This offers up an alternative scenario to the common assumption that we are born with an innate preference for sugar because it would have been evolutionarily beneficial, prompting us to seek out those foods that are the densest source of calories in a world in which calories were supposedly hard to come by. “In evolution,” as the Yale psychologist Linda Bartoshuk told the New York Times in 1989, “we needed the energy of sweet-tasting, sugary foods, especially during times of scarcity.” The research of Le Magnen and others suggests that these preferences have little to do with the presence of famine in our evolutionary history (as discussed on Chapter 14) and everything to do with the absence of these refined carbohydrate foods. We come to prefer these foods, according to the alternative hypothesis, because they induce an exaggerated version of the post-absorption responses to naturally occurring sources of glucose and fructose—either plant foods that are difficult to digest (the kinds of roots, tubers, or fruit eaten by Paleolithic populations) or the protein in meat and the relatively slow conversion of its amino acids into glucose.

Since insulin plays the critical role in our post-absorption responses to particular foods, it’s not surprising that insulin may play the critical role in our determination of palatability. A little-discussed observation in obesity research is that insulin is secreted in waves from the pancreas. The first wave begins within seconds of eating a “palatable” food, and well before the glucose actually enters the bloodstream. It lasts for perhaps twenty minutes. After this first wave ebbs, insulin secretion slowly builds back up in a more measured second wave, which lasts for several hours.*135 The apparent function of the first insulin wave is to prime the body for what’s coming. It takes insulin almost ten minutes to have a measurable effect on blood-glucose levels; it takes twice that long to have any significant effect. Meanwhile, glucose is entering the bloodstream from the meal and continuing to stimulate insulin secretion. When blood sugar is at a maximum, the signal to the pancreas to secrete insulin is also highest, but by this time enough insulin has already been secreted to do the necessary job of glucose disposal. “The pancreas has no idea what’s going on elsewhere in the body,” says University of California, San Francisco, biochemist Gerald Grodsky, who pioneered much of this work. “All it sees is the glucose.” The way we apparently evolved to deal with this systems-engineering problem is the flooding of insulin into the circulation immediately upon beginning a meal; this prepares the body in advance to start taking up the glucose as soon as it appears.