In previous research on 12,630 individuals drawn from the Framingham Heart Study (FHS), we found evidence that one person’s weight gain was associated with weight gain in others to whom he or she was connected. Here, we seek to extend that work by: (1) evaluating its causal basis in a new way, and (2) also examining smoking behavior. In this project, we will evaluate peer effects with respect to health behavior using an “instrumental variable” approach. We propose to exploit known correlations of given variants of genes (alleles) and particular health outcomes, and to use allelic variation between individuals as an instrumental variable (IV). The key idea is that a person (an “ego”) may have peers (or “alters”) who are randomly “assigned” genes predisposing the alters to certain health behaviors, and that this random assignment can be seen as a kind of natural experiment, exposing the ego to peers who either exhibit or do not exhibit the pertinent behaviors. However, while the idea of using genes as instrumental variables in socioeconomic research is intriguing and has generated some excitement among social scientists, a careful review of the necessary assumptions for any such use is in order. Hence, we have three specific aims. First, we will embellish our FHS Social Network dataset describing 5,124 individuals (egos) within a social network of 12,630 people (their possible alters), all of whom are measured repeatedly from 1971 to 2007 as part of the FHS; specifically, we will obtain and merge detailed genetic data to this file. Second, we will carefully evaluate several key assumptions underpinning the use of genes as IVs, especially as pertains to a network setting. We will also need to account for biological issues such as possible “linkage disequilibrium” in genes, and sociological issues such as the tendency to homophily. Third, using a two-stage IV estimation approach, we will assess whether obesity and smoking in an ego’s alters (including friends, neighbors, spouses, siblings, and cousins) are causally related to similar behaviors in the ego. Our hypothesis is that obesity and smoking behaviors in alters will stimulate similar behaviors in egos and that this effect will vary according to the closeness of the relationship between the ego and the alter. We will examine these peer effects in the context of cardiovascular disease, which is responsible for 40% of deaths in the U.S. and incurs costs of over $350 billion annually. Our results could guide policy-makers by shedding light on the ways in which the embeddedness of individuals in social networks is relevant to their health. Our results could also open up new ways of exploiting genetic information to tackle longstanding and difficult problems of causal inference in the social sciences.
Papers published so far based on this work include:
N.A. Christakis, “Social Networks and Collateral Health Effects Have Been Ignored in Medical Care and Clinical Trials, But Need to Be Studied,” British Medical Journal 329: 184-185 (July 2004).
N.A. Christakis and J. Fowler, “The Spread of Obesity in a Large Social Network Over 32 Years,” The New England Journal of Medicine (July 2007).
N.A. Christakis and J. Fowler, “The Collective Dynamics of Smoking in a Large Social Network,” The New England Journal of Medicine, 358(21): 2249-2258 (May 2008).
