Epidemiology of Youth Drug Abuse

Research Findings from September, 1998 Director's Report

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.

Amphetamine, Substance Use and Its Relationship to Depression, Anxiety, and Isolation Among Youth Living With HIV

In a series of three papers, drug use and correlates were examined in youth living with HIV (YLH). Amphetamine use, other HIV-related risk acts, T-cell counts, emotional distress, coping style, and symptoms of HIV were examined in 337 HIV+ youth aged 13 to 24 (20% female; 22% African American, 27% Anglo, 35% Latino) from four cities (LA, NY, San Francisco, and Miami). One third of youth were found to have engaged in amphetamine use in their lifetime, and 21% of youths reported current use (i.e. in the last three months). Compared to non-users, users initiated other drug use at younger ages, used more types of drugs, reported more emotional distress, and employed escape coping significantly more often. Substance use pervaded the lives of these YLH. Among this sample, males had used more drugs, more often, and for longer periods than females. However, there had been major reductions in use. Being male, having high emotional distress, and having fewer negative social supports were significantly associated with greater reductions in substance use. The longer an individual had been diagnosed seropositive tended to be associated with reductions in use (p=.06). Compared to non-users, users also had more sexual partners and more sexual encounters. While users and non-users do not differ on physical symptoms or whether they have been diagnosed with AIDS, users of amphetamines report significantly higher T-cell counts than non-users. Despite poor psychosocial functioning, amphetamine users were found to have higher T-cell counts than other YLH. The continued high-risk profile of transmission acts among users suggests that preventive interventions must target specific drugs used by YLH. As the number of youths infected with HIV rises, secondary prevention programs are needed to help youths living with HIV meet three goals: 1) increase self-care behaviors, medical adherence, and health-related interactions; 2) reduce transmission acts; and 3) enhance their quality of life. Based on modifications of the social action model, a small group intervention was developed and tested in this population. Rotheram-Borus, M.J., Mann, T., and Chabon, B. Amphetamine Use Among Youths Living with HIV. AIDS Education and Prevention, In Press; Rotheram-Borus, M.J., Murphy, D.A., Swendeman, D., Chao, B., Chabon, B., Zhou, S., and Birnbaum, J. Substance Use and its Relationship to Depression, Anxiety, and Isolation Among Youth Living with HIV. International Journal of Behavioral Medicine, In Press; Rotheram-Borus, M.J., and Miller, S. Secondary Prevention for Youths Living with HIV. AIDS Care, 10(1), pp. 17-34. 1998.

Differential Cognitive Functioning in 9-12 Years Olds Relative to Prenatal Cigarette and Marijuana Exposure

In an examination of cognitive performance of 131 9-12 year-old children participating in a Carleton University longitudinal study since birth, discriminant function analysis indicated a dose-dependent association between higher prenatal cigarette exposure in utero and lower performance on global intelligence test scores, with the verbal subtests of the intelligence test discriminating maximally among levels of in utero exposure. In contrast, prenatal marijuana exposure was not associated with global intelligence or the verbal subtests, but rather was negatively related to executive function tasks that require impulse control and visual analysis/hypothesis testing, and with a number of the intelligence subtests requiring these same abilities. The cigarette results extend observations made in this sample and others at earlier ages. The marijuana findings, combined with results observed at earlier ages, lead the authors to suggest that in utero exposure to marijuana may have a negative impact on aspects of neurocognitive competence that fall under the domain of executive function. Fried, P.A., Watkinson, B.M. and Gray, R. Neurotoxicology and Teratology, 20 (3), pp. 293-306, 1998.

Community Epidemiology Work Group

In the Past 6 Months

• Heroin indicators have shown increases in many cities, in some cases overshadowing cocaine.
• Marijuana indicators have continued to escalate across the country.
• Cocaine indicators have continued to level or decline, except for some isolated potentially emerging problems (in Miami and Texas).
• Methamphetamine indicators have increased in most western areas, following some declines last year.
• "Club drugs" have continued to proliferate in the drug scene in several areas.

The 44th meeting of the Community Epidemiology Work Group (CEWG) was held in Boston, Massachusetts on June 23-26, 1998.

