Presented at the Research Society on Alcohol Annual Meeting, June 21, 2009, San Diego, CA.
The Significance of Marijuana Use Among Alcohol-using Adolescent Emergency Department Patients
Version of Record online: 4 JAN 2010
© 2009 by the Society for Academic Emergency Medicine
Academic Emergency Medicine
Volume 17, Issue 1, pages 63–71, January 2010
How to Cite
Chun, T. H., Spirito, A., Hernández, L., Fairlie, A. M., Sindelar-Manning, H., Eaton, C. A. and Lewander, W. J. (2010), The Significance of Marijuana Use Among Alcohol-using Adolescent Emergency Department Patients. Academic Emergency Medicine, 17: 63–71. doi: 10.1111/j.1553-2712.2009.00615.x
Supported by grants from the National Institute of Alcohol Abuse & Alcoholism, R01 AA013385, $2,146,600, 8/02–7/07, and the National Institute of Alcohol Abuse & Alcoholism, K23 AA014934, $953,800, 3/05–2/10.
- Issue online: 4 JAN 2010
- Version of Record online: 4 JAN 2010
- Received May 12, 2009; revision received July 7, 2009; accepted July 16, 2009.
- emergency department;
Objectives: The objective was to determine if adolescents presenting to a pediatric emergency department (PED) for an alcohol-related event requiring medical care differ in terms of substance use, behavioral and mental health problems, peer relationships, and parental monitoring based on their history of marijuana use.
Methods: This was a cross-sectional comparison of adolescents 13–17 years old, with evidence of recent alcohol use, presenting to a PED with a self-reported history of marijuana use. Assessment tools included the Adolescent Drinking Inventory, Adolescent Drinking Questionnaire, Young Adult Drinking and Driving Questionnaire, Center for Epidemiologic Studies Depression Scale, Behavioral Assessment System for Children, and Peer Substance Use and Tolerance of Substance Use Scale.
Results: Compared to adolescents using alcohol only (AO), adolescents who use alcohol and marijuana (A+M) have higher rates of smoking (F = 23.62) and binge drinking (F = 11.56), consume more drinks per sitting (F = 9.03), have more externalizing behavior problems (F = 12.53), and report both greater peer tolerance of substance use (F = 12.99) and lower parental monitoring (F = 7.12).
Conclusions: Adolescents who use A+M report greater substance use and more risk factors for substance abuse than AO-using adolescents. Screening for a history of marijuana use may be important when treating adolescents presenting with an alcohol-related event. A+M co-use may identify a high-risk population, which may have important implications for ED clinicians in the care of these patients, providing parental guidance, and planning follow-up care.
ACADEMIC EMERGENCY MEDICINE 2010; 17:63–71 © 2010 by the Society for Academic Emergency Medicine
Adolescents are among the highest users of alcohol and have some of the highest rates of problematic drinking.1 In the emergency department (ED) setting, significant numbers of both injured2–4 and noninjured5 adolescents have been found to use alcohol. Despite recommendations by both the American College of Surgeons Committee on Trauma6 and the Society for Academic Emergency Medicine7,8 for alcohol screening and intervention among ED patients, emergency physicians frequently fail to detect or refer patients for alcohol treatment.9,10
Several studies have identified screening tools that have shown promise for identifying alcohol-using adolescents in the ED.11–13 However, identification alone may not be sufficient. Pediatric ED (PED) intervention studies for such patients have found that while alcohol-using adolescents may benefit from alcohol interventions in the ED, adolescents with a history of problematic alcohol use experience greater benefit.14,15 Screening for a history of problematic adolescent alcohol and other drug use may thus be especially important in the ED setting.
Some research suggests that adolescent marijuana use may be an important “gateway” to other adolescent substance use, substance use disorders, and other psychosocial problems.16 It is unclear whether asking adolescents about marijuana use is an effective screen for problematic substance use and/or other psychosocial problems.
