include the secretion of saliva, gastric juices, and, not surprisingly, insulin. By the 1970s, these cephalic*136 reflexes had been studied in humans, rats, monkeys, cats, sheep, and rabbits. Le Magnen’s student Stylianos Nicolaidis had demonstrated that rats will secrete insulin in response to the mere taste of a sweet substance, and it doesn’t matter whether it is sugar or a no-calorie sugar substitute. The perceived taste of sweetness is sufficient to stimulate insulin secretion. Just as Pavlov demonstrated that dogs will salivate at the sound of a bell they have learned to associate with feeding, Stephen Woods and his colleagues demonstrated that rats will secrete insulin when confronted with similar eating-related stimuli. (These researchers arbitrarily chose the smell of mentholatum, a mixture of menthol and petroleum jelly, more commonly used as a topical rub for chest colds.) Humans will do the same. This reflexive release of insulin, Nicolaidis suggested, is “pre-adaptive”: it anticipates the effects of a meal or a particular food, and so prepares the body. As Mark Friedman describes it, this cephalic release of insulin also serves to clear the circulation of “essentially anything an animal or a person can use for fuel. Not just blood sugar, but fatty acids, as well. All those nutrients just go away.” Hence, the thought of eating makes us hungry, because the insulin secreted in response depletes the bloodstream of the fuel that the peripheral tissues and organs need to survive.
This cephalic secretion of insulin in preparation for the act of eating provides yet another mechanism that may work to induce hunger, weight gain, and obesity in a world of palatable foods, which could mean, of course, simply those foods that induce excessive insulin secretion to handle the unnaturally easy digestibility of their carbohydrates. The idea was suggested in 1977 by the psychologist Terry Powley, who was then at Yale and is now at Purdue University. Powley was discussing the obesity-inducing effect of lesions in the hypothalamus and speculated that the lesions cause the animal to hypersecrete insulin when just thinking about, smelling, or tasting food, and this amplifies its perception of hunger and palatability. The result would be what Powley called a “self- perpetuating situation”—i.e., a vicious cycle. “Rather than secreting quantities of insulin and digestive enzymes appropriate for effective utilization of the ingested material,” Powley wrote, “the lesioned animal over-secretes and must then ingest enough calories to balance the hormonal and metabolic adjustments.”
Powley did not go so far as to suggest that this same phenomenon was at work in humans, but his then colleague Judith Rodin did. Rodin reported in 1980 that those individuals whose eating behavior is most responsive to the smell or sight of food—a grilling steak, in her experiments—were those who also had the greatest cephalic- phase insulin response. Insulin had to be considered a “major candidate,” Rodin suggested, “for an intervening physiological mechanism that might be responsive to environmental stimuli.” By 1985, Rodin was speculating that the chronic hyperinsulinemia of the obese would also exacerbate this phenomenon. “A feedback loop is suggested by these findings in which hyperinsulinemia in turn leads to increased consumption, which, unless compensated for, could lead to further weight gain,” she wrote. “Because acute hyperinsulinemia can also be produced in some individuals by simply looking at or thinking about food, it, too, can in turn lead to increased consumption and possible weight gain.”
The possibility that insulin determines what Le Magnen called the metabolic background of hunger also explains two observations we discussed in the sections on fattening and reducing diets.
The first is the observation by Ethan Sims that he could stuff his convict subjects with as much as ten thousand calories a day of mostly carbohydrate and they would still feel “hunger late in the day,” and yet subjects fed eight hundred superfluous calories of fat “developed marked anorexia.” On a more familiar level: why is it that most of us can imagine eating a large bag (twenty ounces) of movie popcorn—more than eleven hundred calories if popped in oil,*137 as it typically is—but not so the equivalent caloric amount of cheese: say, fifteen slices of American cheese, or a cup and a half of melted Brie?
