Epidemiology of Substance Abuse

Epidemiology of Substance Abuse

Denise B. Kandel, Ph.D., Chief of Psychiatric Research
Pamela Griesler, Ph.D., Research Scientist IV
Christine Schaffran, M.S., Research Scientist III

The department investigates the natural history of drug involvement, the risk factors and consequences of drug abuse, and the comorbidity of substance abuse with dysfunctional behaviors and psychiatric disorders among adolescents and adults. Cigarette smoking and nicotine dependence have become an increasing focus of the research activities of the department.

Current investigations are based on multiple general population samples, including longitudinal samples, and relational samples of matched parent-child pairs. Six of the samples are national. Five research areas continued to be emphasized this year: the intergenerational transmission of substance use; the impact of maternal smoking in pregnancy and postnatally on offspring smoking and behavior problems; ethnic differences in the etiology of adolescent smoking; the natural history of nicotine dependence; and developmental stages of involvement in drugs. Selected findings are highlighted.

Intergenerational Transmission of Substance Use

Using dyadic data from multiple waves of the National Household Survey on Drug Abuse and the Current Population Survey, parent-child similarity on smoking was examined in different ethnic groups in the American population. Parental smoking increases the risk that children will smoke. Parent and offspring similarity on smoking is stronger for mothers than fathers, when both parents smoke than when only one does, for daughters than sons, is dose related, persists with controls for other covariates, and is of the same magnitude among whites, African-Americans, and Hispanics. The associations between parents and children are especially strong when parents meet criteria for nicotine dependence and when children have ever smoked daily. Parental use of drugs other than tobacco, i.e., alcohol, marijuana, and cocaine, each also independently increases the risk that a child will smoke. Similar patterns were observed with respect to the influence of parental marijuana use on child marijuana use.

The influence of parental drug use on their children's use is not restricted to use of the same drug. Parental use of any drug class increases the risk that a child will use any other drug class as well.

Ethnic Differences in Predictors of Adolescent Smoking

Ethnic differences in the predictors of smoking were examined in a national longitudinal sample of mother and child dyads. There are more common than specific familial, social, and behavioral predictors of initiation and persistence of smoking among youths of different ethnicity. Across ethnic groups, maternal smoking, peer pressure to smoke, and problem behavior are risk factors for smoking initiation; problem behavior and low academic competence are risk factors for smoking persistence. Different risk factors predict different stages of youth smoking. Interpersonal factors are more important for onset of smoking than persistence; intrapersonal factors are more important for persistence.

Effects of Maternal Smoking on Child Problem Behavior and Adolescent Smoking

We had earlier found that prenatal exposure to maternal smoking was related to smoking by female offspring. The consequences of exposure to prenatal maternal smoking for child problem behaviors were examined to determine whether conduct problems could account for this prenatal effect. The developmental trajectories of classes of problem behaviors were examined in children aged 4 to 17 years from 18 consecutive birth cohorts in a national sample. Males had higher mean levels of Hyperactive, Headstrong, and Antisocial problem behaviors than females. Minority children, particularly African-Americans, had higher mean levels of Hyperactive, Antisocial, and Anxious/Depressed problem behaviors than whites. Hyperactive problem behaviors declined with age while Anxious/Depressed problem behaviors increased with age especially among females. White children increased more steeply in Anxiety/Depression than African-Americans. Children whose mothers smoked heavily during pregnancy had the highest levels of Hyperactive, Headstrong, Antisocial, and Anxious/Depressed problem behaviors and the steepest increase in Anxious/Depressed symptomotology compared with children of mothers who did not smoke in pregnancy. Controlling for postnatal smoking, maternal prenatal smoking, and externalizing problem behaviors (summary measure of Hyperactive, Headstrong, and Antisocial), each retained a unique effect on onset of smoking for girls. By contrast, for sons, postnatal, but not prenatal, maternal smoking and Anxious/Depressed problem behaviors predicted smoking onset. These findings highlight the role of early biological risk in the etiology of problem behaviors and cigarette smoking, especially for female offspring.

The Natural History of Nicotine Dependence

The natural history of nicotine dependence was examined in a national sample from retrospective data on age of onset into smoking, rates of daily smoking, age of onset into daily smoking, and rates of nicotine dependence. Early smokers are more likely to progress to daily smoking and daily smokers are more likely to be dependent than non-daily smokers. However, progression to daily smoking is slower for those who onset early (below age 15) than those who onset at a later age.

Developmental Stages of Drug Involvement

In 1998, Dr. Kandel organized a multidisciplinary meeting to asses the Gateway Hypothesis, which postulates that there is a regular sequence of involvement from cigarettes and alcohol to marijuana and to other illicit drugs. The goals were to synthesize what is currently known about stages of drug use in different population groups, specify the effects of drug prevention efforts on sequences of involvement, analyze the advantages of different statistical approaches for identifying stages of involvement, and learn about drug progression from animal models and the neurobiology of addiction. This past year, Dr. Kandel edited a volume based on papers presented at the conference, to be published by Cambridge University Press.

The program of research of this department generates hypotheses to be tested in the laboratory and illustrates the important contributions that epidemiological investigations of substance use can make to biological research.