Estimated Prevalence of Substance Use Disorders Among US Adolescents and Emerging Adults by Substance Class, Severity, and Age, 2022

Abstract

Objective

Substance use disorders (SUDs) often develop during adolescence, forecasting myriad health problems across the lifespan. Implementing responsive clinical services requires information about the prevalence of SUDs by age, substance class, and severity. However, no reports have summarized those data using DSM-5 criteria.

Method

Using 2022 National Survey on Drug Use and Health (NSDUH) data from participants 12 to 25 years of age (n = 26,276), the prevalence and severity of DSM-5 SUDs was estimated across age cohorts (12-13, 14-15, 16-17, 18-20, and 21-25 years) via χ² tests of independence. The Cramer V (φ_c) was also calculated for each outcome to approximate the effect size between age group and substance use outcome.

Results

Although past-year rates for alcohol and cannabis use were higher overall as age cohort increased, the prevalence of disordered use and proportional distribution of SUD severity (mild, moderate, severe) did not differ across age cohorts among those who used alcohol (φ_c = 0.04) and cannabis (φ_c = 0.04) in the past year. Conversely, the prevalence and severity of SUDs generally varied across age groups among those who reported past-year use of less commonly used substances (heroin, methamphetamine).

Conclusion

Meeting criteria for an SUD was common among youth with past-year substance use. Allocation of developmentally appropriate prevention and treatment resources should account for the distribution of mild to severe SUDs across adolescence. The field would likely benefit from further study of these issues in diverse samples.

Plain language summary

This study analyzed data from the 2022 National Survey on Drug Use and Health. It explored the prevalence and severity of substance use disorders among adolescents and young adults aged 12 to 25. The prevalence and severity of substance use disorders among those who use alcohol and cannabis did not differ across age cohorts. It varied across age groups among those who reported past-year use of less commonly used substances (heroin, methamphetamine). The authors emphasized the importance of early screening to identify problematic substance use, since youth can develop mild to severe substance use disorders even in early adolescence.

Substance use disorders (SUDs) are a major public health concern in the United States. Drug overdose is a leading cause of accidental death nationally, with troubling patterns emerging among adolescents and emerging adults. For instance, the number of fatal unintentional overdoses among adolescents doubled between 2019 and 2020.¹ Moreover, early onset-substance use is associated with substantially heightened risk for multiple psychiatric disorders and medical problems into adulthood,² as well as numerous harms during adolescence such as poor academic performance, accidental injuries, neurocognitive impairment, and co-occurring mental health problems.3, 4, 5, 6 Beyond their proximal and distal effects on health, substance use and SUDs also have a negative impact on the economy, with annual cost estimates of $249 billion for alcohol use and $193 billion for drug use.⁷ To prevent or reduce these serious impacts, there is a need to understand the epidemiological landscape of specific SUDs—including by severity—within the critical period ranging from early adolescence through emerging adulthood (ie, 12-25 years of age).

SUDs often emerge in adolescence.⁸ Although not all youth who use substances will experience patterns of use or impairment that warrant an SUD diagnosis, the development of SUDs during the teen years is a robust predictor of adult addiction.⁹ SUDs are operationalized diagnostically based upon the number of observed criteria as outlined in the DSM-5.¹⁰ Disordered substance use is marked by the presence of physiological dependence, recurrent negative consequences, increased tolerance, and loss of control. SUDs also vary in severity, as determined by the number of symptoms present (ie, mild: 2 or 3 symptoms, moderate: 4 or 5 symptoms, severe: 6+ symptoms). SUD severity in adolescence predicts persistence of substance use problems in adulthood and connotes differing treatment needs.¹¹ Mild SUDs can often be treated on an outpatient basis—whether in primary care or in specialty behavioral health clinics—whereas more severe SUDs may warrant more intensive (and expensive) forms of intervention, such as acute inpatient, partial hospitalization, or residential treatment.^12,13 Identifying the scale and scope of specific SUDs by substance class and severity may guide policy decisions regarding resource allocation to improve the availability of prevention and clinical services programming for adolescents and emerging adults with SUDs in ways that are informed by and calibrated to population needs. Such data may also help to guide clinicians’ decision making about the services offered in their practices, as well as to inform education that they provide to patients and families around the potential harms associated with substance use.

Large epidemiological surveys in the United States have measured adolescent substance use and related factors for decades. For example, both the Youth Risk Behavior Survey (YRBS) and Monitoring the Future (MTF) study estimate the national prevalence of substance use in samples of US high school students; however, neither study includes specific information about DSM-5 SUDs.¹⁴ Meanwhile, the Treatment Episode Data Set (TEDS) provides information about SUDs among individuals 12 years of age and older, but only among those who were admitted to a hospital for substance use.¹⁵ Such data do not include several categories of youth, including those who do not require hospitalization and those for whom hospitalization does not occur.

More recently, the National Survey on Drug Use and Health (NSDUH) added DSM-5 criteria, permitting analysis of the relative distribution of mild, moderate, and severe SUDs by substance class for the first time.¹⁶ Yet neither official reports nor secondary analyses have computed this information by substance, age, and severity for youth or have made it available for widespread reference and use. Instead, reports have focused on the general prevalence of meeting criteria for SUD and have stratified the sample into 1 of 3 age groups: 12 to 17 years, 18 to 25 years, and 26 years and older.¹⁷ A limitation of collapsing across age groups is the resulting inability to discern differences in substance use or SUD severity that may vary in meaningful ways at a more granular level temporally, particularly when considering dynamic developmental periods such as adolescence. Therefore, the goal of the current study is to describe the national prevalence and severity of DSM-5 SUDs among US adolescents and emerging adults using age cohorts representing shorter time bands across adolescence (ages 12-13, 14-15, and 16-17 years) and that correspond to pertinent statutory milestones in early adulthood (ages 18-20 and 21-25 years). Using the NSDUH 2022 data, this study had 3 aims: (1) to describe the prevalence of any past-year substance use across age cohorts; (2) to describe the prevalence and severity of DSM-5 SUDs across age cohorts; and (3) to describe the prevalence and severity of SUDs across age cohorts among young people who reported past-year use of each substance.