Studies of older adolescents and young adults support the idea that alcohol and marijuana (A+M) co-use portends worse outcomes than using alcohol only (AO). Stenbacka17 examined alcohol and substance use in a large national registry of Swedish men conscripted for military service at ages 18 and 19 years. At 27-year follow-up, the combination of problematic A+M use in late adolescence was more strongly associated with adult alcohol abuse (risk ratio [RR] = 6.56) and substance abuse (RR = 19.37) than either problematic adolescent alcohol (RR = 3.21) or marijuana (RR = 2.04) use alone.
Flory et al.18 assessed a community sample of 481 older adolescents (19 to 21 years old) who had previously participated in a school-based (grades 6–10) drug use prevention program. The investigators estimated developmental trajectories of substance use based on early versus late onset of A+M use. At adult reassessment, A+M-using individuals were found to have more psychopathology, antisocial personality symptoms, total arrests, and A+M abuse/dependence than AO individuals.
Several studies have compared college students who used AO to those who used A+M.19–21 Together, these studies found that A+M-using students experienced far more problems than AO-using students, including hangovers, doing poorly in school, getting into arguments and physical altercations, driving while drunk, and riding with a drunk driver.
These studies consistently indicate that older adolescents who abuse both A+M display greater impairment than AO-using peers. It is unclear whether the same is true for younger adolescents and what additional factors might be related to A+M use. To date, only one study investigating this question has included younger adolescents. Shillington and Clapp22 used data from the National Longitudinal Survey of Youth and the Young Adult Survey to study adolescents ages 15–21 years.22 The authors found that adolescents using A+M experienced more behavioral problems than their AO-using peers.
Studies have also shown that parent and peer influences are strong predictors of adolescent substance use. For example, levels of parental contact and supervision,23,24 parental norms against drinking,25 and number of deviant peers26 are all known to affect an adolescent’s risk of substance use.
The purpose of this study was to assess whether differentiating AO versus A+M use among younger adolescents presenting to a PED is an effective way to identify teens with impaired functioning. These groups were compared regarding substance abuse and mental health history, as well as parenting and peer factors. It was hypothesized that A+M-using adolescents would have more impairment in all of these areas (i.e., substance abuse and psychosocial functioning) than AO-using adolescents.
This study was part of a larger clinical trial evaluating the efficacy of a brief individual intervention compared to a brief individual intervention plus a family-based intervention. All procedures were approved by the overseeing university and hospital institutional review boards.
Study Setting and Population
The study site was a PED and an adjacent general ED, both of which are Level 1 trauma centers serving a Southern New England catchment area. Adolescents, 13 to 17 years old, treated in a PED between May 2003 and November 2007, were eligible for this study if they had evidence of alcohol use prior to their PED visit. Patients were consecutively approached for enrollment during this time period. Alcohol-positive adolescents were referred to the study by the PED or biochemistry lab staff. Adolescents who were suicidal, were in police custody, had serious injuries or illness requiring hospitalization, were not residing with a parent or guardian, or were non–English speaking were excluded from this study. Some patients did not complete their baseline assessments, and their data are not included here.
Participants were asked about their race and whether or not they considered themselves to be ethnically Hispanic or Latino. Because of the small number of responses in several racial categories, and to be able to make statistically meaningful comparisons, participants were coded into two groups, “white” (non-Hispanic white) and “minority” (Hispanic or nonwhite). Annual household income was divided into nine categories, ranging from less than $5,000 (n = 1) to greater than $150,000 (n = 9).
Research staff were sent an electronic page when an adolescent was identified in the PED either as having consumed any alcohol within 6 hours of his or her admission (PED staff) or as having tested positive (i.e., blood alcohol content [BAC] > 0.0 g/mL) for alcohol via a medical staff–drawn blood test (PED or biochemistry lab staff). Participants were divided into AO and A+M co-use categories based on each subject’s self-reported marijuana use. To be considered a marijuana user, participants answered yes to the question “Have you ever used marijuana?”
