Understanding Health and Well-Being from Naturalistic Driving Behavior

The way humans interact with their environment leaves traces that reflect their health and well-being. Advances in sensing technologies now make it possible to capture these traces passively and unobtrusively. Here, we test the idea of using everyday driving patterns to predict health and well-being. We analyzed everyday driving patterns from 2,658 older adults who also completed self-reported measures of general well-being and health across cognitive, physical, social, and mental domains. Driving behaviors predicted both overall well-being and specific health domains beyond demographic variables. Physical and social health were most strongly associated with driving variables. Furthermore, we identified several driving signatures of health that were highly specific in their predictions. Finally, including driving variables improved out-of-sample predictive performance relative to a demographic-only model. Overall, these findings position driving behavior as a potential new source of personal sensing that can be leveraged to understand and promote health and well-being.