Carrot Health Insights | Healthy Behavior #3: Fat Chance
Published on February 1, 2018
It’s hard to argue with the US Surgeon General’s characterization of obesity: “a public health issue that is among the most burdensome faced by the Nation”. The American obesity problem is bad, and getting worse. Research from the American Public Health Association shows that adult obesity rates have increased from 15% in 1980 to 35% today. For kids and teens, the rate has nearly tripled over the past 30 years. We have a lot of work to do to fix this epidemic. As this study of preschool-aged children shows, though, there is some reason for hope amid all the troubling statistics.
Why are Americans so obese? The answer is pretty obvious: poor dietary choices (too many calories) and not enough physical activity. What is causing this behavior? The Surgeon General summed it up nicely: “behavioral and environmental factors are large contributors to overweight and obesity and provide the greatest opportunity for actions and interventions designed for prevention and treatment.”
In other words, we won’t win this fight by analyzing clinical data alone. Health systems, payers, and other stakeholders must understand the social and behavioral determinants of health (SDoH) — and they must translate this knowledge into highly tailored strategies for obesity prevention and treatment.
To analyze the social and behavioral determinants of obesity, our data science team developed a predictive model using the Carrot MarketView data repository. This process revealed some interesting things — for example, stationery purchasing (scrapbooking?) is associated with high obesity, and single-family homeownership is associated with low obesity. The table below highlights some of the correlated social and behavioral factors:
Carrot Health Obesity Model – Correlative Factors
The Carrot Health Obesity Model is portable. Because it uses consumer data (not Protected Health Information) as inputs, it can be applied to individuals and populations even if there is no clinical or claims data available. For example, we can use the model to predict the obesity rate for each ZIP code in Florida, as the map below shows (red = high, green = low):
Predicted Obesity Rate in Florida (ZIP code level)