The data reported here were collected as part of the baseline adolescent and parent assessments. Baseline assessments for both the adolescents and their parents were completed prior to randomization to the interventions being tested in the larger clinical trial. All participants were required to pass a brief mental status exam before completing the baseline assessments. Parental consent and adolescent assent were obtained for all the adolescents in the study. The consent/assent procedure included assurances that the parents would not be informed of any of the teen’s responses. No adolescent was approached until the BAC was below 0.1 g/mL, he or she could pass a mental status exam, or both. If patients were unable to complete their participation during their PED visit, they were scheduled to return to the hospital within a few days.
The assessment measures were read aloud to adolescents and self-administered by parents. The assessment battery took approximately 45 minutes to complete.
Substance Use and Related Behaviors. The Adolescent Drinking Inventory (ADI)27 is a 24-item self-report measure that focuses on social, psychological, and physical symptoms of alcohol problems. The Adolescent Drinking Questionnaire28 assesses recent drinking frequency, quantity, frequency of high-volume drinking, frequency of intoxication over the prior 3 months, and maximum number of drinks consumed on any one occasion. Two items from the Young Adult Drinking and Driving Questionnaire29 were asked, the frequency of driving after engaging in any drinking at all and riding with an intoxicated driver.
Teen Mood and Behavioral Problems. The Center for Epidemiologic Studies-Depression Scale (CESD)30,31 measures depressive symptomatology. The Behavioral Assessment System for Children (BASC)32 Parent Rating Scale evaluates adaptive and problem behaviors in adolescents ages 14–18 years in community and home settings. The adolescent’s parents completed a BASC.
Peer Factors. The Peer Substance Use and Tolerance of Substance Use33 scale consists of items in which adolescents estimate how many of their friends use alcohol, marijuana, and other drugs. The scale also assesses what the adolescents think their close friends feel about their using marijuana, alcohol, and other drugs. Child-Parent Peer Ratings and Social Skills Inventories34 assess interactions with prosocial and deviant peers. Separate scores were derived for parents and adolescents.
Parenting Factors. Parental monitoring was examined using two scales. Four questions from the Strictness/Supervision Scale35,36 asked both parents and adolescents to what extent the parent knew where the adolescent went at night, how the adolescent spent his or her time, where the adolescent went after school, and who the adolescent’s friends were. Each item was rated on a 4-point scale. Adolescents and parents also completed the Family Management subscale from the Parent/Student Self-Check,34 which assesses parental family management including parental monitoring. Adolescents completed the 14-item scale regarding the parental figure of their choosing (e.g., mother, father).
All dependent variables were first checked for distributional assumptions. Several had skewed distributions and were log-transformed prior to analysis. To assist in clinically intuitive interpretation of results, non–log-transformed data are presented in the tables. Bivariate analyses were performed to determine whether there were any demographic differences between the two groups. The two groups were compared for equivalency of baseline variables using t-tests and Fisher’s exact test. A series of 2 × 2 analysis of variance (ANOVA) tests was performed on all outcome variables, to control for a significant difference among groups by race. To adjust for multiple comparisons, a Bonferroni correction was calculated. The Cronbach alpha, a measure of internal consistency, was calculated for psychometric measures. Values of 0.70 or greater are generally accepted as demonstrating adequate internal consistency. All statistical calculations were performed using SPSS software (version 15.0, SPSS Inc., Chicago, IL).
A total of 242 patients were screened for this study (Figure 1). Fifty-three (21%) were ineligible for the study. An additional 31 patients did not complete the recruitment process (13%). Of the remaining 158 eligible patients, 45 declined to participate. The most commonly cited refusal reasons were lack of interest in the study, they did not believe that their alcohol use was a problem, and/or they did not want to talk about their alcohol use. Recruited and nonrecruited patients were not significantly different in age (15.3 years vs. 15.6 years) or sex (45% vs. 49% male). They did differ in reason for ED visit; recruited patients were more likely to have presented for alcohol intoxication only, while nonrecruited patients were more likely to have presented with the combination of alcohol use and an injury or an “other” reason (χ2 = 7.12, df = 2, p = 0.03).
