There may be a new explanation for why fad diets tend to cluster in pockets - and it may help companies spend their marketing money a little smarter also.

If a new diet or exercise program is clearly working for a friend it carries more weight, pardon the pun, than a charismatic, but less successful pal or a celebrity endorsement. Countless diets and weight-loss programs exist to combat the high rate of obesity among Americans and they probably all work as well as each other, assuming they lead to fewer calories going surplus each day, but they do little good if people don't adopt them. To optimize adoption, the authors simulated the behavior of fictional people created using combinations of physical attributes and personality traits, such as the ability to lose weight and a high or low body mass index. The model distributed traits based on national population averages. 

Based on the simulations, people in social networks linked to someone who successfully lost weight or had a high body mass index produced the largest total weight loss among peers. The networks surrounding a person with a high number of friends -- those who were more charismatic or popular -- produced lower weight-loss totals.

"People want to see that positive influence," says University at Buffalo occupational health researcher Lora Cavuoto. "Understanding how social influence affects people's participation in health programs can lead to better-designed wellness interventions." 

The results support the new approach many weight-loss programs have taken in attracting new participants: Celebrity endorsers with a large reach have taken a backseat to everyday people who benefit from a new diet or workout. 

"Your ties and social contacts may have a bigger effect because you see them every day and you have that close connection," says Cavuoto. "If they can be successful, then that's your best way of getting information out that a program is good."

Engineering doctoral candidate Mohammadreza Samadi and engineering graduate student Mahboobeh Sangachin worked with Cavuto on the study, "Modeling the Spread of an Obesity Intervention through a Social Network," published in the Journal of Healthcare Engineering.