Epidemiology of Youth Drug Abuse

This section lists selected summaries from NIDA funded research projects that investigate the epidemiology, etiology, and prevention research. The summaries provided were selected from recent issues of the Director's Report to the National Advisory Council on Drug Abuse. For a more comprehensive listing of NIDA funded projects see the Director's Report.

Adolescent Conduct Disorder Predictive of HIV Risk Taking Behaviors

A study was conducted to assess the prevalence of conduct disorder among runaway and homeless adolescents and to investigate associations between conduct disorder and HIV risk behaviors. The Diagnostic Interview Schedule for Children and a standardized HIV risk assessment questionnaire were administered to 219 runaway and homeless adolescents recruited from a drop-in center serving high risk youth. One half of the males and 60% of the females were diagnosed with conduct disorder. In multivariate analyses, conduct disorder was the strongest predictor of lifetime use of heroin and/or cocaine and exchanging sex for money, drugs, food or shelter, as well as the number of drugs used and the number of sex partners in the last 3 months. The high rate of conduct disorder in this population, and the association between conduct disorder and both drug and sex-related HIV risk behaviors, indicate a need for interventions that consider the influence of psychiatric diagnosis on high risk behaviors. Booth, R.E. and Zhang, Y. Conduct Disorder and HIV Risk Behaviors among Runaway and Homeless Adolescents. Drug and Alcohol Dependence, 48: pp. 69-76, 1997.

Behavioral Change in HIV+ Adolescents

This study examined whether HIV+ adolescents (N=102) linked to care then change risk and health-related behaviors subsequent to learning their HIV serostatus, and to document the stability of their current behavior patterns over a 6-month period. Over their lifetime, youths engaged in unprotected sexual acts with multiple partners (M=284; median=44; consistent condom protection, 5%) and substance use (21% injecting drug use; 68% hard drugs). When current risk behaviors were assessed twice over two consecutive 3-month periods, almost one third had been sexually abstinent. Among youths who were currently sexually active, most had multiple sexual partners (M=5.7, time 1; 4.9, time 2) and used condoms (72-77% sexual acts protected); most of the youths (63-64%) always used condoms. Use of alcohol (63%), marijuana (41%), hard drugs (36%), and injecting drugs (12%) was substantial. There were gender difference among sexual behavior (females had fewer sexual partners) and substance use behavior (fewer females injected drugs or shared needles). Youths were relatively healthy (M T cells=521.4; 14% T cells < 200; 1.9 diseases and 3.7 physical symptoms in the previous 3 months). There was an exceptionally high rate of adherence, about 66% for appointment over 3-6 months. Rotheram-Borus M.J., Murphy D.A., Coleman C.L., Kennedy M., Reid H.M., Cline T.R., Birnbaum J.M., Futterman D., Levin L., Schneir A., Chabon B., O'Keefe Z., & Kipke M. Risk Acts, Health Care, and Medical Adherence among HIV+ Youths in Care Over Time. AIDS and Behavior, 1, pp. 43-51, 1997.

DSM-IV Criteria for Adolescent Alcohol and Cannabis Use Disorders

In a study comparing DSM-III-Revised and DSM-IV criteria for alcohol and cannabis use disorders among 772 teenagers (63% boys; 77% white) referred to adolescent outpatient drug treatment programs in the Minneapolis, MN area, Winters, Latimer and Stinchfield found that compared to the use of the DSM-III-R criteria, the application of DSM-IV criteria for alcohol and cannabis use resulted in more "abuse" assignments and fewer "dependence" assignments. Study results indicated that the shift in diagnostic assignments when using DSM-IV compared to DSM-III-R was due to a lowering of the abuse threshold rather than a tightening of the dependence criteria. Journal of Studies on Alcohol. In Press.

Cannabis Dependence, Withdrawal, and Reinforcing Effects Among Adolescents with Conduct Symptoms and Substance Use Disorders

Crowley and his colleagues from the University of Colorado assessed 209 adolescents (165 males, 64 female; 13-19 year old), who were referred for substance and conduct problems, to determine if cannabis produced dependence and withdrawal. The subjects were diagnosed with substance dependence, 100%; current conduct disorder, 82.1%; major depression, 17.5%; and attention-deficit/hyperactivity disorder, 14.8%. Most patients claimed serious problems from cannabis, and 78.6% met standard adult criteria for cannabis dependence. Two-thirds of cannabis-dependent patients reported withdrawal. Progression from first to regular cannabis use was as rapid as tobacco progression, and more rapid than that of alcohol, suggesting that cannabis is a potent reinforcer. Data indicate that for adolescents with conduct problems cannabis is not benign, and that the drug potently reinforces cannabis-taking, producing both dependence and withdrawal. However, the authors suggested that findings from this severely affected clinical population should not be generalized broadly to all other adolescents. Crowley, T.J., MacDonald, M.J., Whitmore, E.A., and Mukulich, S.K. Drug and Alcohol Dependence, In Press.

