Social networks affect alcohol drinking patterns

by Kristina Fiore

If you’re surrounded by heavy drinkers, you’re more likely to become one, new research shows.

An analysis of data on alcohol consumption and social networks found that people are 50% more likely to drink heavily if someone they’re directly connected to does so, Nicholas Christakis, MD, PhD, of Harvard University, and colleagues reported in the Annals of Internal Medicine.

“We’ve found that the influence of your friends and people you have connections with can affect your health just as much as your family history or your genetic background,” Christakis said in a statement. “With regard to alcohol consumption, your social network may have both positive and negative health consequences, depending on the circumstances.”

Research has shown that a person’s drinking pattern has many determinants, including biological, social, and cultural factors.

To assess whether alcohol consumption behavior spreads from person to person in a large social network, the researchers looked at data from the Framingham Heart Study on 12,067 people who were followed for 32 years between 1971 and 2003.

Data on social ties were collected during each of the seven waves of the study, and alcohol consumption was self-reported.

The researchers found that people were 50% more likely to drink heavily if a person they were directly connected to drank heavily.

The size of the effect was 36% for those at two degrees of separation, such as a friend of a friend, and 15% for those at three degrees of separation, they added.

It disappeared at four degrees of separation.

Christakis and colleagues also found that people were 29% more likely to abstain from drinking if someone they’re directly connected to abstained.

That effect size was 21% at two degrees of separation and 5% at three degrees, and disappeared at four, they said.

They also found over the course of the study that those who were surrounded by heavy drinkers increased their alcohol consumption by about 70% compared with those who weren’t connected to any heavy drinkers.

Conversely, those surrounded by nondrinkers decreased their alcohol consumption by half.

In multivariate analyses, the researchers found that for every additional connection one made to a heavy drinker, the risk of drinking heavily increased by 18% (95% CI 11% to 25%, P<0.001), and it decreased the likelihood that he or she would abstain by 7% (95% CI 2% to 12%, P=0.009).

On the other hand, each additional abstainer in a person’s life significantly reduced the likelihood that he or she would drink heavily by 10% (95% CI 4% to 15%, P=0.001).

Effects on drinking were also dependent on the relationship. Heavy drinking by a wife, for example, increased the husband’s likelihood of drinking heavily by 196%, while heavy drinking by a husband increased a wife’s chances for drinking heavily by 126%.

That effect was significantly smaller among siblings and didn’t differ whether the contact was a brother or a sister.

But overall, the researchers wrote, it appeared that female contacts are significantly more likely to influence the spread of heavy alcohol consumption.

This could be explained by the fact that increases in drinking behavior are less common among women and are more often associated with dramatic shifts in roles and contexts in life, they theorized.

It could also be that changes in perceived norms toward drinking among women are more powerfully transmitted along social networks because women are perceived as sharing norms for less alcohol consumption. “A woman who changes her behavior would therefore be a stronger stimulus,” the researchers wrote.

Some relationships had absolutely no effect on drinking habits. Neither immediate neighbors nor geographic distance, for example, changed alcohol consumption behaviors, the researchers said.

They acknowledged that the study was limited by self-reported data, and may be too ethnically homogeneous to generalize to a larger population.

In addition, the study could not estimate health outcomes from the amount of alcohol consumed, and the authors noted that moderate alcohol intake has been associated with improved cardiovascular health.

But they concluded that network phenomena might be exploited to spread positive health behaviors.

“Our findings reinforce the idea that drinking is a public health and clinical problem that involves groups of interconnected people who evince shared behaviors,” Christakis said in the statement. “In treating individuals for problematic drinking, we need to look at their social networks to identify and eliminate obstacles to abstaining.”

Kristina Fiore is a MedPage Today staff writer.

Originally published in MedPage Today. Visit for more psychiatry news.