Published on May 23, 2018
If you’re the proud owner of a 1990 Oldsmobile Cutlass Ciera, like the beauty pictured above… you might want to visit your doctor.
Why? There’s nothing inherently unhealthy about the Cutlass Ciera (calm down, Ralph Nader!) But it turns out that people who own this particular car model tend to have poor health. Compared with owners of other vehicles, Cutlass Ciera owners have the ninth-worst rate of high hospital Emergency Department (ED) utilization (24.8%), the third-worst rate of diabetes (22%), and the second-worst rate of tobacco usage (31.9%).
It’s not just Cutlass Ciera owners who are unhealthy. The table below shows the vehicle models that are correlated with the highest observed incidents of the aforementioned health conditions:
10 Worst Cars For Health
This list demonstrates that health outcomes are correlated with financial status. Many of these vehicle models haven’t been manufactured in a long time. The last year for the Cutlass Ciera was 1996 — today, people with Cutlass Cieras are driving 20+ year-old cars worth $1,000 or less. In addition to picking up a correlation with income, this list also picks up a correlation with age, particularly for diabetes and tobacco use. Drivers of Oldsmobiles, Lincoln Town Cars, and Mercury Grand Marquis skew older than average, and diabetes is more prevalent in older individuals.
What about the opposite — which vehicles are correlated with good health? As the following table illustrates, this analysis picks up a very different population. We tend to see more recent vehicle models (the income effect, in the other direction). We also see vehicles whose ownership tends to skew younger, such as the Honda CR-V:
10 Best Cars for Health
Although these lists correlate with income and age, they are more than just proxies for basic demographic variables. Thanks to the magic of data science, we can adjust for those factors, and we can isolate and measure the direct health impact of vehicle ownership.
Vehicle ownership is an example of a Social Determinant of Health (SDoH). By itself, vehicle ownership data isn’t enough to develop an accurate forecast of someone’s future health status and cost. But when combined with thousands of other pieces of information, this collective knowledge can be harnessed — allowing health payers and providers to deepen their understanding of covered populations, and to improve the design, targeting, and effectiveness of interventions.