The CEWG is a network of researchers from 21 U.S. metropolitan areas and selected foreign countries 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. The following are highlights of the meeting:

• Cocaine - While crack cocaine remains the Nation's predominant illicit drug problem, indicator data continue to show leveling off in many urban areas: cocaine-related deaths were down or stable* in 7 of the 8 areas where such information was reported (Philadelphia was the exception); 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 or remained stable* in 14 of the 16 areas where such data were available (Newark and New Orleans had slight increases); cocaine-positive urinalysis percentages declined or remained stable in all 15 CEWG cities in the Arrestee Drug Abuse Monitoring (ADAM) program; and prices and purity remained stable in most areas. Supplies remain abundant in nearly every city. Demographic data continue to show most cocaine users as older, inner-city crack addicts, with two exceptions: in Miami, poison center, school survey, and anecdotal data continue to indicate some adolescents initiating cocaine use in combination with other drugs; and in some Texas communities, white and Hispanic street youth are reportedly using crack as an alternative to poor methamphetamine quality. Cocaine continues to be used in combination with other drugs, such as marijuana in Atlanta, Chicago, Miami, and Philadelphia; heroin in Atlanta, Boston, Philadelphia, St. Louis, and Seattle; and methamphetamine in Denver and parts of Texas. In some areas, such as Boston and Denver, users dissolve crack for injection. Some shifts from crack to cocaine hydrochloride (HCl) are reported: among treatment admissions in Atlanta and Texas; and in street-level sales in New York. Cocaine is reported to be involved in 50 percent of homicides in New Orleans.
• Heroin - Indicators in many cities continue to show increases: heroin-related deaths increased in 6 of the 8 cities where mortality was reported (they remained stable in Honolulu and declined in Seattle); ED mentions increased significantly* in 8 cities; and treatment percentages increased* in 10 of the 16 areas where trend data were available. Heroin now overshadows cocaine in some cities: it ranks first among ED drug mentions in three cities, and it is the primary drug of abuse among treatment admissions in six areas. More than 10 percent of male arrestees tested heroin-positive in six CEWG cities.

  Heroin continues to increase among new and young users in a number of cities because of its easy availability, low price, high purity, and favorable reputation compared with crack. Indicator and anecdotal data suggest increases among young adults in Boston, Chicago, Denver, and San Francisco and among both youth and young adults in Baltimore, Miami, Minneapolis/St. Paul, Philadelphia, San Diego, and parts of Texas. The suburbs are increasingly mentioned: in Baltimore, young white professionals, laborers, and high school students from the suburbs are being noted in the inner city buying drugs; in some Texas suburbs, youth and young adults have been involved in overdose deaths; and a recent Chicago study of injecting drug users included a large proportion of suburban residents. Younger heroin users tend to snort the drug. The percentage of snorters among treatment admissions continues to increase in several cities, including Atlanta, Baltimore, Detroit, Minneapolis/St. Paul, Philadelphia, and San Diego. Atlanta ethnographic reports continue to indicate an increasing number of recently initiated snorters shifting to injection; in New York City, by contrast, snorters do not seem to be shifting to injection, except for a population of young immigrants from the former Soviet Union. Smoking remains relatively rare, but it is reported by 7 percent of admissions in both Miami and San Diego. Young adults in Phoenix reportedly use heroin to "come down" from methamphetamine-induced highs. In San Francisco, heroin-marijuana combinations are sold as "canade." In New York City, many crack sellers are switching to heroin sales due to the high profit potential.

  *mortality, treatment, and ADAM comparisons are for 1996 versus 1997; DAWN comparisons are for 1995 versus 1996, reliable at p<0.05.