Method

Data Source

Publicly available data were analyzed from the 2022 NSDUH survey.¹⁸ NSDUH is a nationally representative, cross-sectional survey of US residents that provides data on substance use, mental health, and related issues. The current analyses were limited to adolescents and emerging adults 12 to 25 years of age (unweighted n = 26,276). This age range reflects conceptualizations of adolescent neurodevelopment extending into the mid-20s,¹⁹ along with ample evidence supporting importance of this timeframe for the onset and escalation of substance use and SUDs.²⁰ Sex/gender, race, and ethnicity were not able to be included in the models to test for potential group differences because of published NSDUH suppression guidelines.²¹ Guidelines to suppress imprecise estimates were based on estimate, sample size, and effective (weighted) sample size. Further details regarding sample design and data collection are available in NSDUH technical reports.¹⁸ The Indiana University Institutional Review Board exempted the current secondary data analyses from review.

Measures

Age Groups

Participants under age 18 years were categorized into 1 of 3 age groups representing 2-year intervals (ie, 12-13, 14-15, and 16-17 years), and emerging adults were assigned to 1 of 2 age groups relative to age 21 years (ie, 18-20 and 21-25 years). This strategy was used to allow a more granular analysis of the prevalence of SUDs during early to mid adolescence, with broader categories of emerging adults formed around the legal age to use substances (ie, age 21 years) including tobacco, alcohol, and cannabis in some states.

Substance Use

Participants reported on their use of various substances over the past year, including alcohol, cannabis, cocaine, heroin, methamphetamine, and misuse of prescription medications (ie, without a prescription or not as prescribed) including pain relivers, stimulants, tranquilizers, and sedatives.²¹ The term “cannabis,” rather than “marijuana, is used throughout this paper” because of the complex xenophobic history of the term “marijuana” in regard to people of Hispanic or Latin American origin or descent.²²

Substance Use Disorders

SUDs were classified by the DSM-5 if 2 or more (of 11) criteria were met in the past 12-month period. SUDs were further classified by DSM-5–defined severity: mild (2 or 3 criteria), moderate (4 or 5 criteria), and severe (6 or more criteria).

Statistical Analyses

Data analysis occurred in the Statistical Package for the Social Sciences (SPSS), version 29. All estimates were weighted to 2020 US Census population estimates to account for selection probabilities and no-response patterns using the person-level weighting variable ANALWT2_C provided by Substance Abuse and Mental Health Services Administration (SAMHSA) in the NSDUH analytic data file.¹⁸ A series of χ² tests of independence were used to describe the following: (1) the prevalence of any past-year use substance use across age cohorts, (2) the prevalence and severity of SUDs across age cohorts, (3) and the prevalence and severity of SUDs across age cohorts among those who endorsed past-year use of each substance. The Cramer V (φ_c) was also calculated for each outcome to approximate the effect size between age groups and the substance use outcome. φ_c can range from 0 (no correlation between the 2 variables) to 1 (perfect correlation between 2 variables).²³ Cohen (1988) provides the following interpretations for φ_c: negligible (0.00 to <0.05), small (0.06 to 0.13), medium (0.14 to 0.21), and large (0.22 to 1.0).²⁴

Results

Participant Characteristics

Table 1 summarizes demographic characteristics of the analytic sample, including the weighted estimates of adolescents and emerging adults 12 to 25 years of age. Half of the sample was male assigned at birth. Emerging adults between 18 and 25 years of age represented the majority of the sample (54.4%), whereas adolescents between 12 and 17 years of age made up 45.5% of the sample. In this study, 5.1% self-reported as Asian, 13.4% Black, 23.5% Hispanic, 5.5% Multiracial, 1.8% native American, 0.4% Native Hawaiian, and 50.2% White.

Table 1.

Summary of Participant Characteristics

	Participants (n = 26,276)		Weighted Estimates (n = 60,512,061)
	n	%	n	%
Sex assigned at birth
Male	12,859	48.9	30,647,220	50.6
Female	13,417	51.1	29,864,842	49.4
Race and ethnicity
Asian	1,353	5.1	3,945,831	6.5
Black	3,532	13.4	8,174,404	13.5
Hispanic	6,185	23.5	14,908,915	24.6
Multiracial	1,436	5.5	1,803,494	3.0
Native American	469	1.8	410,429	0.7
Native Hawaiian	105	0.4	174,478	0.3
White	13,196	50.2	31,094,511	51.4
Age, y
12-13	3,926	14.9	8,474,136	14.0
14-15	4,284	16.3	9,067,219	15.0
16-17	3,759	14.3	8,184,412	13.5
18-20	4,978	18.9	12,728,627	21.0
21-25	9,329	35.5	22,057,668	36.5
Educationa
Less than high school degree	1,888	7.2	3,511,831	5.8
High school degree	4,997	19.0	11,691,899	19.3
Some college/associate’s degree	4,886	18.6	13,807,889	22.8
College graduate	2,536	9.7	5,774,675	9.5
Employment statusb
Full-time	6,217	23.7	14,595,819	24.1
Part-time	5,031	19.1	12,151,798	20.1
Unemployed	1,971	7.5	4,577,777	7.6
Other	7,019	26.7	16,192,194	26.8

Note:

^aData on education are not available for individuals 12 to 17 years of age.