Influences on Adolescent Substance Dependence: Conduct Disorder, Depression, Attention Deficit Hyperactivity Disorder, and Gender

Whitmore and his colleagues from the University of Colorado conducted a study to determine if prevalence and severity of conduct disorder (CD), major depression and ADHD would differ by gender, and that these conditions would associate differentially with severity of substance disorder (SD) in males and females. The subjects were 285 male and 82 female adolescents referred for treatment for comorbid CD and SD. The investigators found that males and females did not differ significantly in severity of substance involvement, major depression, or ADHD, but males had more severe CD. Major depression was the only variable significantly associated with SD severity for females, while for males, severity of CD combined with major depression and ADHD was significantly associated with SD severity. The authors concluded that among referred adolescents, CD, major depression, and ADHD may all be important concomitants of SD in males, while in females, depression may be the primary variable related to SD. Whitmore, E.A., Mikulich, S.K., Thompson, L.L., Riggs, P.D., Aarons, G.A., and Crowley, T.J., Drug and Alcohol Dependence, 47, pp. 87-97, 1997.

Concordance for ADHD Greater for Monozygotic than Dizygotic Twin Pairs

Twin concordance was assessed in a community-based sample of 194 monozygotic and 94 dizygotic male twins, aged 11-12 and with DSM-III diagnoses of ADHD from structured interviews of teachers and of mothers. Concordance rates for ADHD diagnoses were greater for monozygotic twin pairs compared to dizygotic, especially for diagnoses based on mothers' ratings; concordance rates were more similar for teachers' ratings. Thus, the magnitude of the genetic component seems to depend, in part, on the informant (of phenotype). Nevertheless, model-fitting analyses indicated strong additive genetic components, smaller non-shared environmental components and no shared environment components. These data are important because ADHD is recognized as a risk factor for development of substance abuse. Sherman, D.K., Iacono, W.G., & McGue, M.K., J Am Acad Child Adolesc Psychiatry, 36(6), 1997; Sherman, D.K., McGue, M.K., & Iacono, W.G., Am J Psychiatry, 154(4), 1997.

Monitoring the Future Study

Results from the 1997 Monitoring the Future Study were released on December 20, 1997 (Lloyd Johnston, et al.). Following are the most significant findings:

Use of marijuana, cocaine, opiates other than heroin, and alcohol in the lifetime increased from 1996 to 1997 among high school seniors, and lifetime marijuana use increased among 10th graders. More recent use (past year and past month) remained stable with few exceptions. Results for 8th graders provided a glimmer of hope, with no increases, and decreases in a few categories. Attitudes toward substance use were mostly stable between 1996 and 1997. Perceived risk of harm was mostly stable for students in all three grades. Encouraging results were found for 8th graders, with disapproval increasing for selected marijuana, cocaine, alcohol, and cigarette use indicators. All changes presented below are statistically significant.

Illicit Drug Use

• Lifetime use of marijuana/hashish increased from 1996 to 1997 among 12th and 10th graders. Seniors' rate of lifetime marijuana use is higher than any year since 1987 but well below the peak seen in 1979. Past year and past month rates for marijuana remained unchanged for 10th and 12th graders. However, daily marijuana use in the past month increased among seniors but decreased among 8th graders.
• Driven in large part by the rise in marijuana, lifetime use of any illicit drug increased among 12th graders. Past month use of any illicit drug decreased among 8th graders; past year aggregate drug use remained stable for each grade.
• Lifetime use of cocaine, both as crack and powder, increased among 12th graders to reach its highest level since 1990 (highest since 1989 for crack). No changes in cocaine use were found for other recency periods or other grades.
• Past month use of LSD increased among 12th graders between 1996 and 1997.
• Heroin use in the past year decreased from 1996 to 1997 among 8th graders.
• Other opiate use in the lifetime and the past year increased among seniors.
• Stimulant use in the past month increased among 12th graders, but use of these drugs in the past year and in the past month decreased among 8th graders.
• Past month use of tranquilizers increased among 10th graders but decreased among 8th graders.
• Use of inhalants, hallucinogens other than LSD, MDMA, ice, and barbiturates remained stable for each grade level and recency-of-use category.