• Marijuana - Marijuana accounted for more than 10 percent of total ED mentions in four cities, and it was the primary drug of abuse among treatment admissions in Denver, Minneapolis, and Seattle. Treatment admissions remained at elevated levels: percentages increased in five cities, remained stable in nine, and decreased in one. Among adult male arrestees, marijuana-positive findings exceeded cocaine-positives in seven ADAM cities. Despite declines in nine cities between 1996 and 1997, marijuana levels in this population were generally higher than in 1995. Marijuana remained readily available in many CEWG areas, although it was expensive and hard to find in San Francisco. Seizures increased in Boston, Detroit, and Hawaii, while marijuana-related arrests increased in New York City and Washington, DC. Qualitative and quantitative data from most CEWG cities indicate widespread marijuana use among youth. For example, among primary marijuana treatment admissions, the average age at first use was 13.9 in Minneapolis and 14 in Denver. In Washington, DC, the number of marijuana treatment admissions in the 12-17 age group increased 70 percent between 1995 and 1997. In each of the seven cities where ADAM tests juvenile males, the percentage of positive urinalyses was much higher for juveniles than for adults. While alcohol remains the most common secondary and tertiary drug among marijuana treatment admissions, other drug combinations continue to be reported. Blunts are combined with cocaine HCl in Atlanta and Philadelphia ("turbo") and with crack in Chicago ("diablito," "primo," or "3750,") and Miami ("geek joint"). Other combinations include phencyclidine (PCP) in Chicago ("wicky stick" or "donk") and Philadelphia ("love boat" or "dust blunt") and embalming fluid in Minneapolis. In Texas, blunts are dipped in embalming fluid laced with PCP ("fry," "amp," or "water-water") or in codeine cough syrup ("candy blunts").
• Stimulants - Indicators of methamphetamine use - mortality, treatment, and arrestee urinalysis-and ethnographic research continue to show increases in the West. After declining in 1996, mortality figures in San Diego, Hawaii, and Phoenix have risen. Similarly, the 1997 ADAM data indicate increases in all the western CEWG cities, returning the rates of arrestees testing positive to 1994-95 levels. Treatment figures show increases in all reporting western cities, except for Seattle, where they have stabilized, and Phoenix, where they have declined since 1994. Methamphetamine remains the most common primary drug among treatment admissions in Hawaii and San Diego. Smoking remains the primary route of administration for methamphetamine in San Diego and for "ice" in Hawaii, and it has become more common in Denver and San Francisco, although injecting still predominates. Inhalation still predominates in Los Angeles. In San Francisco, "speed" use is increasing among blue collar workers, young professionals, and college students. In Denver, where methamphetamine is often used with crack, some former crack users may have crossed over to exclusive methamphetamine use. Elsewhere in the country, methamphetamine appears in indicators in Minneapolis/St. Paul, Philadelphia, and parts of Texas. The number of treatment admissions has increased in Baltimore, Boston, Detroit, New Orleans, and St. Louis. In Missouri, statewide meth-amphetamine admissions began to outnumber heroin admissions in 1996. State police report increased seizures of the drug in and near Boston, and it is associated with the club or rave scenes in Atlanta, Miami, and New York City.

  Methylphenidate (Ritalin) abuse is reported, mostly among school-aged adolescents, in Boston, Detroit, Minneapolis/St. Paul, Phoenix, Seattle, Washington, DC, and areas of Texas; in Chicago, it is the drug of choice for some stimulant users or is mixed with heroin as a "speedball." Methylenedioxymethamphetamine (MDMA or "ecstasy") availability is reported in eight CEWG areas, primarily as a club drug at raves and dance parties. Increases are reported in Boston. Ephedrine-based products remain a concern, with products such as "herbal ecstasy" widely available at convenience stores and truck stops in many CEWG areas, including Atlanta, Minneapolis/St. Paul, and Phoenix; in Texas, at least eight deaths have been associated with ephedrine since 1993. Seizures of khat, a flowering evergreen shrub also known as "qat" or "Somali tea," continue in Detroit and Minneapolis/St. Paul. Fenfluramine and phentermine (Fen-Phen) are reportedly brought into Texas via legal prescriptions from Mexico.