^bData on employment are not available for individuals 12 to 14 years of age.

Prevalence of Past-Year Substance Use

Table 2 summarizes the prevalence estimates of past-year substance use in the weighted sample. Differences across age cohort and substance use in the past year were strongest for alcohol. The strength between age cohort and alcohol use among youth was large (φ_c = 0.56). Alcohol use was uncommon during early adolescence, with only 5.0% of 12- to 13-year-olds using alcohol in the past year. Alcohol use became more common as age cohort increased, with 14.6% of 14- to 15-year-olds using alcohol in the past year, followed by 31.4% of 16- to 17-year-olds, 51.3% of 18-to 20-year-olds, and 77.2% of 21- to 25-year-olds.

Table 2.

Summary of Past-Year Substance Use in the Weighted Sample

	n	%	χ2 (4)	φc
Alcohol			18,805,012.89	0.56
12-13	419,638	5.0
14-15	1,327,625	14.6
16-17	2,567,290	31.4
18-20	6,524,203	51.3
21-25	17,020,084	77.2
Cannabis			6,255,516.16	0.32
12-13	257,303	3.0
14-15	958,587	10.6
16-17	1,716,638	21.0
18-20	4,406,073	34.6
21-25	8,902,029	40.4
Heroin			35,542.19	0.02
12-13	0	0
14-15	0	0.0
16-17	2,831	0.03
18-20	10,096	0.1
21-25	36,495	0.2
Methamphetamine			107,684.51	0.04
12-13	2,929	0.03
14-15	4,560	0.1
16-17	9,040	0.1
18-20	33,711	0.3
21-25	126,956	0.6
Cocaine			1,233,991.08	0.14
12-13	18,612	0.2
14-15	4,401	0.05
16-17	17,126	0.2
18-20	231,196	1.8
21-25	1,078,859	4.9
Misuse of prescription medications
Pain reliever misuse			120,423.08	0.05
12-13	121,648	1.4
14-15	143,914	1.6
16-17	167,954	2.1
18-20	394,335	3.1
21-25	663,273	3.0
Stimulant misuse			553,387.71	0.10
12-13	20,602	0.2
14-15	63,536	0.7
16-17	134,658	1.6
18-20	345,924	2.7
21-25	887,191	4.0
Sedatives and tranquilizersa			361,288.39	0.08
12-13	11,613	0.1
14-15	47,539	0.5
16-17	70,670	0.9
18-20	237,643	1.9
21-25	569,470	2.6
Collapsed substances
Any opioidsb			126,421.24	0.05
12-13	121,648	1.4
14-15	143,914	1.6
16-17	167,954	2.1
18-20	396,022	3.1
21-25	676,950	3.1
Any central nervous system stimulantsc			1,409,020.08	0.15
12-13	41,447	0.5
14-15	69,764	0.8
16-17	155,187	1.9
18-20	507,335	4.0
21-25	1,727,959	7.8

Note: Number ranges in far-left-hand column are age ranges.

^a“Sedatives and tranquilizers” data are limited to individuals who misused sedatives and/or tranquilizers in the past year.

^b“Opioids” is inclusive of heroin and misuse of pain relievers.

^c“Any central nervous system stimulants” is inclusive of cocaine, methamphetamine, and misuse of prescription stimulants.

The strength between age cohort and cannabis use was also large (φ_c = 0.32). Cannabis use was also uncommon during early adolescence, with 3.0% of 12- to 13-year-olds using cannabis in the past year. Cannabis steadily became more common as age cohorts increased, with 10.6% of 14- to 15-year-olds using cannabis in the past year, followed by 21.0% of 16- to 17-year-olds, 34.6% of 18- to 20-year-olds, and 40.4% of 21- to 25-year-olds.

The strength between age cohort and cocaine use in the past year among youth was moderate (φ_c = 0.14). Although less common overall than alcohol and cannabis use, the prevalence of past- year cocaine use also increased as age cohorts increased, from <1% of 12- to 13-year-olds to nearly 5.0% of emerging adults 21 to 25 years of age.

The strength between age cohort and misuse of prescription medications (pain relivers, stimulants, sedatives, and tranquilizers) was small across medication class (range φ_c = 0.05-0.10), suggesting that prevalence of prescription medication misuse was similar across age cohorts within class. Misuse of pain relievers was most common, with 1.4% to 3.0% of adolescents and emerging adults having misused pain relievers in the past year, followed by 0.2% to 4.0% of youth misusing stimulants, and 0.1% to 2.6% of youth misusing sedative and tranquilizers in the past year.

Less than 1% of adolescents and emerging adults reported having used heroin or methamphetamine in the past year.

Prevalence and Severity of Substance Use Disorders

Table 3 summarizes the prevalence of meeting criteria for an SUD, as well as disorder severity in the weighted sample. Disordered alcohol and cannabis use were the most prevalent SUDs among adolescents and emerging adults. The strength between age cohort and severity of disordered alcohol use among youth was moderate (φ_c = 0.14), suggesting that disordered alcohol use varied across age cohorts. The largest proportion of disordered alcohol use was among 21- to 25-year-olds (18.8%), followed by 18- to 20-year-olds (10.8%), 16- to 17-year-olds (5.7%), 14- to 15-year-olds (2.4%), and 12- to 13-year-olds (1.2%). Across age cohorts, a larger proportion of youth met criteria for a mild alcohol use disorder (0.6%-10.6%) compared to moderate (0.3%-4.6%) and severe (0.2%-3.6%).

Table 3.