The study sample consisted of 113 patients, 51 (45.1%) boys and 62 (54.9%) girls, with a mean (± standard deviation [±SD]) age of 15.3 years (SD ±1.2). The majority of participants (n = 108, 96%) were recruited from the PED; the other five subjects were recruited from the general ED. Participants identified themselves as African American (1.8%), Asian (2.7%), white (67.3%), Hispanic (26.5%), and more than one race (1.8%). Participants presented to the PED for the following reasons: 96 (85% of sample) for alcohol intoxication only; six (5.3%) with intoxication in conjunction with a motor vehicle crash; two (1.8%) each for lacerations, assault-related injuries, falls, and being ill; and one (0.9%) each for a sports-related injury, nonsuicidal psychiatric condition, and an unclassified event. Twelve percent of patients completed their baseline assessment in the ED, 70% on a visit to the study’s office, 13% on a home visit, and 5% in another location.
Fifty-seven percent (64/113) of subjects reported using AO, and 43% (49/113) reported using both A+M. Adolescents who used both A+M were not significantly different from those who used AO in terms of age (15.4 years vs. 15.3 years), sex (53% vs. 39% male), or income class (5.6 vs. 6.1). Subjects did differ on race, however: 77% of the A+M group was white, while 59% of the AO group was white (χ2 = 4.16, df = 1, p = 0.04).
Cronbach alphas in this study were as follows: CESD, 0.90; Peer Substance Use and Tolerance of Substance Use, 0.79 (peer substance use) and 0.91 (peer tolerance of substance use); Child-Parent Peer Ratings and Social Skills Inventories, 0.82 and 0.86 (parent ratings of prosocial peers and deviant peers, respectively) and 0.67 and 0.55 (adolescent ratings of prosocial peers and deviant peers, respectively); Strictness/Supervision Scale, 0.66 (adolescent scale) and 0.88 (parent scale); Student Self-Check, 0.91; and Parent Self-Check, 0.88.
Table 1 presents a comparison of other substance use variables by group. Compared to AO participants, A+M participants smoked more cigarettes per day, had more days of binge drinking (more than five drinks in one sitting), and consumed a greater number of drinks per drinking occasion over the prior 3 months. There was also a trend (p < 0.05) among A+M teens to consume a higher maximum number of drinks at one sitting over the prior 3 months, have higher BACs at the time of their PED visit, to have ridden with an impaired driver, and to have higher ADI scores, compared to AO adolescents. Additional analyses (data not shown in Table 1) revealed that among A+M-using adolescents, 87.2% smoked cigarettes, compared to only 20.3% of AO adolescents (χ2 = 46.56, df = 1, p = 0.000).
|Substance Use and Related Variables||AO Use||A+M Use||F|
|Mean (SD)||n||Mean (SD)||n|
|Cigarettes per day|
|White||0.24 (1.09)||37||6.82 (8.69)||38||23.62 (by +M)*|
|Minority||0.54 (2.08)||26||1.55 (3.59)||11||6.56 (by race)†|
|Total||0.37 (1.57)||63||5.63 (8.11)||49||8.61 (interaction)*|
|Days drinking per month|
|White||1.92 (2.65)||38||4.26 (5.07)||38||2.58 (by +M)|
|Minority||1.87 (4.15)||26||1.64 (2.84)||11||5.57 (by race)†|
|Total||1.90 (3.31)||64||3.67 (4.77)||49||2.34 (interaction)|
