Search   |  Advanced

Research Summary

Does the Alcohol Use of Family and Friends Influence Individual Use?

Health risks such as smoking, obesity, and sexually transmitted disease can travel through social networks. To determine whether and how alcohol use travels through such pathways, researchers analyzed longitudinal data from 12,067 Framingham Heart Study participants assessed every 2–4 years between 1971 and 2003. Social network ties for 5124 principals (i.e., the focal individuals of the network) and self-reported alcohol consumption for principals and their contacts were assessed at each time point. Twenty-two percent of principals were heavy drinkers,* and 15% were abstainers.

  • Principals were 50%, 36%, and 15% more likely to be heavy drinkers if individuals between 1–3 degrees of separation,† respectively, were also heavy drinkers.The likelihood that a principal drank heavily increased by 18% for each heavy-drinking social contact.
  • The likelihood that a principal drank heavily increased by 154% if a female friend started drinking heavily but did not increase significantly if a male friend started drinking heavily.
  • Principals were more likely to drink heavily if their spouse or sibling drank heavily but not if a neighbor or coworker drank heavily.
  • Abstinence in principals was associated with abstinence in social contacts in a pattern similar to the heavy drinking results.
*Defined as consuming, on average, more than 1 drink per day for women and more than 2 drinks per day for men. †Defined as follows: 1=close friend, 2=friend of a friend, and 3=friend of a friend of a friend.


These interesting results suggest that alcohol use behaviors (both heavy drinking and abstinence) are influenced not only by family and close friends but also by more distant social contacts. Thus, public-health and clinical interventions to promote safe alcohol use should target both individuals and social groups. Kevin L. Kraemer, MD, MSc


Rosenquist JN, Murabito J, Fowler JH, et al. The spread of alcohol consumption behavior in a large social network. Ann Intern Med. 2010:152(7): 426–33.