Perceived Harmfulness, Disapproval, and Perceived Availability

• The percentage of 8th graders reporting "great risk" in trying marijuana once or twice decreased. Perceived harmfulness of trying heroin once or twice and taking barbiturates regularly decreased among seniors.
• Disapproval of smoking marijuana occasionally or regularly increased among 8th graders.
• Disapproval of taking cocaine powder occasionally increased among 8th graders.
• Perceived availability of LSD and barbiturates decreased among 10th graders, and perceived availability of amphetamines decreased for 8th and 10th graders.

Alcohol Use

• Alcohol continues to be used at unacceptably high levels. Lifetime and past-year alcohol use increased for 12th graders. However, having "been drunk" in the past month decreased among 8th graders.
• The percentage of 8th graders reporting "great risk" in having five or more drinks once or twice each weekend increased for 8th graders and decreased for 12th graders.

Use of Cigarettes and Smokeless Tobacco

• Between 1996 and 1997, seniors' use of cigarettes in the past month along with daily smoking in the past month increased. Daily cigarette use among seniors is at its highest level since 1979.
• Among 8th graders, daily smoking in the past month deceased, and smoking a half-pack or more cigarettes per day decreased.
• Disapproval of smoking one or more packs of cigarettes per day increased among 8th and 10th graders.
• Lifetime use and past month use of smokeless tobacco decreased among 8th graders.

Community Epidemiology Work Group (CEWG)

The 43rd biannual meeting of the CEWG was held in Phoenix, Arizona on December 9-12, 1997. The CEWG is composed of researchers from 21 metropolitan areas of the United States who meet semiannually to report on patterns and trends of drug abuse in their respective areas; emerging drugs of abuse; vulnerable populations and factors that may place people at risk of drug use and abuse; and, negative health and social consequences. Reports are based on drug abuse indicator data, such as morbidity and mortality information, treatment data, and local and State law enforcement data. Additional sources of information include criminal justice, correctional, medical and community health data, local and State survey information, and research findings from ethnographic studies. The following are highlights from the meetings:

• Cocaine - Crack cocaine continues to dominate as the Nation's primary illicit drug problem, but indicator data show a leveling off in many urban areas: cocaine-related deaths were stable or up slightly in 6 of the 8 areas where such information was reported; emergency department (ED) mentions increased* in only 3 of the 20 CEWG cities in the Drug Abuse Warning Network (DAWN); the percentage of treatment admissions for primary cocaine problems declined slightly or remained stable in 13 of the 15 areas where data were available; and prices remained stable or declined slightly in most areas. Supplies remain abundant in nearly every city. Cocaine continues to be frequently used in combination with other drugs in some cities, including heroin and marijuana in Atlanta and Philadelphia, and methamphetamine in Denver. Demographic data continue to show most cocaine users as older, inner-city crack addicts; only in Miami were any new using populations reported this period. [* DAWN comparisons are for 1994 versus 1996; they are included only when they are reliable at p<0.05.]
  
  
• Heroin - Heroin overshadows cocaine in some indicators: it was the top-ranking ED mention in three cities, and it was the most common primary drug of abuse among treatment admissions in six areas. The rate of ED mentions per 100,000 increased significantly* in eight cities. Mortality, treatment, arrestee urinalysis, and price/purity figures were mixed, with increases in some cities, declines in others, and stable trends in others. In six CEWG cities, noteworthy percentages (˜10 percent) of arrestees, both male and female, tested heroin-positive in the Alcohol and Drug Abuse Monitoring (ADAM) program.
  
  Injecting remains the most common route of heroin administrationÑparticularly in the West, but also in some eastern and midwestern cities. However, it is declining sharply in some cities. Snorting predominates in several eastern and midwestern cities. Smoking is increasing among treatment admissions in Hawaii, San Diego, San Francisco, Atlanta, and Denver. Younger heroin users tend to snort or smoke the drug. Overall, heroin users tend to be males, older than 35, but in some cities, qualitative and quantitative dataÐincluding mortality figures in Miami--indicate increases among those age 18Ð25. Those age 17 and younger still account for relatively few ED mentions or treatment admissions, but even the low numbers, in areas such as Dallas, Minneapolis/St. Paul, San Francisco, and Texas, are a cause for serious concern.
  
