Social contagions spread from person to person, much like viruses. Unlike viruses, social contagions involve the interpersonal spread of attitudes, emotions, and behavior within a specific social network. An important distinction between social contagion and conformity is that social contagion is spread without intent. To date social contagion has been observed in such disparate phenomenon as obesity, smoking, loneliness, and even divorce.
Consuming alcohol is commonly a social activity and it evokes an image of people gathered in a bar or pub to imbibe and make merry. Although there are many known factors influencing alcohol consumption, the present study posed this tantalizing question: is contagion evident among social networks for drinking behavior over time?
Data for the study was collected as part of the Framingham Heart Study, a longitudinal study that collected data from two cohorts (original subjects and their offspring) between 1971 and 2001. The dataset relevant to this study included self-reported alcohol consumption and social network analysis consisting of family, friends, and others. This social network data enabled the authors to examine the drinking behavior of members of each subject’s social network. Data was collected every 2 years from the original cohort and every 4 years for their offspring.
Participants were grouped into three categories based on their weekly alcohol consumption:
- abstainers (no alcoholic drinks)
- moderate drinkers (1-7 drinks by women or 1-14 drinks by men)
- heavy drinkers (more than 7 drinks by women or more than 14 drinks by men).
This study compared the roles of “spontaneous” changes in drinking behavior to that of one’s social connections. Spontaneous changes in drinking behavior may result from advertising, changes in availability of alcohol, stigma, as well as policy changes. Their results parsed the changes over time in alcohol consumption as a function of spontaneous changes and social network influence.
Moderate drinkers had the least influence on their social networks. “We discovered that both abstainers and heavy drinkers have a significant impact on the drinking habits of their social connections and that this influence remained constant over the 30-year data period” (p. 9).
To what degree did the abstainers and heavy drinkers influence moderate drinkers within their network? The influence of abstainers was more attenuated, increasing the probability by 18% that moderate drinkers would become abstainers. However, the influence of heavy drinkers was more substantial: their influence increased the probability by 40% that moderate drinkers would become heavy drinkers (p. 9).
Looking at members of the two extreme drinking categories, “abstainers also had a significant positive influence on heavy drinkers to quit drinking” (p. 10). Unfortunately, the direction of influence could also flow in the opposite direction, with abstainers being swayed by heavy drinkers to start drinking. Likewise, moderate drinkers had a minor influence on abstainers, but no influence on heavy drinkers.
A positive outcome revealed that even though few heavy drinkers (1.8%) will become abstainers, if a heavy drinker is surrounded by abstainers, the likelihood they will stop drinking increases by 50%.
Using historical data on drinking behavior to predict future directionality, the authors found that “heavy drinking will continue to decrease to around 13%, down from 22% in 1975, and abstainers increase to a value very similar to moderate use, of 43%” (p. 11), representing a shift toward the mean for both abstainers and heavy drinkers.
Citation: van den Ende, M. W., van der Maas, H. L., Epskamp, S., & Lees, M. H. (2024). Alcohol consumption as a socially contagious phenomenon in the Framingham Heart Study social network. Scientific Reports, 14(1), 4499.
©Jennie Dilworth, Ph.D