Behavior

That risk-related behaviors are evident early in life, remain stable into adolescence, and are associated with a family history of alcoholism suggests that those behaviors are, at least in part, of genetic origin. To establish that, researchers must use genetically informative study designs.

One approach is to study child or adolescent twins and their parents. Several such studies, which specifically assess the initiation of alcohol use and the transition to alcohol abuse, are being conducted throughout the world. We illustrate with two ongoing studies from Finland.

One, "FinnTwin12," is a study of approximately 2,800 twin pairs and their parents. The twins represent all pairs from five consecutive twin-birth cohorts (1983-1987) who were entered into the study as they reached age twelve (1995-1999), when behavioral ratings by teachers and parents were obtained on all participating pairs.

The ratings include multidimensional scales (i.e., scales that rate various characteristics) of behaviors associated with increased alcoholism risk. Two years later, at age fourteen, the twins were followed up, and, while most reported abstinence, about one-third were then using alcohol.

What predicts drinking or abstaining at age fourteen? Genetic factors played a role only among twin sisters, perhaps reflecting their more accelerated pubertal maturation, and environmental effects shared by twin siblings accounted for most of the variation in drinking or abstaining at this age. Differences that twins attributed to their home environments (e.g., in parental monitoring, support, and understanding) and differences in teachers' ratings of twins' behavior at age twelve (in problem behaviors of aggressiveness, impulsivity, and inattention) differentiated those who were drinking from those still abstaining at fourteen.

But once drinking is initiated, genetic effects become evident in individual differences in frequency and quantity of consumption and in behavioral problems that then result. "FinnTwin16," another study of five consecutive, complete birth cohorts of Finnish twins, illustrates. These twins were first studied as they reached age 16, with follow-up twelve and thirty months later, at ages 17 and 18^1/2. At age 16, about 25 percent had remained abstinent.

Of 2,810 twin pairs, both twins in 459 pairs (16.3%) were abstaining, co-twins in 1,964 pairs (69.9%) had concordantly begun drinking by age sixteen, and only 387 pairs were discordant, with one twin drinking and the other abstaining. Concordance is the co-occurrence of the behavior in the twin pair (e.g., both drinking or both not drinking). Overall concordance exceeded 85 percent, regardless of the twins' gender or zygosity.

There was extremely high familial aggregation for alcohol use or abstinence at age sixteen, additional evidence that genes play little role in abstinence or initiation. But thirty months later, individual abstinence had dropped to 10 percent, concordance among twin pairs had declined considerably, and genetic factors increasingly influenced the frequency and quantity of an adolescent's alcohol consumption. MZ twins were significantly more similar in drinking frequency than were DZ twins. The influence of genetic factors increases over time, with increasing experience with alcohol, and the differences between MZ and DZ twins becomes greater at each follow-up.

Regional residency moderates parental and sibling influences on adolescent drinking. Where abstinence is relatively rare, as in the large cities of Finland, siblings have greater effects on one another. Conversely, the protective effect of parental abstinence on that of their adolescent twin children was more evident in sparsely populated rural areas of the country, where abstinence was more prevalent. And, most interestingly, genetic factors exerted a larger role in urban settings than in rural settings from age 16 through the follow-up at age 18^1/2. Common environmental factors assumed greater importance in rural settings.

Such results suggest that environments moderate the impact of genetic effects across many dimensions of behavior. But what aspects of the environment matter? In an analysis of results at age 18^1/2, we demonstrated that specific characteristics of rural and urban environments moderate the effects of genes on drinking behavior. In areas with proportionately more young adults, genetic effects were nearly five times more evident than in communities with relatively few young adults. Thus, dramatic differences in the magnitude of genetic effects can be demonstrated across communities at environmental extremes of specific risk-relevant characteristics.