 0.35) <.001	0.40 (0.33, 0.48) <.001
To feel good or get high	67.0 (0.96)	76.8 (1.23)	2.39 (2.01, 2.84) <.001	2.08 (1.73, 2.50) <.001
Because of boredom, nothing else to do	29.1 (0.96)	41.2 (1.46)	3.18 (2.68, 3.77) <.001	2.40 (1.96, 2.94) <.001
To fit in with a group I like	7.0 (0.52)	6.3 (0.67)	0.80 (0.58, 1.08) 0.147	0.78 (0.50, 1.21) 0.268
Coping with negative affect reasons
To relax or relieve tension	69.4 (0.95)	84.2 (0.99)	4.24 (3.55, 5.06) <.001	3.44 (2.82, 4.20) <.001
To get away from my problems or troubles	35.2 (1.02)	48.3 (1.37)	3.11 (2.62, 3.70) <.001	2.34 (1.92, 2.86) <.001
Because of anger or frustration	23.5 (0.99)	37.1 (1.53)	4.77 (3.68, 6.19) <.001	3.12 (2.29, 4.25) <.001
Compulsive use reasons
To get through the day	19.6 (0.84)	32.4 (1.28)	5.64 (4.36, 7.30) <.001	3.42 (2.59, 4.52) <.001
Because I am “hooked”—I have to have it	4.3 (0.53)	7.6 (0.96)	6.36 (2.88, 14.05) <.001	1.90 (0.63, 5.71) 0.254
Drug effect reasons
To increase the effects of some other drug(s)	7.4 (0.55)	11.9 (0.98)	3.91 (2.69, 5.68) <.001	1.89 (1.15, 3.11) 0.012
To decrease (offset) the effect of some other drug(s)	4.2 (0.49)	7.2 (0.87)	5.54 (2.80, 10.98) <.001	2.32 (0.90, 5.96) 0.081
Miscellaneous reasons
To seek deeper insights and understanding	29.2 (0.87)	40.8 (1.47)	3.03 (2.51, 3.66) <.001	2.30 (1.84, 2.87) <.001

Notes: n(unweighted)=4,223. SE = standard error. OR = bivariate odds ratio. AOR = adjusted odds ratio from models simultaneously included the noted reason as well as all covariates shown in Table 3. CI = confidence interval. Bold font indicates associations with p<.01.

^aPercentage of all past 12-month marijuana users endorsing each particular reason.

^bPercentage of past 12-month marijuana users endorsing a particular reason who reported any past 12-month solitary marijuana use.

3.3. Time/pandemic interactions (Aim 4)

No significant time interactions (continuous measure or 2021 indicator) with either sociodemographic characteristics or reasons for use on the odds of solitary alcohol or marijuana use were observed.

4. Discussion

Using nationally representative data from 12^th-grade students, the current study found previously observed historical trends of increasing solitary alcohol and marijuana prevalence among users continued in the first year of the COVID-19 pandemic. Heavy use (binge drinking or frequent marijuana use) and lower grades were associated with greater odds of solitary use. Students in less densely populated areas were at greater risk for solitary alcohol use; males, Black students, and those with low religious commitment were at greater risk for solitary marijuana use. Reasons associated with coping with negative affect were associated with solitary use of both alcohol and marijuana; compulsive use reasons were more strongly associated with solitary alcohol than marijuana use.

Overall adolescent alcohol and marijuana use declined in summer 2020 (early months of the COVID-19 pandemic (Miech et al., 2021c), but the current study found no indications of COVID-19 pandemic-related change in solitary alcohol or marijuana use trends in 2021. Among all students, no significant change in solitary use prevalence was observed from 2015-2021; among users, solitary use trends from 2019-2021 showed increases similar to those from 2015-2019. Given the stable associations between negative affect/coping use reasons and solitary use, the consistently increasing prevalence of solitary use of both substances among users is concerning. Both before and during the pandemic, a growing proportion of adolescent alcohol and marijuana users may have been using alone to cope with negative emotions such as depression and stress. This finding is supported by other research finding pandemic-related isolation, stress, and economic hardship was associated with increased coping-related use of both alcohol and marijuana (Patrick et al., 2022).