Summary of Past-Year Substance Use Disorder Severity in Weighted Sample

	Disorder severity						No disordered use		Test statistic
	Mild		Moderate		Severe
	n	%	n	%	n	%	n	%	χ2(12)	φc
Alcohol									3,324,024.68	0.14
12-13	55,062	0.6	22,629	0.3	15,941	0.2	8,380,504	98.9
14-15	143,409	1.6	48,722	0.5	21,025	0.2	8,854,062	97.6
16-17	293,984	3.6	130,516	1.6	38,388	0.5	7,721,524	94.3
18-20	812,939	6.4	298,200	2.3	260,571	2.0	11,356,917	89.2
21-25	2,344,409	10.6	1,010,456	4.6	794,594	3.6	17,908,209	81.2
Cannabis									2,186,395.19	0.11
12-13	38,712	0.5	44,902	0.5	21,557	0.3	8,368,966	98.8
14-15	203,314	2.2	134,179	1.5	145,761	1.6	8,583,965	94.7
16-17	355,733	4.3	179,753	2.2	204,483	2.5	7,444,444	91.0
18-20	698,251	5.5	639,421	5.0	568,521	4.5	10,822,434	85.0
21-25	1,663,822	7.5	1,226,738	5.6	933,646	4.2	18,233,461	82.7
Heroin									27,667.25	0.01
12-13	0	0.0	0	0.0	0	0.0	8,474,136	100.0
14-15	0	0.0	0	0.0	0	0.0	9,067,219	100.0
16-17	2,272	0.03	0	0.0	0	0.0	8,182,141	100.0
18-20	2,846	0.02	0	0.0	7,470	0.1	12,718,312	99.9
21-25	4,553	0.02	4,957	0.02	12,640	0.1	22,035,518	99.9
Methamphetamine									84,801.20	0.02
12-13	0	0.0	0	0.0	0	0.0	8,474,136	100.0
14-15	0	0.0	0	0.0	2,004	0.02	9,065,214	99.9
16-17	661	0.01	0	0.0	2,744	0.03	8,181,007	99.9
18-20	7,531	0.1	3,482	0.03	5,180	0.04	12,712,435	99.9
21-25	16,376	0.1	672	0.003	50,112	0.2	21,990,507	99.7
Cocaine									253,663.83	0.04
12-13	0	0.0	0	0.0	0	0.0	8,474,136	100
14-15	811	0.01	0	0.0	36	0.0004	9,066,371	99.9
16-17	3,118	0.04	0	0.0	0.0	0.0	8,181,294	99.9
18-20	32,964	0.3	2,608	0.02	31,258	0.2	12,661,796	99.5
21-25	90,454	0.4	57,565	0.3	63,690	0.3	21,845,958	99.0
Misuse of Prescription Medications
Pain Reliver Misuse									46,706.62	0.02
12-13	32,714	0.4	8,571	0.1	4,233	0.05	8,428,618	99.5
14-15	26,851	0.3	895	0.01	1,774	0.02	9,037,699	99.7
16-17	18,133	0.2	8,452	0.1	15,805	0.2	8,142,022	99.5
18-20	42,881	0.3	15,253	0.1	21,581	0.2	12,648,912	99.4
21-25	86,648	0.4	41,446	0.2	46,331	0.1	21,883,243	99.2
Stimulant Misuse									84,252.39	0.02
12-13	38,418	0.5	9,138	0.1	8,736	0.1	8,417,844	99.3
14-15	71,099	0.8	13,093	0.1	18,297	0.2	8,964,729	98.6
16-17	67,485	0.8	24,739	0.3	20,328	0.2	8,071,860	98.6
18-20	91,301	0.7	23,563	0.2	16,818	0.1	12,596,945	99.0
21-25	251,011	1.1	12,681	0.1	62,747	0.3	21,731,228	98.5
Sedatives and Tranquilizers Misuse									70,236.96	0.02
12-13	18,832	0.2	4,903	0.1	23,427	0.3	8,426,974	99.4
14-15	26,991	0.3	19,621	0.2	4,785	0.1	9,015,821	99.4
16-17	14,881	0.2	12,730	0.2	12,145	0.1	8,144,656	99.5
18-20	64,451	0.5	9,709	0.1	28,207	0.2	12,626,259	99.2
21-25	113,994	0.5	14,686	0.1	65,989	0.3	21,862,999	99.1
Collapsed Substances
Any Opioidsa									34,134.38	0.01
12-13	47,536	0.6	9,019	0.1	6,978	0.1	8,410,604	99.3
14-15	70,824	0.8	13,462	0.1	18,340	0.2	8,964,592	98.9
16-17	49,636	0.6	10,972	0.1	17,990	0.2	8,105,815	99.0
18-20	99,719	0.8	20,724	0.2	27,014	0.2	12,581,171	98.8
21-25	137,239	0.6	64,549	0.3	58,772	0.3	21,797,107	98.8
Any Central Nervous System Stimulantsb									165,719.81	0.03
12-13	38,418	0.5	9,138	0.1	8,736	0.1	8,417,844	99.3
14-15	71,910	0.8	13,093	0.1	20,302	0.2	8,961,914	98.8
16-17	70,734	0.9	23,450	0.3	21,616	0.3	8,068,612	98.6
18-20	123,800	1.0	29,653	0.2	52,935	0.4	12,522,238	98.4
21-25	317,374	1.4	57,317	0.3	162,761	0.7	21,520,216	97.6

Note: Number ranges in far-left-hand column are age ranges.

^a“Any opioids” is inclusive of heroin and misuse of pain relievers.

^b“Any central nervous system stimulants” is inclusive of cocaine, methamphetamine, and misuse of prescription stimulants.

