The Changing Pattern of Substance Abuse in Urban Adolescents FREE

THE USE of illicit drugs by adolescents in the United States has risen since reaching a nadir in the late 1980s and early 1990s. This increase was first noted in 1993 drug use surveys and the trend has continued through 1995.^{1 - 7} The use of illicit drugs has been shown to be associated with leading causes of morbidity and mortality as well as low educational achievement and increased levels of school dropout in this age group.^{8 - 10} The link between drug use and sexual risk-taking behaviors has been documented^{11 - 12} as has the association of drug use and increased risk for sexually transmitted diseases and human immunodeficiency virus infection.^{13 - 17} For those caring for adolescent patients, a reduction in substance use is therefore a highly sought after clinical end point.

For adolescent health care providers to target care and counseling efforts to the current drug use environment, an evaluation of the prevalence of current drug use patterns as well as an identification of shifts in drug use trends is helpful. The purpose of this study was to determine the prevalence of specific drug use in adolescents attending our urban adolescent health clinic. It is based on blinded and anonymous urine testing and compares recent rates of use and substance "preference" with a similar, previous study.

In the course of routine care, urine specimens were obtained from 1313 patients seen in the Children's National Medical Center (CNMC) Adolescent Medicine Outpatient Clinic, Washington, DC, for a 1-year period (April 1995 to March 1996). Once the appropriate clinical evaluation of the patient was complete, discarded urine specimens had all identifying information removed and destroyed, were coded by age and sex, and assigned a unique study number. All specimens were then tested for the presence of 5 drugs (cannabinoids, phencyclidine [PCP], cocaine metabolites, amphetamines/methamphetamines, and opiates) via fluorescence polarization immunoassay (ADX, Abbott Laboratories, Abbott Park, Ill). Specimens that screened positive for any drug were sent to a reference laboratory for gas chromatography/mass spectrometry confirmation testing. A similar study took place between October 1989 and September 1990. In that study, urine specimens were obtained in a blinded fashion from 1312 patients and were analyzed by an Abbott ADX system for the presence of 4 drugs (amphetamines, cannabinoids, cocaine and PCP). Specimens that screened positive for any drug were set to a reference laboratory for gas chromatography confirmation testing.

Both studies were reviewed and approved by the CNMC Institutional Review Board. The most recent study was also approved by the National Institutes of Health Office for Protection From Research Risks, Rockville, Md. Patient consent was not required for either study.

Results from both years were grouped by substance tested, age, and sex. Statistical analyses were performed using BMDP Statistical Software, Release 7 (SPSS Inc, Chicago, Ill) and StatXact 3 for Windows (Cytel Software Corp, Cambridge, Mass). Descriptive χ² analysis, tests of proportions, and 95% confidence intervals were reported as appropriate. Differences in drug use rates for sex and age groupings for each study year and differences in rates between study periods were analyzed. Differences in rates between study periods for sex and age groupings were also analyzed. The critical α level for all tests of significance was .05. Adjustments to the criteria for significance were made for multiple comparisons.

Between April 1995 and March 1996, 1313 patients presenting for routine health care in the CNMC Adolescent Medicine Outpatient Clinic were tested for the presence of 5 drugs in urine specimens. In this temporal cohort, males composed 33% of the population and females 67%; the mean±SD age of the patients was 16.07±2.14 years (Table 1). In a similar previous study (1989-1990), 1312 patients (Table 1) attending the CNMC clinic were tested for 4 drugs. Males composed 20% of the population and females 80%; the mean age of the population was 16.62±1.88 years.

For the most recent patient group, 179 (14%) were confirmed to test positive for 1 or more drugs. Of those with a positive drug screen, 166 (13%) were confirmed to be positive for cannabinoids (Table 2). Males were significantly more likely to test positive for at least 1 drug than females (18% vs 11%; P<.001). Males were also significantly more likely to test positive for cannabinoids than females (17% vs 10%; P<.001) (Table 3). There was a significant association between age group and confirmed use of 1 or more drugs (P<.001). The 18- to 21-year-old group had a significantly higher rate of overall drug use than the 12- to 14-year-old group and the 15- to 17-year-old group. The 15- to 17-year-old group also has a significantly higher use rate than the 12- to 14-year-old group. There was a significant association between age group and cannabinoid use (P<.001). The 15- to 17-year-old and 18- to 21-year-old groups had significantly higher cannabinoid use rates than the 12- to 14-year-old group (P=.001 for both) (Table 4).

In the earlier cohort, positive urine tests for 1 drug category were found for 65 (5%) of the specimens. Positive urine tests for cannabinoids were found for 30 (2%) of 1265 of the specimens and positive urine tests for cocaine were found for 41 (3%) of 1289 of the specimens (Table 2). There was no significant association between sex and overall drug use. Males, however, tended to have higher cocaine use than females (Table 3). There were significant associations between age group and the use of at least 1 drug, marijuana, and cocaine (P<.001). The oldest age group had significantly higher overall drug, cannabinoid, and cocaine use rates than the younger age groups (Table 4).

