Public Policy That Makes Test Subjects of Us All

http://www.nytimes.com/2009/04/07/science/07tier.html?hpw

Suppose you wanted to test the effects of halving the amount of salt in people’s diets. If you were an academic researcher, you’d have to persuade your institutional review board that you had considered the risks and obtained informed consent from the participants.

You might, for instance, take note of a recent clinical trial in which heart patients put on a restricted-sodium diet fared worse than those on a normal diet. In light of new research suggesting that eating salt improves mood and combats depression, you might be alert for psychological effects of the new diet. You might worry that people would react to less-salty food by eating more of it, a trend you could monitor by comparing them with a control group.

But if you are the mayor of New York, no such constraints apply. You can simply announce, as Michael Bloomberg did, that the city is starting a “nationwide initiative” to pressure the food industry and restaurant chains to cut salt intake by half over the next decade. Why bother with consent forms when you can automatically enroll everyone in the experiment?

And why bother with a control group when you already know the experiment’s outcome? The city’s health commissioner, Thomas R. Frieden, has enumerated the results. If the food industry follows the city’s wishes, the health department's Web site announces, “that action will lower health care costs and prevent 150,000 premature deaths every year.”

But that prediction is based on an estimate based on extrapolations based on assumptions that have yet to be demonstrated despite a half-century of efforts. No one knows how people would react to less-salty food, much less what would happen to their health.
Read more...