  
• Marijuana - Continuing the upward trend in many cities since 1992, marijuana ED rates increased significantly* in 9 CEWG cities and primary marijuana abuse as a percentage of treatment admissions increased slightly to moderately in at least 10 reporting areas since the previous reporting period. Marijuana is now the top-ranking primary drug in at least four areas, and it accounts for substantial proportions (>20 percent) of nonalcohol admissions in at least three others. Among adult male arrestees, marijuana now exceeds cocaine as the most frequently detected drug in eight of the CEWG cities in ADAM; moreover, in each of the seven CEWG cities where ADAM tests juvenile males, the percentage of positive urinalyses is much higher for juveniles than for adults. Treatment demographics have similarly become increasingly youth dominated: the <17 age group now accounts for the largest percentages of marijuana admissions in at least eight areas. That age group also accounted for 20 percent or more of marijuana ED mentions in seven of the CEWG cities in 1996. All age groups, however, are substantially represented in all indicators, both qualitative and quantitative. Youth often consume marijuana with malt liquor. In some cities, joints or blunts are also dipped in PCP, codeine cough syrup, or embalming fluid; sometimes they contain crack or cocaine HCl, or they are laced with heroin.
  
  
• Stimulants - Except for ED mentions, methamphetamine indicators--mortality, treatment, and arrestee urinalysis--and ethnographic research show increases in the West, where the problem has been historically centered. Recent ADAM data show increases** in all the western CEWG cities. San Diego mortality figures also show increases, as do treatment figures in Denver, Los Angeles, and San Francisco. Methamphetamine is the most common primary drug among treatment admissions in San Diego and Hawaii, and it equals heroin as the number-one drug in Arizona. By contrast, ED rates declined significantly by 30-40 percent* throughout the West. They also declined* in Philadelphia, but the numbers are much lower there. The only city with a significant increase* was Minneapolis/St. Paul, but, again, the numbers are sharply lower than in the West. Smoking has recently overtaken inhalation as the primary route of administration in San Diego and Los Angeles. "Ice" smoking also predominates in Hawaii, and it is increasing in Denver and San Francisco. Injecting, however, still predominates in Denver, San Francisco, and Texas. In San Francisco, methamphetamine remains widespread among the gay and club scenes. Denver drug dealers sometimes cut crack or heroin with methamphetamine. Elsewhere in the country, methamphetamine appears in indicators in Minneapolis/St. Paul and in the rural areas surrounding St. Louis; availability is reported in Atlanta and New Orleans; and it is associated with the club or rave scenes in Boston, Baltimore, Miami, and New York City. [**ADAM comparisons are for first half of 1996 versus first half of 1997.]
  
  Methylenedioxymethamphetamine (MDMA or "ecstasy") availability is reported in 12 CEWG areas, primarily as a club drug at raves and dance parties. Increases are reported in Boston and Miami. Ephedrine-based products remain a major concern, with products such as "herbal ecstasy" widely available at convenience stores and truck stops in many CEWG areas, including Minneapolis/St. Paul and Arizona. Methylphenidate (Ritalin) abuse is reported among school-aged adolescents in Boston and Washington, DC, and it is the drug of choice for some stimulant users in Chicago. Seizures of khat, a flowering evergreen shrub also known as "qat" or "Somali tea," continue in Minneapolis/St. Paul.
  
  
• Depressants - Gamma-hydroxybutyrate (GHB) has been involved in poison control cases in Boston, Miami, and Texas, and is suspected in deaths in Miami; it is also part of the club scene (but not reported as a significant problem) in Baltimore, Honolulu, New York City, and areas of New Jersey; in Atlanta, it has become common as a synthetic steroid. Another club drug, ketamine ("Special K" or "Vitamin K"), is available in Boston (where youth both smoke and inject it), Honolulu, Miami, Minneapolis/St. Paul (where adolescents and young adults snort it and sometimes sprinkle it on tobacco or marijuana), New York City (where recent legislation has classified it as a controlled substance), and Washington, DC. Flunetrazepam (Rohypnol) availability has been sharply curtailed in Miami since State and Federal legislative measures were enacted. In Texas, however, it continues to be reported by treatment admissions, especially youth, in border areas. It continues to be reported in "date rape" incidents in Atlanta, Minnesota (where large quantities have been seized), and the Washington, DC, area; and it is used as a club drug in Atlanta and Honolulu. It has received media attention in some cities, such as San Diego and Seattle, but is not a widespread problem there. Clonazepam (marketed as Klonopin in the United States and Rivotril in Mexico), is sold and abused as flunetrazepam in Miami; on the Texas-Mexico border, juveniles widely use it in combination with beer, just as they had used flunetrazepam before the import ban. Opiate addicts use it to enhance the effects of methadone in Atlanta, Boston, and Minneapolis/St. Paul (where availability has declined).
  