• Depressants - Gamma-hydroxy butyrate (GHB), referred to as a "stove-top drug" in Phoenix because of its ease of manufacture, has been involved in poisonings and over-doses in Boston, Detroit, Miami, and New Orleans. It is also part of the club scene in Baltimore, San Francisco, and Texas, and is popular as a synthetic steroid in Atlanta. Another club drug, ketamine ("Special K" or "vitamin K"), was involved in overdose cases in New Orleans. Some ketamine availability was reported in Boston (where police report it is becoming more prevalent), Detroit, Miami (where it was involved in DUI cases along with other drugs), Minneapolis, Seattle, and Washington, DC. Flunitrazepam (Rohypnol) availability and use appears to have decreased substantially in some CEWG areas: Detroit (where hotline inquiries have declined since 1996), Miami (where it has been replaced by GHB combined with alcohol), and Texas (where submissions to the crime laboratory declined 71 percent between 1996 and 1997); however, seizures and anecdotal data in New Orleans indicate that distribution and abuse is increasing. Clonazepam (marketed as Klonopin in the United States and Rivotril in Mexico) is frequently used as a substitute for flunitrazepam in Texas; in Atlanta it is used to enhance the effects of methadone; and police in Boston report it frequently accompanies heroin use. Benzodiazepines such as diazepam are appearing in crack houses in Atlanta and are used by heroin and cocaine abusers in Miami to self-medicate for withdrawal symptoms. Treatment admissions for depressants in Chicago, while remaining small proportionally, have more than doubled recently.
• Hallucinogens - Contrary to declining ED and treatment indicators, ethnographic data indicate increased lysergic acid diethylamide (LSD) use among youth in Atlanta and Boston and increased phencyclidine (PCP) use in Chicago and San Francisco. LSD also remains popular among youth in Detroit, Miami, and Minneapolis. Treatment admissions, however, remain low for all hallucinogens. In Washington, DC, PCP treatment admissions decreased by more than 50 percent between 1995 and 1997 (1.4 percent of all admissions). In addition to LSD and PCP, psilocybin mushrooms were reportedly available in Boston, Minneapolis, and Seattle, while some availability of mescaline was reported in both Baltimore and Boston.
• Viagra - The potency pill sildenafil citrate (Viagra) may be appearing in the club scene. Los Angeles treatment officials describe "Viagra parties" at gay bars and report that three young men have died after combining the drug with "poppers," a recreational nitrate. Researchers are also concerned about two other areas of potential abuse: sale of the drug over the Internet; and a black market for the drug in other countries, such as Japan and the United Kingdom, where the drug is not approved for prescription sale.

Age of First Use: Its Reliability and Predictive Utility

In a study of the early-onset issue, researchers at Rutgers University posed three questions: (1) Is age of first licit use a predictor of differences in alcohol and drug use intensity during the period (age 20) when normative patterns of use reach a peak?, (2) Is age of first licit use a predictor of differences in use intensity in young adulthood (age 30) when most individuals have moderated their use? and (3) How consistent are adolescents in retrospectively recalling age of first use? Subjects (N=839) from the Rutgers Health and Human Development Project provided four waves of longitudinal data spanning the age range from 15 to 31. The vast majority of adolescents exhibited a sequential pattern of drug use initiation consistent with that found in previous studies. Retrospective recall of age of onset revealed a fair degree of relative agreement but a lack of absolute agreement; that is, as adolescents became older, recalled ages on onset exhibited an upward shift approximately equal across most individuals. Age of first licit use as recalled at age 18 did not predict differences in alcohol or drug use intensity at age 20. Age of first illicit use was a fairly strong predictor of drug use at 20 but a weak predictor of alcohol use at that age. Age of first licit use and age of first illicit use did not predict difference in useor use consequences at age 30. Regardless of age of onset, illicit drug use and heavier alcohol use constitute an adolescence-limited phenomenon for most individuals. The findings suggest that intervention efforts need to be aimed not only at delaying the onset of illicit use in adolescence but also at reducing use levels among young adult users by facilitating the maturing out process. Labouvie, E., Bates, M. E., and Pandina, R. J. Age of First Use: Its Reliability and Predictive Utility. Journal of Studies on Alcohol, 58(6), pp. 638-643, 1997.