The strength between age cohort and severity of disordered cannabis use among youth was small (φ_c = 0.11), suggesting that disordered cannabis use varied modestly across age cohorts. The largest proportion of disordered cannabis use was among 21- to 25-year-olds (17.3%), followed by 18- to 20-year-olds (15.0%), 16- to 17-year-olds (9.0%), 14- to 15-year-olds (5.3%), and 12-to 13-year-olds (1.2%). Across age cohorts, a larger proportion of youth met criteria for a mild cannabis use disorder (0.5%7.5%) compared to moderate (0.5%-5.6%) and severe (0.3%-4.3%).

Fewer than 2% of adolescents and emerging adults met criteria for disordered use across other substance classes. The strength between age cohort and severity of disordered use across all other substance use classes (heroin, methamphetamine, cocaine, and misuse of prescription medications) among youth was negligible (range φ_c = 0.01 to 0.03), suggesting that other disordered substance use was similar across age cohorts.

Prevalence and Severity of Substance Use Disorders Among Those Who Use Substances

Table 4 summarizes the prevalence of meeting criteria for a SUD, along with frequencies of disorder severity, in a subsample restricted only to those who reported use of substances in the corresponding substance class in the past year. The strength between age cohort and severity of disordered alcohol use among youth who used alcohol in the past year was negligible (φ_c = 0.04), suggesting that although the overall number of young people who used alcohol and had an alcohol use disorder increased by age cohort (Figure 1A), the proportion of disordered alcohol use severity was similar across age cohorts (Figure 1C). An estimated 16.1% to 24.4% of adolescents and emerging adults who used alcohol in the past year met criteria for an alcohol use disorder. Across age cohorts, a larger proportion of youth met criteria for a mild alcohol use disorder (11.5%-13.8%) compared to moderate (3.7%-5.9%) and severe (2.5%-4.7%).

Table 4.

Summary of Severity of Substance Use Disorders Among Those Who Used Substances in the Past Year in the Weighted Sample

	Disorder severity						No disordered use		Test statistic
	mild		Moderate		Severe
	n	%	n	%	n	%	n	%	χ2 (12)	φc
Alcohol (n = 27,858,838)									144,403.60	0.04
12-13	55,062	13.1	22,629	5.4	15,941	3.8	326,006	77.7
14-15	143,409	10.8	48,722	3.7	21,025	1.6	1,114,468	83.9
16-17	293,984	11.5	130,516	5.1	38,388	1.5	2,104,401	82.0
18-20	812,939	12.5	298,200	4.6	260,571	4.0	5,152,493	79.0
21-25	2,344,409	13.8	1,010,456	5.9	794,594	4.7	12,870,625	75.6
Cannabis (n = 16,240,632)									82,841.48	0.04
12-13	38,712	15.0	44,902	17.5	21,557	8.4	152,133	59.1
14-15	203,314	21.2	134,179	14.0	145,761	15.2	475,333	49.6
16-17	355,733	20.7	179,753	10.5	204,483	11.9	976,670	56.9
18-20	698,251	15.8	639,421	14.5	568,521	12.9	2,499,880	56.7
21-25	1,663,822	18.7	1,226,738	13.8	933,646	10.5	5,077,823	57.0
Heroin (n = 49,423)									18,696.16	0.44
12-13	0	0.0	0	0.0	0	0.0	0	0.0
14-15	0	0.0	0	0.0	0	0.0	0	0.0
16-17	2,272	80.3	0	0.0	0	0.0	559	19.7
18-20	1,762	17.5	0	0.0	7,171	71.0	1,163	11.5
21-25	4,553	12.5	4,957	13.6	8,447	23.1	18,539	50.8
Methamphetamine (n = 177,196)									21,759.96	0.20
12-13	0	0.0	0	0.0	0	0.0	2,929	100.0
14-15	0	0.0	0	0.0	2,004	44.0	2,555	56.0
16-17	661	7.3	0	0.0	2,744	30.4	5,635	62.3
18-20	7,531	22.3	3,482	10.3	5,180	15.4	17,519	52.0
21-25	16,376	12.9	672	0.5	48,745	38.4	61,163	48.2
Cocaine (n = 1,350,192)									41,586.72	0.10
12-13	0	0.0	0	0.0	0	0.0	18,612	100.0
14-15	811	18.4	0	0.0	36	0.8	3,553	80.8
16-17	3,118	18.2	0	0.0	0	0.0	14,008	81.8
18-20	32,964	14.3	2,608	1.1	31,258	13.5	164,365	71.1
21-25	90,454	8.4	57,565	5.3	63,690	5.9	867,150	80.4
Misuse of prescription medications
Pain reliever misuse (n = 1,491,123)									46,275.64	0.10
12-13	32,714	26.9	8,571	7.0	4,233	3.5	76,129	62.6
14-15	26,851	18.7	895	0.6	1,774	1.2	114,394	79.5
16-17	18,133	10.8	8,452	5.0	15,805	9.4	125,564	74.8
18-20	42,881	10.9	15,253	3.9	21,581	5.5	314,620	79.8
21-25	86,648	13.1	41,446	6.2	46,331	7.0	488,848	73.7
Stimulant misuse (n = 1,451,912)									43,452.39	0.10
12-13	5,120	24.9	794	3.9	5,689	27.6	8,999	43.7
14-15	7,857	12.4	2,809	4.4	5,888	9.3	46.982	73.9
16-17	6,302	4.7	3,227	2.4	10,992	8.2	114,139	84.8
18-20	16,044	4.6	6,702	1.9	16,818	4.9	306,360	88.6
21-25	62,841	7.1	10,090	1.1	59,457	6.7	754,802	85.1
Sedatives and tranquilizers (n = 936,935)									70,445.60	0.16
12-13	7,858	67.7	0	0.0	2,680	23.1	1,074	9.2
14-15	13,350	28.1	1,929	4.1	4,785	10.1	27,475	57.8
16-17	6,547	9.3	4,839	6.8	9,218	13.0	50,067	70.8
18-20	20,754	8.7	6,732	2.8	14,796	6.2	195,361	82.2
21-25	57,599	10.1	10,843	1.9	47,800	8.4	453,228	79.6
Collapsed substances
Any opioidsa (n = 1,506,485)									47,459.89	0.10
12-13	32,714	26.9	8,571	7.0	4,233	3.5	76,129	62.6
14-15	26,851	18.7	895	0.6	1,774	1.2	114,394	79.5
16-17	18,133	10.8	8,452	5.0	15,805	9.4	125,564	74.8
18-20	44,567	11.3	15,253	3.9	21,653	5.5	314,548	79.4
21-25	91,200	13.5	41,988	6.2	55,157	8.1	488,604	72.2
Any central nervous system stimulantsb (n = 2,501,692)									10,619.28	0.04
12-13	5,120	12.4	794	1.9	5,689	13.7	29,844	72.0
14-15	8,668	12.4	2,809	4.0	7,892	11.3	50,395	72.2
16-17	9,550	6.2	1,938	1.2	12,280	7.9	131,419	84.7
18-20	49,398	9.7	12,792	2.5	52,935	10.4	392,210	77.3
21-25	192,431	11.1	57,219	3.3	159,769	9.2	1,318,540	76.3