  
• Hallucinogens - Lysergic acid diethylamide (LSD) is reportedly available in many CEWG cities and prices have become somewhat lower. However, indicators are declining in most areas. Rates of LSD-related ED mentions declined* in every CEWG area with the exception of San Diego and New Orleans. Similarly, phencyclidine (PCP)-related ED mentions declined* in six CEWG cities, but rose slightly in three. Primary hallucinogen users generally continue to constitute small percentages of total treatment admissions. Most hallucinogen users are young, suburban and middle class. In numerous areas, such as Chicago, Philadelphia, St. Louis and Texas, PCP is frequently used in combination with other drugs, primarily marijuana. In Boston and Seattle, LSD and other hallucinogens are often regarded as club drugs and used in the rave scene.

The Emergence of Narcotic Addict Types and Prediction of Addiction Career Outcomes

Based on the retrospective self-reports about activities, perceptions, and experiences in early adolescence (12-14 years), researchers classified 255 narcotic addicts into four distinct groups using risk factor information from five descriptive domains: family, peer deviance, personal deviance, psychological status, and protective factors. The four groups (Marginal Deviants, Socially Damaged, Early Drug Abusers, and Behavioral Deviants) were differentiated primarily by the extent of early drug use, other behavioral deviance, and family dysfunction. The predictive utility of the typology was examined in terms of outcome over the first 10 years of the addiction career, including age at first narcotic addiction, amount of time incarcerated, and percentage of time addicted while in the community. Nine of 10 risk factors were statistically significant in differentiating addicts from nonaddicts. Narcotic addicts, as adolescents, were more likely to be characterized by greater family, peer, personal, and emotional difficulties and to be less protected from negative influences in their lives than were never-addicted individuals from the same neighborhood. The extent of preaddiction deviance during adolescence was the key dimension in differentiating types in relation to outcome. And, although family dysfunction was a prominent feature in classifying individuals who were comparatively low in deviance, this characteristic added little in the way of predictive relevance to the outcome measures examined. Nurco, D. N., et al. The Early Emergence of Narcotic Addict Types. Am. J. Drug Alcohol Abuse, 23(4), pp. 523-542, 1997.

Precursors and Consequences of Gang Membership and Delinquency

Social scientific knowledge about youth gangs and how their members differ from nongang youths is limited and controversial. Knowledge of the etiology of gangs and the temporal relationship between gang membership and criminal behavior is especially limited. This study examines the relationship between gang membership and delinquency using self-report data from a survey of youths. This model-based study was conducted to explain gang membership in a population of 11,000 secondary school students in Colorado. Three models that can account for gang membership were tested: selection, social facilitation, or enhancement. Lower self-esteem, lower perceived academic ability, lower psychosocial health, and weaker bonds with institutions preceded gang membership which supports a selection model. Furthermore, greater drug use, greater delinquency, greater fear of harm, and being armed were both precursors and consequences of gang membership (facilitation and selection models). "Wannabes" were positioned half way between nonmembers and members and had the lowest level of self-esteem of any group. Being in a gang actually seemed to improve the self-esteem of the members, at least initially. Findings were consistent with theories that gang membership results from a lack of social integration. Dukes, R.L., Martinez, R.O., and Stein, J.A. Youth & Society, 29, pp. 139-165, 1997.

Alcohol Expectancies as Potential Mediators of Parent Alcoholism Effects on the Development of Adolescent Heavy Drinking

Children of Alcoholics (COAs) are at heightened risk for early onset of alcohol use, for heavy drinking, and for alcohol-related impairment. However, little is known about the mediating mechanisms that account for this risk. In a study of adolescents' expectancies about alcohol effects as potential mediators of their growth over time in heavy drinking, participants (n=428) were from an ongoing longitudinal study of parent alcoholism and data were collected in three annual in-person interviews with adolescents and their parents. Latent growth curve modeling analyses showed that COAs started the study at higher levels of heavy drinking and increased their heavy drinking at a steeper rate than did non-COAs over the course of the study. However, expectancies about alcohol effects could not explain these parent alcoholism effects. Although expectancies were related to adolescents' initial drinking levels, they did not prospectively predict growth over time in adolescent alcohol use. Thus, other mediating mechanisms (e.g., personality, parent and peer socialization) are necessary to explain parent alcoholism risk for escalation in heavy drinking during adolescence. Colder, C.R., Chassin, L., Stice, E., and Curran, P. Journal of Research on Adolescence, 7, pp. 359-374, 1997.

