"There's lots of room for different factors to influence the behavior of two people who are married," Grant says. "One spouse could work at a place where the co-workers go out for a drink after work. Or one spouse could be a regular churchgoer, while the other prefers to sleep."

Another aspect of the environment is the drinking behavior of one's partner. The researchers found that the impact of the partner's drinking depends on whether it's examined along with non-random mating. Once the researchers accounted statistically for the fact that "like marries like," they saw that the additional influence of the partner's behavior tended to reduce the likelihood of problem drinking. Grant says that although non-random mating means that a person with genetic risks for alcohol problems will tend to marry another with a propensity toward alcohol dependence, it appears that when one spouse begins to abuse alcohol, the other might actually reduce alcohol intake.

"We don't really know how this works," she explains. "It is possible that an individual decreases his or her alcohol consumption in reaction to the other's excessive alcohol use. Maybe one person is responsible for getting the kids up and out for school in the morning, for example."

Grant says she hopes soon to study how spouses might influence not only each other's risk of alcohol dependence but also other psychiatric disorders, such as depression and how those factors interact. And she says as more is learned about these risks, it's important to let people know what they're up against.

"Education is a key to reducing risk for alcohol dependence," she says. "Regardless of genetic risks, there are other detrimental environmental factors associated with alcohol, including reduced educational attainment and income, fewer social and neighborhood support networks, higher rates of divorce and single parenthood and exposure to other psychiatric problems. We need to make people aware of all of their risks, so they can take steps to protect themselves."