Note: Number ranges in far-left-hand column are age ranges.

^a“Any opioids” is inclusive of heroin and misuse of pain relievers.

^b“Any central nervous system stimulants” is inclusive of cocaine, methamphetamine, and misuse of prescription stimulants.

Figure 1.

Prevalence (Count, %) of Alcohol Use Disorder and Cannabis Use Disorder, by Age and Severity, Among US Young People Who Reported Past-Year Use of Alcohol and Cannabis

Note:(A) Estimated count (×100,000) of young people with and without an alcohol use disorder, by age cohort and severity, among people who used alcohol in the past year. (B) Estimated count (×100,000) of young people with and without a cannabis use disorder, by age cohort and severity, among people who used cannabis in the past year. (C) Proportional distribution of alcohol use disorder severity, by age cohort, among people who used alcohol in the past year. (D) Proportional distribution of cannabis use disorder severity, by age cohort, among people who used cannabis in the past year.

The strength between age cohort and severity of disordered cannabis use among youth who used cannabis in the past year was also negligible (φ_c = 0.04), suggesting that although the overall number of youth who used cannabis and met criteria for a cannabis use disorder increased by age cohort (Figure 1B), and was higher than for alcohol, the proportion of disordered cannabis use severity was similar across age cohorts (Figure 1D). Among those who used cannabis in the past year, 43.0% to 50.4% of youth met criteria for a cannabis use disorder.

Among adolescents and emerging adults who used heroin in the past year, the strength between age cohort and severity of disordered heroin use was large, suggesting that the proportional distribution of heroin use disorder severity varied across age cohorts. The largest proportion of disordered heroin use was among 18- to 20-year-olds who used heroin (88.5%), followed by 16- to 17-year-olds (80.3%), and 21- to 25-year-olds (49.2%). The majority of 18- to 25-year-olds who met criteria for a heroin use disorder were classified as severe, whereas most 16- to 17-year-olds met criteria for a mild heroin disorder.

The strength between age cohort and severity of disordered methamphetamine use among youth who used methamphetamine in the past year was moderate (φ_c = 0.20), suggesting that disordered methamphetamine use was became more common as age cohorts increased. An estimated 51.8% of 21- to 25-year-olds met criteria for a methamphetamine disorder, followed by 48% of 18- to 20-year-olds, 44% of 14- to 15-year-olds, and 37.7% of 16- to 17-year-olds. No 12- to 13-year-olds met criteria for a methamphetamine use disorder. Across age cohorts, a larger proportion of youth who met criteria for a methamphetamine disorder met were categorized as severe (15.4%-44.4%) compared to mild (0.0%-22.3%), and moderate (0.0%-10.3%).

The strength between age cohort and severity of disordered cocaine use among those who used cocaine in the past year was small (φ_c = 0.10), suggesting that disordered cocaine use varied modestly across age cohorts. The largest proportion of disordered cocaine use was among 18- to 20-year-olds (28.9%), followed by 21- to 25-year-olds (19.6%), 14- to 15-year-olds (19.2%), 16- to 17-year-olds (18.2%), and 14- to 17-year-olds (<1.0%). No 12- to 13-year-olds met criteria for a cocaine use disorder. A larger proportion of youth met criteria for a mild cocaine use disorder (8.4%-18.4%) compared moderate (0.0%-5.3%) and severe (0%-13.5%).

Misuse of Prescription Medications

The prevalence of sedatives and tranquilizer disorder was the highest prescription medication disorder among adolescents and emerging adults who used the corresponding substance in the past year. The strength between age cohort and severity of disordered sedatives and tranquilizer misuse was moderate (φ_c = 0.16), suggesting that disordered sedative and tranquilizer misuse was more common among younger cohorts. An estimated 90.8% of 12- to 13-year-olds who misused sedatives and tranquilizers in the past year met criteria for disordered use, followed by 42.2% of 14- to 15-year-olds, 29.2% of 16- to 17-year-olds, 17.8% of 18- to 20-year-olds, and 20.4% of 21- to 25-year-olds. Across age cohorts, a larger proportion of youth met criteria for mild (8.7%- 67.7%) or severe (6.2%-23.1%) sedative use disorder compared to moderate (0.0%-6.8%).