Sequence of Drug Use among Serious Drug Users: Typical vs Atypical Progression

The "typical" sequence of drug use onset identified in previous studies (alcohol, then marijuana, followed by other illicit drugs) was found to be much less prevalent among serious drug users than in samples of high school students. Sequence of drug use was examined in a secondary analysis of two samples of serious drug users: one of 152 men and one of 133 women. The data were collected in ethnographic studies of adult drug users and distributors in New York City between 1984 and 1987. The proportions of drug users following specified patterns of drug use onset were compared to proportions obtained in previous research in samples of high school youth, and serious drug users. Only 33% of the serious drug users followed the typical sequence, compared to 75-93% of subjects in previous studies of high school students. The serious drug users were more likely to have used marijuana before using alcohol, and more likely to have used other illicit drugs before using marijuana. In addition, atypical sequencing was associated with earlier initiation of the use of illicit drugs other than marijuana and greater lifetime drug involvement. These findings suggest that for a large number of serious drug users, marijuana does not play the role of a gateway drug. Prevention efforts that focus on alcohol and marijuana may be of limited effectiveness for youth who are at risk for serious drug abuse. Mackesy-Amiti, M.E., Fendrich, M., and Goldstein, P.J. Drug and Alcohol Dependence, 45, pp. 185-196, 1997.

Violence, Suicidality, and Alcohol/Drug Use Involvement in Adolescent Females with a Psychoactive Substance Use Disorder and Controls

This study had three aims: (1) to determine the relationship between behavioral dysregulation, negative affectivity, and familial impairment with violence and suicidality (i.e., severity of ideation and attempts) in a sample of adolescent females with a psychoactive substance use disorder and controls; (2) to determine whether these relations are mediated by internalizing (depression/anxiety) and externalizing (nonviolent antisocial behavior) symptomatology; and (3) to determine whether severity of alcohol/drug use involvement moderates the relations between the mediating variables with violence and suicidality. Multiple behavioral, psychiatric interview, and self-report measures were used to index behavioral dysregulation, negative affectivity, familial impairment, internalizing and externalizing symptomatology, and violence and suicidality in 161 14- to 18- year old adolescent females with a psychoactive substance use disorder and in 80 controls. Structural equation modeling was used to determine the proposed relations. Results indicated that behavioral dysregulation, negative affectivity, and familial impairment were related to violence, whereas only familial impairment was related to suicidality. Internalizing symptomatology mediated the relation between familial impairment and suicidality, and was related to violence, where as externalizing symptomatology mediated the relations between behavioral dysregulation, negative affectivity, and familial impairment with violence. Severity of alcohol/drug use involvement did not moderate the relations between internalizing or externalizing symptomatology with suicidality or violence. Nevertheless, the relation between internalizing symptomatology and suicidality was stronger in females with a greater degree of alcohol/drug use involvement, compared with those with a milder degree of involvement. Therefore, from a prevention standpoint, behavioral dysregulation, negative affectivity, familial impairment, as well as internalizing and externalizing symptoms, may serve as clinical "points of intervention" for altering the development of violence and suicidality in high risk and substance abusing youth. Mezzich, A.C., Giancola, P.R., Tarter, R.E., Lu, S., Parks, S.M., and Barrett, C.M. Alcoholism Clinical and Experimental Research, 21 (7), pp. 1300-1307, 1997.

Associations Between Attention Deficit Hyperactivity Disorder (ADHD) and Psychoactive Substance Use Disorders (PSUD)

In a sample of siblings of ADHD and non-ADHD probands, ADHD, conduct disorder, anxiety disorders, and male gender in the sibling were associated with higher rates of PSUD. There was also a statistical interaction between ADHD and conduct disorder in which nearly all siblings with both disorders had PSUD and a particularly early onset of that disorder. Using DSM-III-R structured diagnostic interviews and blind raters, a four-year follow-up of siblings from ADHD (N=152) and control families (N=117) was conducted. The mean age of the siblings was 17.3 (range 9 to 41) and roughly half were males. Information on PSUD was obtained in a standardized manner blind to the proband's clinical status. Cox proportional hazards models evaluated baseline diagnoses of ADHD and other psychiatric disorders (conduct, anxiety, and mood) as potential predictors of PSUD. These Cox analyses simultaneously controlled for confounding variables such as age, gender, intactness of family, and SES. These findings linking ADHD and PSUD could have significant public health significance. Since ADHD is a prevalent, childhood-onset disorder that is often characterized by impulsive behavior, it could represent a large group of youth at high risk for illicit drug use. Considering that the earliest age at onset of PSUD in our sample was 12 years and that the latest age at onset of ADHD is 7 years, there may be a window of several years where prevention and early intervention efforts could be targeted. Milberger, S., Biederman, J., Faraone, S.V., Wilens, T., and Chu, M.P. American Journal on Addictions, 6, pp. 318-329, 1997.