The Impact of Maternal Drinking During and After Pregnancy on the Drinking of Adolescent Offspring

The impact of prenatal maternal drinking on alcohol consumption in adolescent offspring was examined among boys and girls separately. A prospective longitudinal sample of 185 mother-firstborn child dyads was used to examine the impact of maternal self-reported lifetime and current drinking, controlling for potential confounding factors. In this representative general population sample, maternal drinking during pregnancy, particularly continuous moderate to heavy consumption, had a significant positive effect on adolescent daughters' current drinking, but a slight negative effect on sons' lifetime drinking. The sex-specific prenatal effect on current drinking persisted with controls for prenatal maternal cigarette smoking, current maternal drinking, child-rearing practices (i.e. maternal child closeness, monitoring and a rule against drinking) and adolescent's problem behaviors in childhood. Prenatal maternal smoking was also associate with elevated rates of adolescent drinking, particularly current drinking. Of the child-rearing variables, only a rule against drinking decreased adolescent drinking. Thus, selected prenatal factors may constitute risks for alcohol consumption among adolescent daughters. Griesler, P.C., Kandel, D.B. Journal of Studies on Alcohol, 59(3), pp. 292-304, 1998.

Ethnic Identity has Protective Effect on Risks for Drug Use

Five year follow-up data were collected from structured interviews with 555 male and female Puerto Rican adolescents who were originally interviewed in 1990 in New York City. At time of follow-up, the average age of the respondents was 19. The interview data were analyzed to determine the relationship of multiple drug risks, Puerto Rican identity, and drug use. In addition, the risk/protective and protective/protective paradigms for examining interactive effects of ethnic identity and drug risks on drug use were assessed. Each risk and two ethnic variables were related to drug use. Regressions showed that cultural knowledge, being culturally active, group attachment, and identification with Puerto Ricans offset the impact of risks on drug use. Ethnic variables also enhanced the protective effect of other protective factors. The findings substantiate expanding risk-buffering models to include ethnic identity and the protective role of ethnic identity for Puerto Rican Youth. Brook, J.S., Whiteman, M., Balka, et al. Drug Use Among Puerto Ricans: Ethnic Identity as a Protective Factor. Hispanic Journal of Behavioral Sciences, 20(2), pp. 241-254, 1998.

Maternal Smoking in Pregnancy, Child Behavior Problems, and Adolescent Smoking

This study used a longitudinal sample of mother-child dyads to examine the role of child behavior problems in explaining the effect of maternal prenatal smoking on adolescent daughter's smoking. Maternal smoking during pregnancy is associated with higher levels of child behavior problems, particularly among girls. Childhood behavior problems increase the likelihood of lifetime smoking among daughters but do not explain the effect of prenatal maternal smoking on their current smoking. Maternal smoking in pregnancy, especially heavy use of a pack or more a day, retains a unique effect on girls' current smoking with controls for current maternal smoking, child behavior problems, and maternal monitoring of the child. The effect of maternal prenatal smoking is suggestive of a biological component, which may have direct or indirect influences on adolescent smoking. The small number of cases in the study calls for the replication of these findings in large samples that would incorporate prospective measures of prenatal nicotine exposure from mother and father and additional biological and psychosocial covariates. Griesler, P.C., Kandel, D.B., Davies, M. Journal of Research on Adolescence, 8(1), pp. 159-185, 1998.

Heavy Caffeine Use and the Beginning of the Substance Use Onset Process: An Illustration of Latent Transition Analysis

This chapter has two objectives, the introduction of a relatively new methodology, latent transition analysis (LTA) and the demonstration of its usefulness in alcohol prevention research. LTA is an extension of latent class theory that allows the user to estimate and test models of stage-sequential development. Researchers may be more accustomed to thinking in terms of strictly quantitative development, in which change can be characterized by increases and decreases in a particular variable, such as test scores, self-esteem, or amount of alcohol consumed per week. Stage-sequential development is distinguished from quantitative development by the involvement of qualitatively different stages. Individuals develop by passing through these stages. In this study, LTA was used to investigate whether heavy use of caffeine would be a predictor in the early part of the substance use onset process. Heavy use of caffeine was defined in several different ways, in terms of both coffee and cola soft drinks. By all the definitions of heavy caffeine use, those in the high-caffeine-risk latent class were more likely to have initiated the substance use onset process by seventh grade, and those who had not initiated by seventh grade were more likely to do so by eighth grade. The increased risk of onset related to caffeine use was greater for coffee than for cola. For coffee, the probability of onset was greater for the more extreme definitions of caffeine risk. The level of caffeine consumption required to elevate risk was surprisingly low. Increased risk was found for as little caffeine consumption as 6 cups of coffee in one's lifetime and 6 cola drinks in the past week. The results replicated well and consistently across two cross-validation subsamples. In K. Bryant, M. Windle, and S.West (Eds.), The Science of Prevention: Methodological Advances from Alcohol and Substance Abuse Research (pp. 79-99). Washington, DC: American Psychological Association, 1998.