Comparing the 2 cohorts, there was a significant increase in the proportion of males and a significant decrease in mean age for the most recent study group (Table 1). Patients tested recently were significantly more likely to have confirmed positive urine test results for at least 1 drug (P<.001) and for cannabinoids in particular (P<.001), and significantly less likely to have urine specimens positive for cocaine than those patients screened in 1989 to 1990 (P<.001) (Table 2). To explore the effect of sex and age differences across study periods, sex and age groupings were analyzed individually to determine shifts in drug use between study years. The increase in overall drug use and cannabinoids in particular was consistent across both sex and age groupings. The decrease in cocaine use was also consistent across both sex and age groupings (Table 3 and Table 4).

This study demonstrates an increase in positive urine test results in patients seen in our adolescent ambulatory clinic between 1989 and 1990 and 1995 and 1996, with a strong shift toward cannabinoid use and a shift away from cocaine use; this shift was consistent across both sex and age groupings. This upward trend in marijuana usage is similar to reported drug-use trends reported by national surveys.^{6 - 7 ,18} In our patients, this increase in marijuana use has been offset by a significant decrease in cocaine use, even greater than that seen in national survey.^{5 - 6} The higher use of drugs in general and marijuana in particular by males and older adolescents found in our study is also similar to the findings of national surveys.^{1 - 3 ,5 - 7}

The findings of this study are unique because the results are based on actual use rather than self-reported data. While most other studies use responses to anonymous surveys, it is our experience that such historical self-reports are not totally accurate.¹⁹ The body of available literature concerning the validity of self-reported substance abuse demonstrates variability in results. Factors such as type of drug, testing environment, and subject characteristics affected the outcome of validity testing.^{20 - 24} Maisto et al,²⁵ in a review of self-reported substance abuse literature, noted that the accuracy of drug abusers' self-report of drug use is typically not high. Using urinalysis in addition to self-reported substance use may provide a means for a more precise estimate of true prevalence. In addition, the development of more accurate self-report measurement instruments may be enhanced by validating these instruments with urinalysis.

However, with the advantage of direct measurement of drugs in the urine comes the disadvantage of the limitations imposed by our testing methods. Urine screening can usually detect drugs in the urine only for a specific period (7-30 days for marijuana, 1-3 days for cocaine, 1-4 days for opiates, and 2-5 days for frequent users of PCP).²⁶

The significantly higher rates in overall drug use and marijuana use in particular are most disturbing, however, because of the particularly rapid rise in drug use by the youngest adolescents. Both the 12- to 14-year-old group and the 15- to 17-year-old group experienced a greater than 4-fold increase in the identification of any drug in their urine. This is troubling because it suggests that adolescents in our community are initiating drug use at an earlier age than a decade ago.

It is interesting that the overall increase in substance use has been somewhat "softened" in our patient population by the shift away from cocaine to marijuana. The cocaine epidemic of the late 1980s and early 1990s was felt to be linked to the spread of a number of sexually transmitted diseases, including syphilis and human immunodeficiency virus infection.^{13 - 14 ,16 - 17} Before we are able to take any solace from this shift, however, we must acknowledge that in the 1980s a similar increase in marijuana use in our patient population signaled an increase in all drug use, including cocaine.

Practitioners need to be aware that drug use patterns in adolescents can shift relatively abruptly and counseling should be targeted to current drug use patterns. The need for such counseling becomes increasingly important because of the established link of drug use to other health risk behaviors.^{8 - 10 ,27 - 28} It is likely that this also holds true for our patients and is of particular concern with our younger adolescent patients. Earlier initiation of risk behaviors will certainly increase the risk of adverse health outcomes. From the established level of documented substance use in our younger adolescents, it is clear that interventions seeking to reduce drug use, premature sexuality, and other risk behaviors must target elementary and junior high school aged children.²⁹ This is a process that must go forward as soon as possible and be as broad-based as possible. The future of a whole generation is at stake.

Accepted for publication October 9, 1997.

This research was supported by grant R01 DA08019-03 from the National Institute of Drug Abuse, Washington, DC.

Editor's Note: One nice aspect of this study is the correlation between self-report and urine screen results; the not-so-nice part is the clear increase in substance abuse by adolescents. They use drugs but they tell the truth.—Catherine D. DeAngelis, MD

Reprints: Lawrence J. D'Angelo, MD, MPH, Children's National Medical Center, Adolescent and Young Adult Medicine, 111 Michigan Ave NW, Washington, DC 20010.