A Test of Socioeconomic Status As A Predictor of Initial Marijuana Use

The current study examined the effects of socioeconomic status (SES) on adolescent marijuana use using data from a national longitudinal survey of subjects aged 11 to 17 (N = 1,725). Both direct and indirect measures of SES (e.g., Hollingshead's measure of SES, urbanicity, neighborhood problems) were used to determine the extent to which SES predicts marijuana use among adolescents. For males, this study found a nonlinear relationship between the Hollingshead measure and marijuana use that had not been identified in previous research. For females, the Hollingshead measure was insignificant when alcohol use and having friends who use marijuana were added to the model. This finding suggests that the effects of SES on marijuana use are mediated by alcohol use and friends' use of marijuana. Weekly alcohol users were much more likely than nonusers to initiate marijuana use for both males and females. Other significant variables for both sexes included having a job, having friends who use marijuana, and having used some alcohol in the past year. For males, grade point average (GPA), commitment to friends, urbanicity, time spent with friends, and peer strain were also significant predictors of initial marijuana use. For females, prior victimization and low school aspirations were significant. In sum, these findings suggest that psychosocial risk factors for marijuana use are substantially different for males than for females and that future researchers need to test for nonlinear relationships between SES and adolescent substance use. Miller, D.S., and Miller, T.Q. Addictive Behaviors, 22 (4), pp. 479-489.

Risk and Protective Factors as Predictors of Adolescent Alcohol Involvement and Transitions in Alcohol Use: A Prospective Analysis

Determinants of initial alcohol use may differ from predictors of accelerated or problematic consumption. Social influences may be strong predictors of initial drinking; however, later stages of problem drinking may be linked developmentally to intrapersonal deficits. This study prospectively examined the influence of chronic and changing risk and protective status in predicting adolescent alcohol involvement and transitions in alcohol use. Data were obtained from a three-wave cohort (N=823) of 8th- 10th grade nonintervention students participating in a school-based drug abuse prevention trial. Cognitive, attitudinal and social influence measures were dichotomized using empirical cut-offs to designate risk or protective status. Using a conceptually based assignment scheme, additive risk indices were created assessing chronic (averaging across time) and changing features of competence, psychological and interpersonal functioning, cognitive-affective and social influences. Three chronic and change protective indices were created tapping competence, psychological, and interpersonal functioning. Controlling for initial drinking and gender, chronic risk for social influence and psychological functioning and increased risk for social influences and competency predicted subsequent drinking behavior. Chronic psychological protection attenuated subsequent drinking. Using categorical measures of drinking behavior to designate nonuse, experimental or moderate-heavy use, chronic social influence and competency risk were associated with an increased likelihood of accelerated drinking, whereas improved psychological functioning diminished the likelihood of increased drinking behavior. Findings underscore the need for implementing prevention strategies that reinforce developmentally appropriate skills and enhance personal competence and psychological functioning as effective barriers against initial and more problematic alcohol use. The unique contribution of protective forces also underscores that risk reduction and protection enhancement are complementary processes and are both required to offset social influences for alcohol consumption. Scheier, L.M., Botvin, G.J., and Baker, E. Journal of Studies Alcohol, 58, pp. 652-667, 1997.

Predicting Dropping Out of High-School Based on Data from a 19-Year Longitudinal Study of Family Lifestyles

In this study, nested latent-variable causal models were contrasted to compare the direct and indirect relationships of distal family and child and proximal adolescent factors to dropping out of high school. The sample included 194 conventional and nonconventional families that have been participating in a 19-year longitudinal study, the Family Lifestyles Project. The findings showed that dropping out of high school is a multiply-determined process, with early influences beginning in childhood, that involves family as well as child and adolescent factors. Among the hypothesized predictors of dropping out were early child drug exposure and adolescent drug use. Early child drug exposure by adult models directly predicted later adolescent drug use which in turn predicted more high school dropout. The indirect effect of early drug exposure on later dropping out was also significant. Early family nonconventionality when combined with a higher commitment to lifestyle values was associated with a lower probability of dropping out. Garnier, H.E., Stein, J.A., and Jacobs, J. American Educational Research Journal, 34, pp. 395-419, 1997.

Intra-Family Substance Use

Analyses that specifically model the within-level and between-level covariance matrices are used to examine family levels of substance use and predictors of family substance use. Data from 1135 participants and 430 families (430 target adolescents, 137 siblings, and 568 parents) were included in the analyses, with family cluster sizes ranging from two to five. Adolescents had a mean age of 16.15 years. Between-level variation in substance use scores represented approximately 38% of the total variation in substance use scores, illustrating the importance of a common environment, such as the family, in accounting for levels of substance use. Contextual variables, or family-level predictors included marital status, SES, and biological relatedness. Intra-class correlation coefficients for biologically related families were higher for alcohol, cigarettes, marijuana, and illicit drugs, than they were for non-biologically related families with all three family-level variables significantly predicting the between-level substance use variance. In general, those families characterized as more economically disadvantaged, of a single parent nature, and non-biologically related, had higher family-levels of substance use. Duncan, T.E., Duncan, S.C., Hops, H., and Alpert, A. Drug and Alcohol Dependence, 46, pp. 167-180, 1997.