|Days binge drinking per month|
|White||0.45 (0.28)||38||2.90 (5.24)||38||11.56 (by +M)*|
|Minority||1.15 (3.85)||26||1.05 (1.15)||11||1.32 (by race)|
|Total||0.73 (2.46)||64||2.48 (4.70)||49||3.14 (interaction)|
|Number of drinks per sitting/past 3 months|
|White||4.92 (1.72)||38||5.74 (1.64)||38||9.03 (by +M)*|
|Minority||4.65 (1.74)||26||6.00 (1.61)||11||0.00 (by race)|
|Total||4.81 (1.72)||64||5.80 (1.62)||49||0.54 (interaction)|
|Maximum number of drinks/past 3 months|
|White||7.90 (4.68)||38||11.68 (6.18)||38||4.83 (by +M)†|
|Minority||8.23 (6.54)||26||9.64 (2.25)||11||0.52 (by race)|
|Total||8.03 (5.47)||64||11.22 (5.59)||49||1.02 (interaction)|
|Calculated BAC in PED|
|White||0.24 (0.10)||38||0.26 (0.09)||38||6.76 (by +M)†|
|Minority||0.21 (0.09)||26||0.29 (0.08)||11||0.06 (by race)|
|Total||0.23 (0.10)||64||0.27 (0.09)||49||1.97 (interaction)|
|Number of times rode with impaired driver|
|White||0.63 (1.82)||38||3.95 (7.80)||38||6.18 (by +M)†|
|Minority||1.19 (2.99)||26||1.18 (1.40)||11||0.18 (by race)|
|Total||0.86 (2.36)||64||3.33 (6.97)||49||2.51 (interaction)|
|Adolescent Drinking Inventory|
|White||10.79 (6.58)||38||18.42 (11.82)||38||4.65 (by +M)†|
|Minority||11.58 (8.97)||26||12.55 (8.58)||11||1.63 (by race)|
|Total||11.11 (7.58)||64||17.10 (11.40)||49||2.79 (interaction)|
Regarding other illicit substance use, 11 of the 49 (22%) A+M subjects reported such use, compared to one of the 64 (2%) AO subjects. Three A+M subjects also used stimulants; two admitted to “designer drug” use; and one each endorsed cocaine, opiate, and “other medication” use. Three A+M subjects used multiple other substances, including designer drugs, stimulants, sedatives, cough medicines, and hallucinogens. The lone AO subject reported “other medicine” use.
Table 2 examines depressed mood and behavior variables by group. Parents of A+M teens reported that their teens had significantly more externalizing problem behaviors than parents of AO teens. There was a trend for A+M teens to report higher depression (CESD) scores than AO teens. Additionally, there was a trend for parents of A+M teens to report their teens having more internalizing problem behaviors.
|Mood and Behavior Variables||AO Use||A+M Use||F|
|Mean (SD)||n||Mean (SD)||n|
|White||8.92 (7.24)||38||15.26 (12.10)||38||4.29 (by +M)†|
|Minority||10.69 (10.15)||26||13.09 (8.63)||11||0.01 (by race)|
|Total||9.64 (8.51)||64||14.78 (11.36)||49||0.87 (interaction)|
|Externalizing behaviors (BASC)|
|White||56.73 (13.46)||37||66.64 (17.13)||36||12.53 (by +M)*|
|Minority||53.04 (13.42)||24||66.91 (20.27)||11||0.26 (by race)|
|Total||55.28 (13.46)||61||66.70 (17.68)||47||0.35 (interaction)|
|Internalizing behaviors (BASC)|
|White||50.62 (9.62)||37||55.44 (12.86)||36||4.03 (by +M)†|
|Minority||49.54 (8.55)||24||54.82 (18.03)||11||0.11 (by race)|
|Total||50.20 (9.16)||61||55.30 (14.02)||47||0.01 (interaction)|
Analysis of peer variables by group is presented in Table 3. Adolescents who used A+M reported greater peer tolerance of substance use than AO adolescents. There was a trend for A+M adolescents to report more peer substance use, more deviant peers (both teen and parent report), and fewer prosocial peers (teen report only) compared to AO teens.