The strength between age cohort and opioid use disorder among those who misused pain relievers in the past year was small (φ_c =0 .10), suggesting that disordered use of pain relievers varied modestly across age cohorts. The largest proportion of disordered opioid use disorder among those who misused pain relievers was among 12- to 13-year-olds (37.4%), followed by 21- to 25-year-olds (26.3%), 16- to 17-year-olds (25.2%), 14- to 15-year-olds (20.5%), and 18- to 20-year-olds (20.2%). Across age cohorts, a larger proportion of youth met criteria for a mild (13.1%-26.9%) opioid use disorder compared to moderate (0.6%-7.0%) and severe (1.2%-9.4%).

The strength between age cohort and stimulant use disorder among those who misused prescription stimulants in the past year was also small (φ_c = 0.10), suggesting that disordered use of prescription stimulants also varied across age cohorts. The largest proportion of disordered stimulant use was among 12- to 13-year-olds (26.3%), followed by 14- to 15-year-olds (26.1%), 16- to 17-year-olds (15.2%), 21- to 25-year-olds (14.9%), and 18- to 20-year-olds (11.4%). Across age cohorts, a larger proportion of youth met criteria for mild (10.9%-26.9%) and severe (6.7%-27.6%) stimulant use disorder compared to moderate severity (1.1%-4.4%).

Discussion

As leaders across health care, government, and other sectors consider how best to address the ongoing addiction crisis, it is critical to have an up-to-date understanding of the population rates of SUDs—including by severity—for adolescents and emerging adults. Such data add nuance to more commonly reported prevalence data (eg, “ever used,” “used in the past year”) and provide information regarding an important consequence of substance use. More recently, NSDUH became one of the first large epidemiologic surveillance studies to assess for SUDs and the severity of disordered use; however, official reports have collapsed adolescents and emerging adults across broad age groups (eg, 12-17 and 18-25 years).¹⁷ The current study offers the first computation of DSM-5 SUD prevalence estimates among US adolescents and emerging adults using more granular age cohorts (ages 12-13, 14-15, 16-17, 18-20, and 21-25 years) to characterize patterns of SUD severity during this dynamic developmental period.

Prevalence and Severity of Substance Use Disorders

Consistent with other samples,^20,25,26 the majority of adolescents and emerging adults did not meet criteria for SUD, including most youth who reported past-year use of most substances. Generally, the prevalence of SUDs did not vary across age with the exception of alcohol and cannabis use disorders. Meeting criteria for an alcohol use disorder became more common with increasing cohort age, with highest rates observed among 21- to 25-year-olds, but a sizeable proportion of young people had developed an alcohol use disorder before reaching the legal US drinking age of 21 years. An estimated 1 in 20 adolescents 16 to 17 years of age met criteria for an alcohol use disorder (36.5% of which were classified as moderate or severe), and 1 in 10 emerging adults 18 to 20 years of age met criteria for an alcohol use disorder (40.7% of which were classified as moderate or severe). Although overall alcohol use among young people has trended downward in recent years, these findings highlight that although most youth who use alcohol do not report corresponding alcohol use disorder symptoms, among those who do, it is common for them to experience enough symptoms to warrant classification as a moderate to severe disorder.

Meeting criteria for a cannabis use disorder also became more prevalent as age group increased. Findings revealed that nearly 1 in 20 adolescents 14 to 15 years of age met criteria for a cannabis use disorder (57.9% of which were classified as moderate or severe). The prevalence of a cannabis use disorder also doubled in the next age group, with 1 in 10 adolescents 16 to 17 years of age meeting criteria for a cannabis use disorder (51.9% of which were classified as moderate or severe). These results show that disordered cannabis use is common, even among adolescents who are up to 7 years younger than the legal age for purchasing cannabis in states where cannabis is legalized, and that half or more of high school–aged youth with a cannabis disorder experience symptoms in the moderate to severe range. It is possible that this pattern reflects the increased acceptance of cannabis in society, reduced perception among youth that cannabis is harmful,²⁷ and increased access to high-potency cannabis and cannabinoid products sold through storefronts.²⁸ This information is important not only for health care professionals, who are often tasked with identifying and treating SUDs, but also parents, caregivers, and young people themselves who may underestimate the potential harms associated with alcohol and cannabis use during adolescence and emerging adulthood.²⁹

Prevalence and Severity of Substance Use Disorders Among Those Who Use Substances

The current study is also among the first to examine the population prevalence and severity of substance use disorders specifically among adolescents and emerging adults who report that they use substances. Although it is valuable to know the overall prevalence of SUDs, inclusion of people who do not use substances in those estimates may mask important information about patterns of SUDs among those who do. Results generally suggested that the prevalence of SUDs and proportional distribution of SUD severity among youth who use substances varied across age groups, with 2 exceptions, namely, alcohol use disorder, cannabis use disorder.

Although heroin, cocaine, and methamphetamine use in the past year were generally rare across age groups (<1%), meeting criteria for SUD was common among adolescent and emerging adults who engaged in use over the past year. This finding underscores the high addiction potential of these substances. This pattern may also reflect a subset of youth who have a greater predisposition to development of SUDs (eg, fewer protective factors, greater risk factors) or who have pre-existing/co-occurring patterns of substance use disorders for other substance classes (eg, alcohol, cannabis). Regardless, this finding emphasizes the importance of efforts aimed at preventing the onset and escalation of the use of these substances to avoid harms associated with corresponding SUDs. Such work is especially important considering the ubiquity of substances contaminated with synthetic opioids such as fentanyl, which is a major cause of substance use–related mortality in young people.¹ The pattern for SUDs across age groups were generally reversed among disordered use of prescription medications (pain reliver misuse, stimulant misuse, sedative and tranquilizer misuse), with the prevalence of SUDs being higher among younger age groups compared to older age groups. It is possible that this reversed pattern reflects the availability of prescription medications in the home or other social environments (eg, school) where similar aged peers may have greater access to prescription medications. Alternatively, it may also reflect an escalation of stimulant and opioid misuse across age cohorts. It is possible that adolescents build a tolerance towards prescription forms of stimulants and opioids in early adolescence, and transition to nonprescribed opioids and stimulants as they get older. This finding highlights disparities in substance use disorders across age groups that are normally overlooked when collapsing all adolescent age groups together.

