Wouldn't that match this study? Low income Americans driink either a lot or not any coffee at all. High income Americans drink some coffee. So you would expec

Wouldn't that match this study? Low income Americans driink either a lot or not any coffee at all. High income Americans drink some coffee. So you would expect no coffee consumption and high coffee consumption to correlate with a higher risk than moderate consumption. This is exactly what the study found.