People with obese friends are more likely to be obese themselves, The Independent reported on July 26 2007. It said that researchers “have found that obesity is socially contagious”. Reporting the same story, the Daily Mail said, “if a person’s friend becomes obese, their own chances of piling on the pounds are almost trebled”. The stories were based on a 32-year study carried out in the US which reported a link between social networks and obesity.
After reading these stories you might be left with the impression that the link with obese friends is “causal” or that to remain thin people should not choose obese friends. On its own, this study is not designed to make these sorts of conclusions. The question as to what mechanism might be causing the link or how the social, environmental and behavioural influences are interacting remains unanswered.
The study was conducted in the US by Nicholas Christakis and colleagues form the Department of Health Care Policy at Harvard Medical School and was published as a special article in the peer-reviewed journal the New England Journal of Medicine.
This is a re-analysis of data from a large, long-running cohort study. The researchers looked at the data about weight gain for more than 12,000 people who had been involved in the Framingham Heart Study and their children. Their weight and height data had been repeatedly collected over 32 years, beginning in 1971, on the children or enrolled people in the original Framingham study.
This study began in 1948 with just over 5,000 people. There are now three generations of data, with few people leaving the study, allowing the researchers to draw a very large network of interconnections between people and to relate changes in body weight to various aspects of the connections – “ties” – between people or the distance –“degrees of separation” – between them.
The researchers were able to detect groups or “clusters” of obese people who were linked to each other. A person’s chance of becoming obese was increased by 57% if he or she had a friend who became obese over a given period of time. The chance of becoming obese was less if an adult had a brother or sister who had become obese – a 40% increase – and if a husband or wife had become obese it was 37%.
The researchers concluded that by analysing this large web of data, the biological and behavioural traits associated with obesity appear to spread through social ties.
This is a very large, well-conducted study using new and interesting techniques to describe the relationships between connected people over time. The analysis of the social influences has demonstrated links which have been quantified. The researchers acknowledge that these links and the underlying mechanisms will need further study.
If we accept that this study shows that friends do not become obese because they adopt each other’s lifestyle, then further explanations of how the effect shown in this social networking study occurs are required.