|Peer Variables||AO Use||A+M Use||F|
|Mean (SD)||n||Mean (SD)||n|
|Peer substance use|
|White||0.86 (0.75)||38||1.59 (0.84)||38||6.63 (by +M)†|
|Minority||0.88 (0.86)||26||1.03 (0.75)||11||2.42 (by race)|
|Total||0.86 (0.79)||64||1.46 (0.84)||49||2.85 (interaction)|
|Peer tolerance of substance use|
|White||2.10 (0.67)||38||2.54 (0.69)||38||12.99 (by +M)*|
|Minority||1.84 (0.88)||26||2.57 (1.01)||11||0.50 (by race)|
|Total||2.00 (0.76)||64||2.55 (0.76)||49||0.80 (interaction)|
|Deviant peers (teen report)|
|White||1.81 (0.59)||38||2.66 (0.90)||38||5.40 (by +M)†|
|Minority||2.25 (0.85)||26||2.16 (0.66)||11||0.04 (by race)|
|Total||1.99 (0.73)||64||2.55 (0.87)||49||8.28 (interaction)*|
|Prosocial peers (teen report)|
|White||3.67 (0.67)||38||3.08 (0.65)||38||4.41 (by +M)†|
|Minority||3.39 (0.79)||26||3.36 (0.55)||11||0.00 (by race)|
|Total||3.55 (0.73)||64||3.14 (0.64)||49||3.82 (interaction)|
|Deviant peers (parent report)|
|White||1.64 (0.72)||37||2.51 (0.97)||37||8.60 (by +M)*|
|Minority||2.10 (1.13)||23||2.46 (1.22)||11||0.91 (by race)|
|Total||1.82 (0.92)||60||2.47 (1.01)||48||1.53 (interaction)|
|Prosocial peers (parent report)|
|White||3.68 (0.89)||37||2.85 (0.88)||37||1.56 (by +M)|
|Minority||2.75 (1.14)||24||3.05 (1.19)||11||2.99 (by race)|
|Total||3.31 (1.09)||61||2.91 (0.95)||48||7.00 (interaction)†|
Table 4 presents a comparison of parental monitoring variables by group. On the Strictness/Supervision Scale, parents of A+M teens reported less parental monitoring than parents of AO teens. On the Family Management Scale of the Student Self Check, A+M teens reported less parental monitoring by their primary caregiver than AO teens. There was also a trend among A+M-using teens to report less parental monitoring on the Strictness/Supervision Scale than AO teens.
|Parenting Variables||AO Use||A+M Use||F|
|Mean (SD)||n||Mean (SD)||n|
|Strictness/Supervision Scale (teen)|
|White||3.34 (0.54)||38||2.94 (0.55)||38||4.17 (by +M)†|
|Minority||3.14 (0.53)||26||3.07 (0.61)||11||0.09 (by race)|
|Total||3.26 (0.54)||64||2.97 (0.56)||49||1.93 (interaction)|
|Strictness/Supervision Scale (parent)|
|White||3.49 (0.45)||37||3.10 (0.64)||37||7.12 (by +M)*|
|Minority||3.02 (0.76)||24||2.66 (0.98)||11||10.49 (by race)*|
|Total||3.31 (0.63)||61||3.00 (0.74)||48||0.01 (interaction)|
|Family Management Scale (PSC)|
|White||7.15 (1.53)||33||5.90 (1.73)||32||5.41 (by +M)†|
|Minority||6.96 (1.98)||19||6.17 (2.53)||8||0.01 (by race)|
|Total||7.08 (1.69)||52||5.95 (1.88)||40||0.27 (interaction)|
|Family Management Scale (SSC)|
|White||7.67 (1.74)||38||6.80 (1.39)||38||7.44 (by +M)*|
|Minority||7.32 (1.87)||25||6.25 (1.80)||11||1.57 (by race)|
|Total||7.53 (1.79)||63||6.67 (1.49)||49||0.08 (interaction)|
Because the two groups differed by race, 2 (AO vs. A+M) × 2 (white vs. minority) ANOVAs were conducted, using all outcome variables. A main effect of race was observed on the parent report of parental monitoring on the Strictness/Supervision Scale [F(1,88) = 5.41, p = 0.022], with minority parents reporting significantly less monitoring of their teens than white parents. There was a significant race × marijuana status interaction for cigarette smoking [F(1,108) = 8.61, p = 0.004] and deviant peers [F(1,109) = 8.28, p = 0.005].
To elucidate the effect of each factor on the other factors, simple-effects analyses examining the unique effects of one factor at each level of the other factor were performed. Simple effects analysis of cigarette smoking revealed that white A+M adolescents smoked significantly more cigarettes per day than minority A+M teens [F(1,108) = 11.77, p = 0.001]. Among the AO teens, there was no significant difference by race. Simple-effects analysis of deviant peers revealed that minority AO teens associated with more deviant peers than white AO adolescents [F(1,109) = 5.05, p = 0.027]. Among A+M teens, there was no significant difference by race.