These results highlight significant differences in SUD prevalence between adolescent and emerging adult age cohorts that are frequently aggregated in surveillance data. As efforts continue to build a more robust clinical infrastructure for youth with SUDs, current findings support expanding the accessibility and availability of developmentally appropriate services designed to address the full spectrum of SUD severity, especially for alcohol and cannabis, in adolescents.^30,31 For instance, to meet the needs of youth with mild to moderate SUDs (ie, the majority of adolescents with SUDs), it may be prudent to scale-up implementation of brief interventions, integrated primary and behavioral health care, expanded outpatient services, services delivered in nonclinical settings (eg, schools, juvenile justice), and promising digital therapeutics that may be self-directed or delivered in tandem with clinical support.³²

Although a greater number of young people have less severe SUDs, there are still tens to hundreds of thousands of youth with more severe, complex, or persistent SUDs. For many of these youth, there is often a need for higher levels of care, such as partial hospitalization or residential treatment programs to help them achieve recovery and remission. It is critical that such programs incorporate empirically supported interventions, including evidence-based psychotherapies and medications when appropriate and available, to increase the likelihood of positive outcomes.³³

Ongoing research to establish new treatments for SUDs and to test existing medications rigorously in diverse samples of adolescents with varying degrees of SUD severity is vital to better inform which treatments should be offered to which patients. To date, only buprenorphine/naloxone has received US Food and Drug Administration (FDA) indication for the treatment of non-tobacco SUD in adolescents (16 years and older). No medication has received FDA indication for treatment of cannabis or alcohol use disorders in adolescents, although some medications with FDA approval in adults are used off label in adolescents.³⁴

Consistent with recommendations from the American Academy of Pediatrics Committee on Substance Abuse, these results also emphasize the importance of early screening to identify problematic substance use, because youth develop mild to severe SUDs even in early adolescence, with overall numbers increasing across the teen years into early adulthood. These results also highlight the need for comprehensive diagnostic assessment for all youth who have used alcohol or cannabis in the past year, given the high prevalence of disordered use among youth who used alcohol and cannabis in the past year regardless of age.

First, this study did not include race and/or ethnicity in the analyses as a result of NSDUH suppression rules, which are in place both to better ensure reliable estimates and to protect participant confidentiality.¹⁸ When multiple cross-sections of a dataset are computed simultaneously (eg, age range × substance × race/ethnicity), the subset of individuals who can fall into a given cell becomes increasingly small, especially when an outcome is comparatively rare within a population (eg, methamphetamine use disorder). Thus, even when we attempted to collapse age and race/ethnicity categories into fewer groups, many of the cross-sectional cells for racial or ethnic categories other than non-Hispanic White did not meet data suppression standards. Calculation of SUD class and severity estimates by race/ethnicity would likely require recruiting greater numbers of people of color in future epidemiological studies of SUDs.^35,36 The absence of these data limits our ability to evaluate whether the need for particular prevention and treatment services might vary based on these demographic characteristics, which are often correlated with geography and, therefore, service availability and public policy. Similarly, cell sizes across age categories and sex/gender were too small to interpret across several key outcomes—particularly for specific SUD prevalence and severity—and are therefore not presented in the current study. It should also be noted that the prevalence of SUDs observed here may be underestimates, especially among emerging adults, because NSDUH participant eligibility criteria exclude active-duty military, individuals living in institutions, and unhoused individuals not living in shelters, which are populations known to have higher rates and severity of substance use and SUDs.²¹ Responses may also be contingent on recall and social biases such that participants may have minimized and underreported their use, symptoms, or related impairments; however, SAMHSA provides evidence from a number of studies, including cross-validation of biospecimen data with self-report, that appear to support the overall validity of the NSDUH data collection approach.

DSM-5 criteria should be incorporated into other large-scale epidemiological studies that routinely or exclusively involve adolescents in surveys (eg, Monitoring the Future [MTF], Youth Risk Behavior Surveillance System (YRBS] survey) to provide additional data on the prevalence of SUDs and to facilitate investigation of SUD severity in relation to other risk factors, protective factors, and outcomes. As the United States makes major investments in SUD prevention and treatment across the lifespan, including adolescence, it will be useful to monitor longitudinal shifts in patterns of SUDs, along with other related factors such as access to and engagement in clinical services, as a function of age and SUD severity.

CRediT authorship contribution statement

Zachary W. Adams: Writing – review & editing, Writing – original draft, Supervision, Methodology, Funding acquisition. Trey V. Dellucci: Writing – review & editing, Writing – original draft, Software, Formal analysis. Jon Agley: Writing – review & editing, Software, Methodology, Funding acquisition, Conceptualization. Kristina Bixler: Writing – original draft, Formal analysis. Maggie Sullivan: Writing – original draft. Jesse D. Hinckley: Writing – review & editing, Writing – original draft, Funding acquisition, Conceptualization. Leslie A. Hulvershorn: Writing – review & editing, Funding acquisition, Conceptualization.