Ruby - you have a good point about health in the 1960s.
I think that the author is arguing that you can't make any causal inferences from observational epidemiology! And the only way to truly understand biological/causal relationships is through RCTs. I thought this was a pretty dramatic point but it speaks to the media and public's weakening confidence in the science of epidemiology. It also speaks to the modern era of "black-box" epidemiology.
Here is a great quote from Kreiger's 1994 paper "Epidemiology and the web of causation":
"Epidemiologists widespread practice of treating disease mechanisms as a 'black box' will soon render the discipline incapable of making meaningful contributions to understanding disease etiology"
And this was written 20 years ago!!