This study’s findings are consistent with those of other studies of older adolescents and college students,18–21 which found that older adolescents and young adults who used both A+M reported more psychosocial, family, and peer risk factors than those who use AO.17 In this study, younger adolescents who use both A+M have a significantly greater likelihood of increased alcohol and tobacco use compared to AO-using teens. Additionally, these adolescents were more likely to report less parental supervision and a greater number of deviant peers.
The findings of this study have important ramifications for ED clinicians who care for adolescents who present to the ED with an alcohol-related event. A positive response to questions about both A+M use should alert the clinician that the patient may be at higher risk for a number of untoward outcomes than if only alcohol use is endorsed. In particular, it is likely that an A+M-using patient will be a heavier drinker and cigarette smoker than an AO-using patient. The clinician should probe as needed for other risky behaviors. The ability to rapidly screen for high-risk patients may help ED clinicians utilize their time and resources more efficiently and efficaciously.
Spirito et al.15 found that adolescents with problematic drinking had a greater response to ED alcohol interventions than adolescents without problematic drinking. More intensive interventions, especially those that increase parental involvement with and monitoring of their teen, may be necessary for successful intervention with adolescents using both A+M.37 Identifying high-risk adolescent drinkers is thus important not only in terms of post-ED visit follow-up care, but may also help determine which ED patients are mostly likely to benefit from ED interventions.
Brief screening tools are particularly appealing to ED clinicians, given the busy, hectic nature of the ED clinical environment. Single screening question tools have been shown to successfully identify alcohol-using college students at high risk for alcohol-related injury38 and alcohol misuse among primary care patients.39 A two-item depression screening tool has also been validated in a PED population.40 Given the burgeoning emphasis on screening and brief intervention with referral to treatment (SBIRT) by the Society for Academic Emergency Medicine,7 the American College of Surgeons,6 and other professional medical organizations, effective screening methods are a growing need. A single question about marijuana use, asked of alcohol-positive ED adolescent patients, may be an important way for ED clinicians to identify particularly high-risk adolescents with problematic alcohol use. The 2007 Youth Risk Behavior Surveillance survey found a 38.1% lifetime prevalence of marijuana use by adolescents in the United States.41 This high prevalence of marijuana use in national surveys and in this study lends plausibility to the idea that asking about its use may be a high-yield screening question.
The results of this study differ from that of Shillington and Clapp,22 the only other study that included younger adolescents. While both studies found higher rates of psychosocial problems in A+M-using adolescents, the previous study found no difference in drinking problems after controlling for demographic variables, co-morbid A+M use, other substance use variables, and psychological factors (religiosity and impulsivity). In contrast, this study found greater tobacco and alcohol use among A+M-using teens. A possible explanation for the observed differences between the two studies includes the methods used to identify marijuana use, the settings, and differences in the psychosocial variables included in the analyses. Another difference between the two studies is the age of the study populations. In this study, participants were younger (mean age = 15.4 years) than in the previous study (mean age = 17.1 years). It is known that adolescent substance use patterns vary significantly by age.42–44 Marijuana use among 18- to 20-year-old adolescents is three to four times that of 12- to 14-year old adolescents.42,43 It is possible that adolescents who use marijuana at an earlier age have worse substance use outcomes.
The fact that adolescents in this study are younger than those in the previous studies, but had similar patterns of problems as studies with older subjects, supports the idea that problematic young adult substance abuse often begins in adolescence. For example, using data from the National Longitudinal Survey of Youth, Muthén and Muthén45 found that the strongest predictor of heavy drinking at age 18 was early onset of drinking. Hingson and Zha46 examined data from the National Epidemiologic Survey of Alcohol and Related Conditions and found that earlier age of onset drinking significantly increased the likelihood of adult alcohol use disorders and alcohol-related problems. Early detection and intervention for substance use in adolescence may thus be important in forestalling the development of adult substance abuse.