The degree to which solitary use is indicative of high-risk use may differ between alcohol and marijuana. While the proportion of those using each substance alone was substantial (more than half of marijuana users and one-third of alcohol users in 2021), consistent with other studies (Creswell et al., 2015a), the current study found solitary use was reported more often for marijuana than alcohol. Qualitative research suggests young adults associate alcohol with larger social settings, while marijuana is associated with smaller social or solitary settings (Walton, 2020). The current study found using to have a good time with friends was negatively related to solitary alcohol use but unrelated to solitary marijuana use, suggesting adolescent solitary alcohol users—but not marijuana users—are less likely to use to enjoy social contexts. When more than half of those using a substance report doing so alone to some degree, solitary use could be said to be normative. If a behavior is normative among a particular group, it poses the question of whether or not it is still a meaningful risk factor. The current study found adolescent solitary use was associated with greater risks for binge drinking and frequent marijuana use. Associations between using alone and higher-risk use highlight the importance for clinicians and service providers to better understand factors associated with solitary use for both substances.

The current study found both similarities and differences in sociodemographic associations with adolescent solitary alcohol and marijuana use, suggesting possible differences in sociocultural contexts underlying these behaviors. Solitary alcohol use was associated with residing in a non-MSA (vs. large MSA). Lower population density has been associated with increased risk for adolescent alcohol (but not marijuana) use (Chan et al., 2016; Gale et al., 2012; Rural Health Information Hub, 2020), lower health and mental health service availability (Holzer et al., 2000), and fewer opportunities to have interactions with a large number of individuals in daily life (such interactions may provide a buffer against depression) (Stier et al., 2021). Regarding multivariable associations observed only for solitary marijuana use, risk factors included being male, Black, and low religious commitment. Males are more likely than females to become dependent on marijuana (Agrawal & Lynskey, 2007; Coffey et al., 2003; Cuttler et al., 2016); a higher likelihood of solitary use as adolescents may be a factor in such associations. Regarding racial/ethnic differences, Lee et al. (2021) found Black and Asian adolescents experienced higher peer/close friend disapproval of marijuana use; such disapproval might lead to a higher likelihood of solitary use. Finally, regarding religious commitment, adolescents with higher religiosity are not only less likely to initiate marijuana use, but also less likely to become persistent users after initiation (Ulmer et al., 2010). The protective nature of religiosity may also extend to high-risk forms of use including solitary use.

The current study supported and extended prior research on associations between reasons for use and solitary use. Earlier research using smaller samples (Creswell et al., 2014; Skrzynski et al., 2021; Tomlinson & Brown, 2012) identified associations between adolescent solitary alcohol use and depressive symptoms, negative affect, negative reinforcement, and social discomfort. One study using data from five U.S. regions identified associations between solitary alcohol and marijuana use and coping motives among American Indian adolescents (Guo et al., 2021). Using nationally-representative data focusing on both alcohol and marijuana use, the current study’s results emphasize solitary use as a risk factor for both substances. Experimentation was negatively related to both solitary alcohol and marijuana use; using to fit in was not associated with either behavior. These results are expected, because prior research has found experimentation to be negatively associated with higher-risk use, and using to fit in to be either not associated or negatively associated (Johnston & O’Malley, 1986; Patrick et al., 2011a, 2011b, 2019; Terry-McElrath et al., 2009, 2017). The current study found adolescent solitary alcohol and marijuana use were both associated with reasons suggesting using to cope with negative affect: relax, problems/troubles, anger/frustration, and get through the day. Prior studies have observed historical shifts in reasons for use have been trending towards more coping-related use (Patrick et al, 2019); this matches the current study’s observed increasing trends in solitary use. However, the current study also found solitary use was positively associated with other common reasons for use including feel good/get high and boredom. Taken together, these results suggest adolescent solitary users are likely to be established users who perceive substance use reduces negative affect or boredom and leads to feeling better or getting high. There is a continued need for interventions that address strategies and skills to prevent or reduce negative coping behaviors such as substance use (Conrod et al., 2010; Kaminer et al., 2002).

4.1. Limitations

The current findings are subject to limitations. Data were self-reported from a cross-sectional high school student sample. Causality between sociodemographic characteristics, reasons of use, and solitary use cannot be determined. Findings may not generalize to individuals who drop out of school prior to 12^th grade. The solitary use measures did not allow examination of solitary use frequency, reasons for solitary use specifically were unavailable, and there was a non-trivial degree of missing data for overall/solitary use. Survey response rates in 2021 were somewhat lower than previous years and may contribute unknown bias.

4.1. Conclusions

Solitary alcohol and marijuana use are associated with higher risk use as well as coping-related reasons for use. Understanding when and why adolescents use substances when they are alone is an important area for continued research.

Highlights.

• Solitary alcohol and marijuana use increased among 12^th grade students who use.

• Solitary use was associated with binge drinking and frequent marijuana use.

• Solitary use was associated with lower school grades.

• Solitary use was associated with using to cope with negative affect.