Problem and Conventional Behaviors Among American Indian Adolescents

Investigators from the National Center for American Indian and Alaska Native Mental Health Research at the University of Colorado Health Sciences Center report two related factor analytic studies that take Problem-Behavior Theory as a starting point. They examined the latent structure of problem and positive behaviors in a sample of 1894 American Indian adolescents (n=1894 in the first study and 2250 in the second study) and found a 2-factor 2nd-order structure in which problem behaviors (alcohol use, drug use, antisocial behavior, risky sexual behavior) and positive behaviors (school success, cultural activities, competencies, community-mindedness) represented two relatively uncorrelated aspects of behavior. The positive behaviors construct contributed significant incremental construct validity in the prediction of psychosocial outcomes (e.g., depression, competencies), relative to the problem behavior construct alone. Factor structures differed across gender primarily for alcohol use and school success. The second-order alcohol use factor was more closely related to "problem drinking" for girls, but for boys it was more closely related to "negative consequences following drinking". "Doing schoolwork carefully" was more closely related to school success for girls than for boys, reflecting others' reports that girls tend to attribute their successes in school to their own hard work whereas boys tend to attribute their school successes to greater intellectual ability. Across four sampled communities, the structures differed only slightly, suggesting commonalities on such dimensions as positive behaviors, possibly from consistent messages adolescents receive about appropriate ways to act. Results highlight the need to include a focus on positive behaviors in prevention/promotion activities, and community members need to understand not only how to circumvent the processes that result in maladaptive outcomes but also how to promote the development of successful adolescents. Mitchell, C.M., and Beals, J. The Structure of Problem and Positive Behavior Among American Indian Adolescents: Gender and Community Differences. American Journal of Community Psychology, 25, pp. 257-288, 1997; and Mitchell, C.M. and O'Nell, T.D. Problem and Conventional Behaviors Among American Indian Adolescents: Structure and Validity. Journal of Research on Adolescence, 8, pp. 97-122, 1997.

Affectivity and Impulsivity: Temperament Risk for Adolescent Alcohol Involvement

Researchers at Arizona State University studied the joint effects of impulsivity, positive affectivity, and negative affectivity on adolescent alcohol use and alcohol-related impairment. Participants were 427 adolescents (aged 12-18, mean age=14.6) from the third wave of an ongoing study of adolescent children of alcoholics and demographically matched controls. Data were gathered using in-person interviews with adolescents and their parents. Results showed that impulsivity moderated the effects of positive affectivity on both alcohol use and alcohol-related impairment. Impulsive adolescents who were also characterized by low levels of positive affectivity showed higher alcohol involvement and more alcohol-related impairment than did either impulsive adolescents with high levels of positive affectivity or non-impulsive adolescents. These results suggest that multiple dimensions of temperament (and their interaction) must be taken into account in trying to predict alcohol involvement, and that positive affectivity is important above and beyond negative affectivity. Practically, the results suggest that positive affectivity is a potentially important target for intervention programs. Although intervention programs often focus on teaching adolescents how to cope with negative affect, less attention has been paid to interventions that increase positive affect, and strategies that target positive affect may be helpful additions to prevention programs. Colder, C.R., and Chassin, L. Psychology of Addictive Behaviors, 11, pp. 83-97, 1997.

