Good Calories, Bad Calories

The common belief is that ex-smokers gain weight because they eat more once they quit. They will, but according to studies only in the first two or three weeks. After a month, former smokers will be eating no more than they would have been had they continued to smoke. The excess of calories consumed is not enough to explain the weight gain. Moreover, as Judith Rodin, now president of Rockefeller University, reported in 1987, smokers who quit and then gain weight apparently consume no more calories than those who quit and do not gain weight. (They do eat “significantly more carbohydrates,” however, Rodin reported, and particularly more sugar.) Smokers also tend to be less active and exercise less than nonsmokers, so differences in physical activity also fail to explain the weight gain associated with quitting.

The evidence suggests that nicotine induces weight loss by working on fat cells to increase their insulin resistance, while also decreasing the lipoprotein-lipase activity on these cells, both of which serve to inhibit the accumulation of fat and promote its mobilization over storage, as we discussed earlier (see Chapter 22). Nicotine also seems to promote the mobilization of fatty acids directly by stimulating receptors on the membranes of the fat cells that are normally triggered by hormones such as adrenaline. The drug also increases lipoprotein-lipase activity on muscles, and this may explain the steep rise in metabolic rate that occurs immediately after smoking. All of this fits with the observations that smokers use fatty acids for a greater proportion of their daily fuel than nonsmokers, and heavy smokers burn more fatty acids than light smokers. In short, nicotine appears to induce weight loss and fat loss not by suppressing appetite but by freeing up fatty acids from the fat cells and then directing them to the muscle cells, where they’re taken up and oxidized, providing the body with some excess energy in the process. When smokers quit, they gain weight because their fat cells respond to the absence of nicotine by significantly increasing lipoprotein-lipase activity. (There’s also evidence that the weight-reduction drug fenfluramine—the “fen” half of the popular weight-loss drug phen/fen, which was banned by the FDA in 1997—works in a similar manner, by decreasing lipoprotein-lipase activity in the fat tissue.)

This alternative hypothesis of obesity and its physiological perspective on hunger forces us to rethink virtually all our cherished notions about how weight changes and why. By this hypothesis, any long-term variations in weight, appetite, and energy expenditure—even our inclination to exercise or go for a walk—are likely to be induced at a fundamental level by changes in the regulation of fat metabolism and the partitioning and availability of metabolic fuels in the body. These in turn are driven, first and foremost, by changes in insulin secretion and how our fat and muscle tissue respond to that insulin. In this sense, insulin becomes what researchers who study hibernation and other seasonal weight variations in animals refer to as the adjustable regulator. Increase or decrease the circulating levels of insulin, and weight, hunger, and energy expenditure increase or decrease accordingly. It’s insulin that regulates the equilibrium between the forces of fat deposition and the forces of fat mobilization at the adipose tissue.

What’s been clear for almost forty years is that the levels of circulating insulin in animals and humans will be proportional to body fat. “The leaner an individual, the lower his basal insulin, and vice versa,” as Stephen Woods, now director of the Obesity Research Center at the University of Cincinnati, and his colleague Dan Porte observed in 1976. “This relationship has also been shown to occur in every commonly used model of altered body weight, including…genetically obese rodents and overfed humans. In fact, the relationship is sufficiently robust that it exists in the presence of widespread metabolic disorder, such as diabetes mellitus, i.e., obese diabetics have elevated basal insulin levels in proportion to their body weight.” Woods and Porte also noted that when they fattened rats to “different proportions of their normal weights,” this same relationship between insulin and weight held true. “There are no known major exceptions to this correlation,” they concluded. Even the seasonal weight fluctuations in hibernators agree with this correlation; the evidence suggests that annual fluctuations in insulin secretion drive the yearly cycle of weight and eating behavior, although this has never been established with certainty.