The ethnic and racial characteristics of this sample were such that subjects had to be grouped into two categories (white vs. minority), which affects the interpretability of this study. Additional research is needed to examine racial differences among AO- and A+M-using adolescents. Because of the cross-sectional nature of this data, the precise relationship between marijuana use and the observed associations cannot be determined. It is also possible that those adolescents who participated in this study are different than those who declined to participate or did not complete their baseline assessment batteries, which may also limit the generalizability of this study’s findings. Another limitation is that all of the study measures rely on adolescent or parent self-report. It is possible that the responses were not accurate or that the participants provided socially desirable answers. However, if this occurred, it is likely that this study’s results underestimate the true association. Another possibility is that how truthfully patients answered the questions may have been influenced by having been recruited in the PED. The adolescents may have feared repercussions if they admitted to use. However, most of the baseline assessments occurred after discharge from the PED, which may have limited this potential problem.
There are numerous areas for future research, which would further the understanding of how A+M co-use contributes to substance-related and psychosocial problems. Previous studies have found significantly higher rates of substance-related violence and problems with the legal and law enforcement systems among adults who use A+M than those that use AO.20,21 This relationship remained robust even after controlling for potential confounders. It is unclear if this association is true for younger adolescents as well. Another important, unanswered question is whether the apparently worse clinical condition of A+M (compared to AO) using younger adolescents is causally due to marijuana use. Is marijuana truly the “gateway” to other substances and psychosocial problems? Or is marijuana use an important marker of persons at high risk for substance and/or psychosocial problems? For example, the relationship between marijuana and tobacco use, as seen in this and other studies,47 is well known. What is less clear is the causal (if at all) nature of the relationship. Understanding these relationships is critical to successfully designing and implementing effective screening, treatment, and prevention programs. Finally, prospective validation of comorbid A+M use as a screen for problematic alcohol use and poor psychosocial functioning is needed.
Adolescents who use both alcohol and marijuana are at greater risk for other substance use than adolescents who only use alcohol. These adolescents are also at higher risk for other behavioral and psychosocial problems. Screening alcohol-positive adolescents for history of marijuana use may thus be an efficient method of identifying particularly high-risk adolescents. These patients and their families may need more intensive interventions, both while in the ED and after discharge, to address substance use, parenting, and other psychosocial problems. Further research is needed to investigate this relationship between alcohol and marijuana use and psychosocial functioning.
- 1U.S. Department of Health and Human Services. Tenth Special Report to the U.S. Congress on Alcohol and Health. 2000. Available at: http://pubs.niaaa.nih.gov/publications/10report/intro.pdf. Accessed Oct 13, 2009.
- 6American College of Surgeons, Committee on Trauma. Resources for the Optimal Care of Injured Patients: 2006. Chicago, IL: American College of Surgeons, 2006.
- 8Preventive care in the emergency department, Part I: Clinical preventive services--are they relevant to emergency medicine? Society for Academic Emergency Medicine Public Health and Education Task Force Preventive Services Work Group. Acad Emerg Med. 2000; 7:1036–41., , .
- 24Alcohol Use Among Adolescents. Thousand Oaks, CA: Sage Publications, 1999, p 126..
- 27Screening for adolescent problem drinking: validation of a multidimensional instrument for case identification. Psychologic Assess. 1989; 1:61–3., .
- 30The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977; 1:385–401..
- 32The Clinician’s Guide to the Behavior Assessment System for Children (BASC). New York, NY: Guilford, 1992., .
- 34Intervening in Adolescent Problem Behaviors. A Family Centered Approach. New York, NY: Guilford, 2003., .
- 37Reaffirming the power of parental influence on adolescent smoking and drinking decisions. Adolesc Fam Health. 2001; 2:37–43., .
- 41Youth risk behavior surveillance--United States, 2007. MMWR Surveill Summ. 2008; 57:1–131., , , et al.
- 42National Institute of Drug Abuse. NIDA InfoFacts: High School and Youth Trends. Available at: http://www.nida.nih.gov/Infofacts/HSYouthtrends.html. Accessed Sep 26, 2009.