Effects of Stress on Delinquency and Drug Use Similar for Males and Females

In a test of Agnew's revised individual-level strain theory, which postulates a mediating role for negative relationships and resulting negative affect (primarily anger) that propels adolescents toward deviant adaptations, investigators at the National Opinion Research Council (NORC) assessed the effects of stressful life events on male and female adolescents' subsequent delinquent and drug-using behaviors. They hypothesized the relationships would vary by sex because of previously observed differences in males' and females' reactions to stress. They used structual equation modeling with two waves of data from 11-to 17-year old adolescents in the High Risk Youth Study (N = 803), a sample that includes heavy representation of offsprings of parents with psychological disorders. Modeling variables reflecting stressful life events, school and family attachment, grades, and indicators of delinquency and substance use, they found that stressful life events have a similar, short-term impact on delinquency and drug use among females and males and that changes in life events were associated with greater delinquency and drug use. The authors presented several possible reasons for the failure to find a sex difference, including the possibility that stress associated with parental psychological disorders may affect males and females similarly. Hoffman, J.P., and Su, S.S. The Conditional Effects of Stress on Delinquency and Drug Use: A Strain Theory Assessment of Sex Differences. Journal of Research in Crime and Delinquency, 34 (1), pp. 46-78, 1997.

Drug Use and Parenting in Adolescence

This project at the University of Washington conducted longitudinal analyses of 241 adolescent mothers' use of 3 substances -- marijuana, alcohol and tobacco -- from the time of pregnancy through 1 year post-partum. Variables measured for each substance included use, verified with urinalysis; intentions to use; attitudes and perceived norms about using; and specific outcome and normative expectations about use of each substance. Substance use increased dramatically from pregnancy to 6 months post-partum, leveling off between 6 and 12 months post-partum. Consistent with the Theory of Reasoned Action model, as substance use increased after pregnancy, so did intentions to use, favorable attitudes toward use, perceived norms regarding use, and beliefs favorable to use. Changes in specific outcome and normative beliefs were observed and that, over time, the young mothers saw bad outcomes of using substances (such as negative effects on their health) as less likely, and positive outcomes (such as helping them to forget their problems) as more likely. They also perceived less disapproval from their families, friends, and doctors after their babies were born. These findings suggest that young mothers are heeding warnings about the danger to their babies of using substances during pregnancy, but are less convinced that substance use has negative effects on parenting or on their own health. The contrast between the abundance of messages warning about substance use during pregnancy and the very few messages aimed at reduce substance use among parents may unwittingly reinforce a notion that substance use is not very harmful when one is not pregnant. A promising intervention approach for young mothers may be to capitalize on their concerns for their babies' health. Morrison, D.M., Spencer, M.S., and Gillmore, M.R. J. Res. Adoles., 8, pp. 69-95, 1998.

Sibling Antisocial Behavior, Substance Use, and Intra-family Conflict

Conventional covariance structure analysis, such as factor analysis, is often applied to data obtained in a hierarchical fashion, such as siblings observed within families. Multivariate modeling of such data, however, is most frequently done as if the data were obtained as a simple random sample from a single population. An alternative specification is presented that explicitly models the within-level and between-level covariance matrices in family antisocial behavior. Data from the National Youth Survey included 1076 adolescents from 450 households. The age of participants ranged from 11 to 17 years with a mean age of 13.9 years. Antisocial behavior included theft, aggression, vandalism, and minor violations. Predictors of family antisocial behavior included life transitions, parent marital status, and whether the family was receiving public assistance. Results showed homogeneity in antisocial behavior within sibling clusters and heterogeneity across families. Between-family variation in antisocial behavior represented approximately 31% of the total variation in antisocial behavior scores. Families experiencing greater numbers of life transitions had higher family levels of antisocial behavior. Findings highlight potential pitfalls of ignoring issues of dependence and demonstrate the importance of examining family-level clustering of specific problem behaviors, such as antisocial behavior.