The simple explanation is that the insulin induced by the carbohydrates serves to deposit both fats and carbohydrates (fatty acids and glucose) as fat in the adipose tissue, and it keeps those calories fixed in the adipose tissue once they get there. As long as we respond to the carbohydrates by secreting more insulin, we continue to remove nutrients from our bloodstream in expectation of the arrival of more, so we remain hungry, or at least absent any feeling of satiation. It’s not so much that fat fills us up as that carbohydrates prevent satiety, and so we remain hungry.
The second observation is the carbohydrate craving associated with obesity. Here the metabolic background of hunger is established by chronic hyperinsulinemia rather than the immediate insulin secretion during a carbohydrate-rich meal. In both cases the insulin induces hunger or prevents satiety. In the case of hyperinsulinemia and obesity, however, this happens even between meals, when the cells should be living off a fuel mixture of predominantly fatty acids. Instead, the insulin traps the fat in the fat tissue, and it signals the cells to burn glucose. As far as the body is concerned, the elevated insulin is the indication that we’ve just eaten—“high levels of insulin herald the ‘fed’ state,” as George Cahill put it—and the signal that carbohydrates are available to be burned. But in this case, they’re not. Now the homeostatic system that evolved to maintain blood sugar in a healthy range establishes an internal environment in which the cells are primed to burn glucose for fuel, and only glucose can satisfy that demand, yet there’s no expendable glucose in the system. High insulin levels even prevent the liver from releasing the glucose that’s stored there as glycogen. As a result, it’s glucose that we crave. Even if we eat fat and protein—our cheese slices, for instance—the hyperinsulinemia will work to store these nutrients rather than allow them to be used for fuel.
The practical implication of this situation is critical to how we perceive the dietary treatment of obesity, or simply the maintenance of a healthy weight, in a world of inexpensive, easily digestible carbohydrate-rich foods. Among the more pessimistic arguments wielded against carbohydrate-restricted diets is that all diets fail eventually because the subjects inevitably fall of the diet, just as they do calorie-restricted diets. But this argument is based on the assumption that all diets work by limiting the calories consumed. It also ignores any physiological difference between a craving for carbohydrates and the hunger that results from semi-starvation. The latter is caused by the absence of sufficient calories to satisfy physiological demands. The craving for carbohydrates is more closely akin to an addiction, which is how it was described by the British clinician Robert Kemp in 1963. It is the consequence of hyperinsulinemia, which in turn is caused initially by the presence of carbohydrates in the diet, just as an addiction to nicotine or cocaine or any other addictive substance is caused by the use of these substances. There is nothing inherently natural about such addictions. The hunger that accompanies calorie restriction is an unavoidable physiological condition; the craving for carbohydrates is not.
Sugar (sucrose) is a special case. Just like cocaine, alcohol, nicotine, and other addictive drugs, sugar appears to induce an exaggerated response in that region of the brain known as the reward center—the nucleus accumbens. This suggests that the relatively intense cravings for sugar—a sweet tooth—may be explained by the intensity of the dopamine secretion in the brain when we consume sugar. When the nucleus accumbens “is excessively activated by sweet food or powerful drugs,” says Bartley Hoebel of Princeton, “it can lead to abuse and even addiction. When this system is under-active, signs of depression ensue.” Rats can be easily addicted to sugar, according to Hoebel, and will demonstrate the physical symptoms of opiate withdrawal when forced to abstain.
Whether the addiction is in the brain or the body or both, the idea that sugar and other easily digestible carbohydrates are addictive also implies that the addiction can be overcome with sufficient time, effort, and motivation, which is not the case with hunger itself (except perhaps in the chronic condition of anorexia). Avoiding carbohydrates will lower insulin levels even in the obese, and so ameliorate the hyperinsulinemia that causes the carbohydrate craving itself. “After a year to eighteen months, the appetite is normalized and the craving for sweets is lost,” said James Sidbury, Jr., about the effects on children of his carbohydrate-restricted diet. “This change can often be identified within a specific one to two week period by the individual.”
If the more easily digestible carbohydrates are indeed addictive, this changes the terms of all discussions about
