drinkers. Smoking and exercise also showed a dose-response relationship
to the amount of caffeinated coffee consumed. Risk factor
levels among drinkers of decaffeinated coffee were more like those
of caffeinated coffee than non-drinkers. These data illustrate the
clustering of atherogenic behaviors with coffee drinking and highlight
their potential importance in interpreting the growing body of
literature about coffee and health. (Am J Public Health 1990;
80:1310-1313).