Could there be third correlated variable namely income that correlate with both western diet and health care costs? Or even better records of disease and deaths?