If a story’s opening headlines and decks include the line, “…or is it just bad science?” you can be pretty sure the writer’s answer is yes. And if it’s written by Gary Taubes, who has an appetite for underscoring what he regards as wrong and maybe even shamefully wrong science, the answer is of course. There is much to admire in his long Sunday NYTimes Magazine piece on the flaws of epidemiology and the tendency to ascribe to its reports more certainty than they merit. He leads on the rise and fall of hormone replacement therapy, takes on observational cohort studies in general, and the uncertainty inherent in uncontrolled studies that easily leave cause and effect ambiguous.

The story is sensible in its facts. The tone is accusatory, suggesting that public health authorities ought long ago to have wised up to the failure of many studies to account for confounding variables before telling people – say – to be sure to eat their cruciform vegetables, or swallow that statin. Taubes’s advice: be skeptical of all cohort studies. Wait for a controlled, blinded test of its implications.

Two extra items to note:

1. Matthew Nisbet, the American University man well known for his analyses of how the framing of information biases its reception, has a sharp critique of the Taubes piece at his blog site, Framing Science. It is, he writes, “disastrous and irresponsible” in its choice of illus and headers that imply that the topic and problem is careless human experimentation and big science. Nisbet suggests, rather, that the main malefactors in turning epidemiology into overhyped advice are not within the scientific community but – gasp – the media and the public.

2. In a case of dueling epidemiology stories, a long takeout package by Andreas von Bubnoff in the Los Angeles Times tackles much the same material as does Taubes, with similar advice that the public be more skeptical. This one, however, has less the tone of an exposé than a straightforward explainer of the limits of observational science. It is more likely to inform — but less likely to set off discussions around the water cooler.

Speaking of Skepticism: The LA Times’s Healthy Skeptic series brings a no-surprise-here verdict by Chris Woolston on something called Enzyte, sold on TV with the help of Smiling Bob for “male enhancement.” It doesn’t work.

LATE ADDITION: (Sept 20) Turns out Robert Lee Hotz at the Wall Street Journal had a punchy column Sept. 14 on the flaws to be found in many medical papers. By his source’s calculation, most published research findings are wrong. Lee writes well, including this gem: “In a wilderness of knowledge, it can be difficult to distinguish error from fraud, sloppiness from deception, eagerness from greed…” etc. Indeed.

-CP