A new study has found that relationships between coworkers can predict Body Mass Index (BMI) and that corporate email traffic data can now be used to map a social network that identifies connections influential to one’s likelihood of being obese. Published in the peer-reviewed journal,
, the study focuses on testing scalable methods for mapping workplace social networks and investigating how well these network maps forecast BMI. The study mapped a corporate social network at a multi-site company using survey data and readily available data from email traffic. Based on the pioneering science developed at Harvard University by Nicholas Christakis, MD, PhD, MPH, and James Fowler, PhD, this research was conducted jointly by social network analytics firm Activate Networks and the world’s largest well-being improvement company, Healthways (NASDAQ: HWAY).
“Social influence in the workplace profoundly affects many aspects of our lives, including our health,” said Dr. Christakis. “We have found that social influence is one of the powerful factors, if not the most powerful measurable factor, affecting such health behaviors as weight gain, weight loss, smoking cessation, exercise, mood, and even altruism. This study contributes importantly to our understanding of the power of social networks at work, and it does so by tracing the email communications among people.”
Activate Networks and Healthways utilized corporate email traffic data to identify social connections that proved to predict body mass index (BMI) of individuals in the workplace. The study used data from email traffic that already exists in most organizations, instead of requiring self-reported data, which can often be incomplete, although the study had access to both types of data. The result of the research was a map of the organization that reveals which employees are likely to be the most influential to their peers, and which employees would be more likely to adopt a particular health trait, in this case being either relatively thin or heavy. The study also found that network maps derived from email traffic were comparable to traditional network maps derived from self-reported survey data.