Latent Variable Modeling of Longitudinal and Multilevel Substance Use Data

Multilevel Latent Growth Modeling (MLGM) was used to analyze longitudinal and multilevel data for adolescent and parent substance use measured at four annual time points. An associative LGM model was tested for alcohol, cigarette and marijuana use with a sample of 435 families (435 target adolescents, 203 siblings, and 566 parents). Adolescents (targets and siblings) had a mean age at T1 of 13.69 years. Sufficient homogeneity in substance use within families was found, with the proportion of between variation, or intraclass correlation, for the four repeated measures, ranging from .26 to .30 for alcohol, .26 to .29 for cigarettes, and .28 to .33 for marijuana. Results revealed a significant upward trend in the development of alcohol, cigarette and marijuana use among families. The shapes of the family-level developmental curves for each substance were also strongly related. Younger family members had higher initial levels of use and developed in their use of substances at a faster rate than older family members. Males had higher levels of alcohol use than females. Single and step-parent families had higher family-levels of substance use, whereas families characterized as less educated and more economically disadvantaged had higher levels of cigarette use and developed in their use of marijuana at a greater rate. Duncan, T.E., Duncan, S.C., Alpert, A., Hops, H., Stoolmiller, M., and Muth, N. B. Latent Variable Modeling of Longitudinal and Multilevel Substance Use Data. Multivariate Behavioral Research, 32, pp. 275-318, 1997.

Latent Transition Analysis Shows Caffeine Use is Correlated to Initiation of Drug Use

This chapter introduces a relatively new methodology, latent transition analysis (LTA) and demonstrates its usefulness in substance abuse prevention research. LTA can be used to evaluate various aspects of stage-sequential models. It can help the user to assess whether a hypothesized set of stages is a realistic representation of a data set, and can be used to compare different models of the stage transition process. LTA is a latent variable model for longitudinal panel data.. It was measured in 7th grade and again in 8th grade students. Using a cross-validation procedure developed for LTA, the model accepted was an 8-stage model: "no use", "alcohol only", "tobacco only", "alcohol + tobacco", "alcohol + drunk", "alcohol, drunk, advanced", "alcohol, tobacco, advanced", and "alcohol, tobacco, drunk, advanced". In this study, caffeine risk was defined as low through high use. Results showed that those in the "high risk" group based on caffeine use at 7th grade, by any definition of caffeine risk, were 1.5 to 2.5 times less likely than those in the "low-risk" group to be in the "no use" stage in 7th grade. Those in the "high risk" group were also 2.4 to 3.9 times more likely than those in the "low risk" group to be in the highest drug use stage at 7th grade. Finally, those in the "high risk" group were 1.2 to 1.9 times more likely than those in the "low risk" group to transition out of the "no use" drug stage between 7th and 8th grades. The results of this study suggest that high use of caffeine is associated with an increased probability that an individual will initiate the onset process. Collins, L.M., Graham, J.W., Rousculp, S.S., & Hansen, W.B. Heavy Caffeine Use and the Beginning of the Substance Use Onset Process: An Illustration of Latent Transition Analysis. In K. Bryant, M. Windle, and S. G. West (Eds.) The Science of Prevention: Methodological Advances from Alcohol and Substance Abuse Research. pp. 79-99. Washington, D.C.: American Psychological Association, 1997.

Young Adults' Drug Use: A 17 Year Longitudinal Inquiry of Antecedents

This longitudinal study examined the interrelation of personality and peer factors on young adults' drug use. This research also examined the influence of the interaction of personality and peer factors on drug use. The samples of 756 males and females was interviewed four times between the mean ages of 6 and 22. Personality attributes in childhood were related to peer factors in early adolescence which, in turn, were related to personality traits in later adolescence. These traits were linked with the selection of peers and, ultimately, drug use in young adulthood. Additionally, the adolescent and young adult domains had direct effects on young adults' drug use. Significant interactions indicated that a few protective childhood personality traits buffered the risk factor of the adverse effects of deviant peers in adolescence on young adults' drug use. The presence of protective characteristics from one domain in childhood enhanced the preventive effect of later protective traits from the other domain in young adulthood on drug use. Brook, J.S., Balka, E.B., Gursen, M., Brook, D.W., Shapiro, J., Cohen, P. Psychological Reports, 1997.