OBESITY, DEFINED AS 20% OR more than the ideal weight or body mass index (BMI, calculated as weight in kilograms divided by the square of height in meters) of 30 or more, has reached epidemic levels in the United States, affecting more than 30% of adults. Annual direct costs for treating obesityrelated medical illnesses have been estimated at nearly $51.6 billion; the annual US expenditure on weight reduction exceeds $30 billion. It was recently estimated that the prevalence of obesity in US adults increased by 8% during the past decade. In the 21st century, obesity may be the number 1 US public health problem (http://www .surgeongeneral.gov/topics/obesity/). Morbid or severe obesity was traditionally defined as a weight of 45 kg or more or 100% over ideal body weight defined by standard life insurance tables. More recent classification systems define morbid obesity as a BMI of 40 or more or a BMI of 35 or more in the presence of comorbidities. The prevalence of severe obesity based on US 19992000 population data was estimated to be 3.1% in men and 6.7% in women. Inapublishedreportofmortalityrates in patients who were morbidly obese, mortality was 12 times higher in men aged 25 to 34 years and 6 times higher in men aged 35 to 44 years vs men with healthy weight of the same age. Because even modest weight loss (10%-15% of initialweight)usuallyresults inimprovement or resolution of multiple medical comorbidities, surgical treatment of severe obesity appears to be cost-effective by eliminating use of medications and absenteeism from work in patients who were previously morbidly obese. Gastric Restrictive Operations There are 3 types of gastric restrictive operations: stapled gastroplasty, gastric banding, and conventional Rouxen-Y gastric bypass. In stapled gastroplasty, the stomach is partitioned close to the gastroesophageal junction creating a small-capacity upper gastric pouch with a small calibrated outlet leading from the upper pouch to the remainder of the digestive tract (FIGURE, A). Early weight loss results after banded gastroplasty have been acceptable with a mean excess weight (difference between preoperative and ideal weight) loss reported in the range of 60% (or approximately 30% of initial preoperative weight lost). The early morbidity rate (within the first 30 days) for banded gastroplasty is less than 10% and the perioperative mortality rate is less than 1.0%. However, many patients regain a substantial portion of their lost weight between 3 and 5 years, postoperatively. Vertical-banded gastroplasty has been shown to adversely alter postoperative eating behavior toward sweets and ice cream, which, in part, explain inferior weight loss results in comparison with Roux-en-Y gastric bypass; a nearly 80% failure rate has been reported 10 years following vertical-banded gastroplasty. Gastroplasty has recently fallen into disfavor due to poor weight loss maintenance and a 15% to 20% rate of reoperation for either stomal outlet stenosis or severe gastroesophageal reflux. Gastric banding uses a premeasured prosthetic device to restrict oral intake (Figure, B). The circumference of the band is generally in the range of 5.0 cm, similar to the measurement used for gastroplasty. Kuzmak introduced an inflatable band in which the diameter of the band can be adjusted by infusion of saline through a subcutaneous reservoir. Complication rates with the early techniques of gastric banding were relatively high due to stenosis, erosion, or both of the band, frequently requiring reoperation. Weight loss results and complication rates with the inflatable bands are better than those observed after the earlier banding techniques. Although erosion of the inflatable bands is relatively uncommon, usability of the subcutaneous reservoir deteriorates over time. Weight loss with gastric banding has been less consistent than weight loss reported after banded gastroplasty and gastric bypass. Gastric bypass combines gastric restriction with a small amount of subclinical malabsorption. Although absorption of iron and vitamin B12 is decreased after gastric bypass, malabsorption of protein, carbohydrate, and fat has not been reported after gastric bypass performed with short Roux limb lengths. In gastric bypass, the upper stomach is completely closed off, thereby excluding more than 95% of the stomach, all of the duodenum, and 15 to 20 cm of proximal jejunum from digestive continuity. The Roux-en-Y technique (Figure, C) is currently the preferred method of gastric bypass. Weight loss results with the Rouxen-Y technique have been superior to thoseobservedafterothergastric restrictive operations in multiple clinical comparisons. Mean excess weight loss in gastric bypass patients typically ranges from 65% to 75%, which corresponds to loss of approximately 35% of initial weight. However, there is some degree of recidivism between 3 and 5