Why We Get Fat

Gary Taubes argues that the conventional explanation for why people get fat is wrong. He contends that obesity is not caused by overeating or sedentary behavior but by dietary carbohydrates, which drive the hormone insulin to promote fat accumulation. The book's first part dismantles the prevailing "calories-in/calories-out" paradigm, and the second presents an alternative rooted in the hormonal regulation of fat tissue.

Taubes opens with Hilde Bruch, a German pediatrician who moved to New York in 1934 and was startled by the number of obese children she encountered during the Great Depression. At her Columbia University clinic, Bruch treated obese children who ate excessively but could not sustain eating less. Taubes cites George Cahill, a former Harvard Medical School professor, who summarized the book's central mechanism: "Carbohydrate is driving insulin is driving fat" (10).

In the first part, Taubes catalogs populations worldwide that were both extremely poor and beset by high rates of obesity. The Pima, a Native American tribe in Arizona, had widespread obesity by the early 1900s after transitioning from affluence to poverty and subsisting on government rations of white flour and sugar. Similar patterns appeared among the Sioux, women in postwar Naples, Zulu women in South Africa, and populations across Latin America and Africa. Malnutrition and obesity coexisted in the same families, a pattern Taubes argues the energy-balance hypothesis cannot explain without implausibly assuming impoverished mothers were starving their children to overeat.

Taubes examines undereating as a treatment. In the Women's Health Initiative, 20,000 women instructed to eat a low-fat diet reportedly consumed 360 fewer calories per day, yet after eight years lost an average of only two pounds. A 1959 review by psychologist Albert Stunkard and his colleague Mavis McLaren-Hume found results "remarkably similar and remarkably poor" (35). Bruce Bistrian of Harvard Medical School abandoned calorie-restricted treatment because patients always regained weight, concluding that undereating is "a way of temporarily reducing the most obvious symptom" (39).

Exercise fares no better. The obesity epidemic coincided with an explosion of leisure-time physical activity; health-club revenues grew seventeenfold between 1972 and 2005, yet Americans grew fatter. The 2007 joint guidelines of the American Heart Association and the American College of Sports Medicine conceded that evidence for exercise as a weight-control tool is "not particularly compelling" (44). A 1989 Danish study in which sedentary subjects trained to run marathons found that men lost five pounds of body fat over 18 months, and women lost none.

Taubes calculates that gaining 50 pounds over 25 years requires overeating by only 20 calories per day, making conscious calorie balancing implausible. Fat distribution differs by sex, runs in families, and can be altered by disease independent of food intake, suggesting hormones and genes determine how calories are partitioned. Taubes argues that the first law of thermodynamics is frequently misapplied: It describes what happens when someone gets fatter but not why, making the claim that overeating causes obesity a tautology. He further notes that energy intake and expenditure are dependent variables: Restricting food causes animals to reduce expenditure and increase hunger, so they regain weight when restriction ends.

The second part presents Taubes's alternative model. He describes George Wade's experiments at the University of Massachusetts: Female rats whose ovaries were removed became obese, but when placed on strict calorie restriction after surgery, they still became just as fat while becoming sedentary. Removing the ovaries eliminated estrogen, which normally suppresses lipoprotein lipase (LPL), an enzyme that pulls fat from the bloodstream into fat cells. Without estrogen, LPL pulled more fat from the bloodstream into storage. The animals did not get fat because they overate; they overate because they were getting fat.

From this and other research, Taubes formulates three "laws of adiposity." The first holds that body fat is carefully regulated by hormones and enzymes. The second states that a regulatory defect as small as 20 misdirected calories per day can produce obesity over decades. The third states that whatever makes a person fatter will also make that person overeat, because the body compensates for calories diverted into fat, just as growing children overeat because growth hormones demand it.

Taubes traces this framework to Gustav von Bergmann, who coined lipophilia (tissue affinity for fat) in 1908, and Julius Bauer of the University of Vienna, who argued that hormones regulate fat accumulation directly. This view was widely accepted in Europe by the late 1930s but vanished after World War II, when Louis Newburgh of the University of Michigan promoted the competing claim that obesity results from a "perverted appetite" (82).

The biological centerpiece of the book is insulin. Fat continuously flows in and out of fat cells, stored as triglycerides and mobilized as fatty acids. Insulin activates LPL on fat cells, suppresses hormone-sensitive lipase (HSL)—an enzyme that breaks down stored triglycerides inside fat cells so fatty acids can be released—and promotes glucose uptake, fat cell creation, and liver conversion of fatty acids to triglycerides. Nearly every other hormone works to release fat from storage; cortisol is a partial exception, promoting both storage and release through different mechanisms. Insulin overrides them all. Taubes cites the 1965 work of Salomon Berson and Rosalyn Yalow, researchers who pioneered the measurement of hormone levels in the blood, who found that releasing fat requires "only the negative stimulus of insulin deficiency" (125).

Because carbohydrates drive insulin secretion, Taubes identifies refined flour, starches, sugars, and liquid carbohydrates as the primary dietary causes of obesity. Fructose, found in table sugar, high-fructose corn syrup, and fruit, is metabolized in the liver, which converts much of it to fat and over time causes insulin resistance—a condition in which cells become less responsive to insulin's effects, forcing the pancreas to secrete still more. These fattening carbohydrates are also the cheapest calories available, explaining the link between poverty and obesity. Anthropological evidence shows that when isolated populations adopted Western foods, invariably sugar and white flour first, rates of obesity and diabetes followed.

Taubes refutes the claim that high-fat, low-carbohydrate diets cause heart disease. A 2001 Cochrane Collaboration assessment found "only limited and inconclusive evidence" (184) that modifying dietary fat prevents cardiovascular disease. Clinical trials comparing Atkins-type diets with low-fat diets consistently showed the high-fat groups lost more weight, raised HDL cholesterol, and lowered triglycerides. In the Stanford A TO Z study (2007), the Atkins group outperformed calorie-restricted, Ornish, and Zone diet groups; lead researcher Christopher Gardner, a 25-year vegetarian, called the results "a bitter pill to swallow" (192). Taubes connects these findings to metabolic syndrome, a cluster of interrelated metabolic abnormalities driven by insulin resistance, which leads to heart disease, diabetes, and potentially cancer and Alzheimer's disease.

Taubes notes that carbohydrate restriction for obesity is not new. The French lawyer and food writer Jean Anthelme Brillat-Savarin concluded in 1825 that starches, flours, and sugars caused obesity, and London undertaker William Banting's 1862 account of his weight loss on such a diet became an international bestseller. Through the mid-20th century, leading medical institutions prescribed carbohydrate-restricted diets as standard treatment. This consensus collapsed after the 1960s, when authorities concluded that dietary fat causes heart disease and classified obesity as an eating disorder.

In the final chapters, Taubes recommends permanently avoiding refined grains, starches, and sugars while eating freely of meat, fish, eggs, vegetables, and fat. He warns that carbohydrate cravings may function like an addiction, citing research by Bartley Hoebel of Princeton University showing that sugar activates the brain's reward center similarly to cocaine and nicotine. He stresses physician guidance for people with diabetes or on blood-pressure medications. An appendix from the Duke University Medical Center targets fewer than 20 grams of carbohydrates per day, and an afterword addresses why some carbohydrate-rich populations remain lean, which Taubes attributes to traditionally low sugar consumption.