The hypothesis that human obesity is caused by deficient thermogenesis has been proposed by many investigators throughout the 20th century. Supporting evidence was obtained from epidemiologic studies of dietary intake, animal models with aberrant brown adipose tissue (BAT) function, and genetic studies of human polymorphisms of genes involved in BAT function. Supporting evidence was also obtained from clinical studies of the thermogenic effect of meals, but these measures capture only a short portion of the day and may miss some of the thermogenic effect. To capture the effects throughout the day and to move the studies out of the metabolic ward, investigators have used the doubly labeled water (DLW) method to measure total daily energy expenditure. DLW studies have not supported the above hypothesis. Increases in total energy expenditure (TEE) during overfeeding have been small (0.9 +/- 0.8 MJ/d) and account for an average of only 18 +/- 18% of the excess energy intake. Most of this increase is in the resting metabolic rate. Moreover, these studies showed little variation in the changes in resting metabolic rate or in thermogenesis from meals during overfeeding. Instead, the component that is most variable and that accounts for the variability in weight gain during overfeeding is the energy expended in physical activity. This component of TEE deserves greater attention in future studies. These studies of thermogenesis have shown the importance of clinical research as part of a comprehensive approach to understanding the etiology of human obesity.
