Customize your JAMA Network experience by selecting one or more topics from the list below.
Use of alcohol, marijuana, and other drugs at an early age (eg, before age 18 years) increases the likelihood of drug use disorder,1,2 which may disrupt brain development.3 A recent US report4 showed decreased incidence of some drugs, such as marijuana and tobacco, among people aged 12 to 17 years, while the contrary is true among those aged 18 to 25 years. This suggests an increase in the mean age at initiation of some drugs, although we have found no confirmation of this in recent literature. In this study, we examine recent trends in the mean age at initiation for 18 internationally regulated drugs (including alcohol and tobacco), focusing on the critical neurodevelopmental period (ages 12-21 years), using data from the National Survey on Drug Use and Health (NSDUH).
Each year from 2004 to 2017, staff from the NSDUH drew a random, nationally representative sample of the US population ages 12 years and older through multistage sampling to assess drug-use and drug-associated behaviors. The NSDUH staff used audio computer-assisted self-interviews to collect data on alcohol and drug use, including age at first use, after obtaining informed consent/assent. We analyzed publicly available, deidentified NSDUH data that included 338 268 individuals aged 12 to 21 years who completed self-interviews on their initiation of alcohol, tobacco, and other drugs. Because all data came from deidentified, publicly available files, the Washington State University institutional review board ruled that the plan to analyze these data was not human subjects research.
In this study, we identified individuals who initiated drug use within 12 months prior to assessment. We analyzed 18 drugs, including alcohol and tobacco products (Figure), to estimate the year-by-year drug-specific mean age at initiation among individuals aged 12 to 21 years. Analysis weights account for the complex survey design of the data. We conducted joinpoint regression using log-linear model to estimate the slope of analysis-weighted, drug-specific mean ages between 2004 and 2017. Finally, meta-analysis with a DerSimonian and Laird random effects estimator was used to summarize year-by-year mean age estimates for initiation of each drug.5 All analyses were conducted using Stata SE15 (StataCorp), with the statistical significance level at .05. Data analysis occurred from April 2019 to June 2019.
Data are from the National Survey on Drug Use and Health, 2004-2017 (N = 84 317 adolescents and young adults). Individuals who initiated drug use are individuals aged 12 to 21 years with first-time drug use within 12 months prior to the survey. PCP indicates phenylcyclohexyl piperidine.
aBecause of survey methodological changes, mean ages from 2015 to 2017 are not included.
bP < .001.
cDifferences are significant (ecstasy, β = 0.003; P = .03; hallucinogens, β = 0.004; P = .004; heroin, β = 0.005; P = .02; inhalants; β = 0.004; P = .02; smokeless tobacco, β = 0.001; P = .002; stimulants, β = 0.005; P = .007; lysergic acid diethylamide [LSD], β = −0.014; P = .04).
Between 2004 and 2017, 84 317 adolescents and young adults initiated use of any drug. Estimated year-by-year mean ages, slopes, and annual percentage changes for each drug are visible in the Figure. Of the 18 drugs analyzed, we observed an increase in the mean age at initiation for 12 drugs since 2004 (including cigar use, β = 0.003; P < .001; cocaine, β = 0.005; P < .001; ecstasy, β = 0.003; P = .03; hallucinogens, β = 0.004; P = .004; heroin, β = 0.005; P = .02; inhalants; β = 0.004; P = .02; marijuana, β = 0.002; P < .001; smokeless tobacco, β = 0.001; P = .002; stimulants, β = 0.005; P = .007; and tobacco cigarettes, β = 0.006; P < .001). The mean age at initiation increased from 2004 to 2014 for alcohol (β = 0.005; P < .001), but there was no increase in this mean age at initiation after 2014 (β = −0.004; P = .06). The mean age at lysergic acid diethylamide (LSD) initiation increased from 2004 to 2009 (β = 0.013; P = .03) and declined significantly after 2009 (β = −0.014; P = .04). No statistical difference in the mean ages at initiation for crack cocaine, methamphetamines, opioids, phenylcyclohexyl piperidine (PCP), sedatives, and tranquilizers was noted.
The overall summary mean age at initiation and the mean percentage change over the years of data for each drug are presented in the Table. Adolescents and emerging adults began to use inhalants during early adolescence to midadolescence (mean age, 15.4 [95% CI, 15.2-15.6] years), the earliest mean age compared with those of other drugs. The mean ages at initiation for cocaine and crack cocaine were 18.0 (95% CI, 17.7-18.2) years and 18.0 (95% CI, 17.8-18.2) years, respectively, indicative of initiation in late adolescence or the beginning of young adulthood. There were no significant changes in the trends in the mean ages at initiation of crack cocaine, methamphetamines, opioids, PCP, sedatives, and tranquilizers.
Our findings show that the mean age at initiation is increasing for most drugs under study. For example, the mean age at alcohol initiation was approximately 16 years in 2004. By 2017, the mean age had risen to about 17 years—a full year after the minimum age for an unrestricted driver’s license in the United States. For alcohol and LSD, the stable estimates in recent years suggest that trends toward later ages at initiation may have already ended. Study limitations include self-reported data and recall bias. The NSDUH survey does not yet include vaping and e-cigarette use.
Drug use, including marijuana, appears to be especially harmful to developing brains. This study provides evidence that, since 2004, a decreasing proportion of adolescents initiate drugs at 15 years and younger, the age group typically classified as early onset.6
Accepted for Publication: August 18, 2019.
Corresponding Author: Karl C. Alcover, PhD, Elson Floyd College of Human Medicine, Department of Medical Education and Clinical Sciences, Washington State University, 665 N Riverpoint Blvd, Spokane, WA 99202 (firstname.lastname@example.org).
Published Online: March 2, 2020. doi:10.1001/jamapediatrics.2019.6235
Correction: This article was corrected on April 6, 2020, to replace an “Elson Floyd College of Human Medicine” with “Elson S. Floyd College of Medicine”; add a degree, MHI, after the name of author Christopher L. Thompson, MD, MHI; specify that Dr Thompson’s affiliation is called “Peds AI” and not “Peds AI Clinic”; and add 2 missing notations denoting significance to the graphs for cigar and cocaine in the Figure.
Author Contributions: Dr Alcover had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Alcover.
Acquisition, analysis, or interpretation of data: Both authors.
Drafting of the manuscript: Both authors.
Critical revision of the manuscript for important intellectual content: Both authors.
Statistical analysis: Alcover.
Administrative, technical, or material support: Alcover.
Supervision: Both authors.
Conflict of Interest Disclosures: None reported.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Alcover KC, Thompson CL. Patterns of Mean Age at Drug Use Initiation Among Adolescents and Emerging Adults, 2004-2017. JAMA Pediatr. 2020;174(7):725–727. doi:10.1001/jamapediatrics.2019.6235
Coronavirus Resource Center