Another paper on sibling behavior demonstrates a more appropriate specification which explicitly models the within-level and between-level covariance matrices of sibling substance use and intra family conflict. Participants were 267 target adolescents (mean age=13.11 years) and 318 siblings (mean age=15.03 years). The level of homogeneity within sibling clusters, and heterogeneity among families, was sufficient to conduct a multilevel covariance structure analysis (MCA). Parent and family-level variables consisting of marital status, socioeconomic status, marital discord, parent use and modeling of substances, were used to explain heterogeneity across families. Marital discord predicted intra family conflict, and single-parent families and families with higher levels of parent displayed substance use had higher levels of sibling substance use. Duncan, T.E., Alpert, A., and Duncan, S.C. Multilevel Covariance Structure Analysis of Sibling Antisocial Behavior. Structural Equation Modeling, 5, pp. 211-228, 1998; Duncan, T.E., Alpert, A., Duncan, S.C., and Hops, H. Journal of Psychopathology and Behavioral Assessment, 18, pp. 347-369, 1997.

Social Context Predictors of Adolescent Substance Use Development

This study examined the form of growth in alcohol, cigarette, and marijuana use among adolescents and covariates influencing this growth. Participants were 664 male and female adolescents (ages 14 to 17 years) assessed at three time points. A common trajectory existed across the developmental period with significant increases in all three substances. Second-order multivariate extensions of the basic latent growth modeling framework indicated that associations among the individual differences parameters representing growth in the various substance use behaviors, could be adequately modeled by a higher-order substance use construct. Inept parental monitoring, parent-child conflict, peer deviance, academic failure, gender, and age, were significant predictors of initial levels and the trajectory of substance use. Results indicate considerable similarity in the development of alcohol, cigarettes and marijuana during adolescence, and suggest that it may be possible to reduce the upward trajectory of adolescent substance use if we improve the prevalence of effective parental monitoring, reduce parent-child conflict and associations with deviant peers, and increase academic success. Duncan, S.C., Duncan, T.E., Biglan, A., and Ary, D.V. Contributions of the Social Context to the Development of Adolescent Substance Use: A Multivariate Latent Growth Modeling Approach. Drug and Alcohol Dependence, 50, pp. 57-71, 1998.

Using Developmental Processes to Predict Substance Use Outcomes

Latent growth models have been used to describe developmental growth and to identify factors that influence growth in predictable ways. However, there are few examples of research that use characteristics of growth trajectories as predictors of developmental outcomes. This report provides an illustration of this type of study, where the development of ego resiliency (a sequence of stages of functioning across the domains of personal relationships, impulse control, moral develop-ment and cognitive style) over the adolescent years is modeled as a predictor of alcohol and tobacco use in early adulthood. Ego development was assessed six times during a ten year interval in a sample of 123 adolescents. Previous work with these data show that growth in ego development can be described as a curvilinear trajectory over this ten year period. This paper extends prior work to examine the extent to which ego level at three different points along the ten year trajectory (the initial assessment, midway through the study, and the final assessment), the rate of growth at these three points, and the extent to which curvature in ego growth predicted subjects' substance use status in the final year of the study. Results show that steady growth in ego development during early adolescence is associated with nonsmoking status in early adulthood. In addition, youths with steady ego development are more likely than those with delayed development to report moderate drinking habits. Sayer, A.G. and Allison, T.J. Using Developmental Processes to Predict Substance Use Outcomes. The Methodology Center Technical Report Number 98-25, College of Health and Human Development, The Pennsylvania State University, 1998.

Modeling of Longitudinal and Multilevel Alcohol Use Data

Using Multilevel Latent Growth Modeling (LGM), levels of alcohol use and development of alcohol use over four years were examined among individuals (adolescents and parents) nested within families. The sample comprised 435 families (435 target adolescents, 203 siblings and 566 parents) assessed annually for four years. The effects of family status (single-parent, two-parent intact, and stepparent families) and socioeconomic status (SES) on family levels and development of alcohol use were examined. Approximately 29% of the total variation in initial levels of alcohol, and 37% of the developmental trajectories for alcohol use could be explained by family membership. Stepparent families, and less educated and more economically disadvantaged families, had higher family levels of alcohol use and developed in their use of alcohol at a faster rate. Findings suggest that the alcohol use of individuals in the same family is more alike than that of individuals from different families and that family alcohol use may be influenced by family-level variables such as family composition or SES. Duncan, T.E., Duncan, S. and Hops, H. Latent Variable Modeling of Longitudinal and Multilevel Alcohol Use Data. Journal of Studies on Alcohol, 59, pp. 399-408, 1998.