Rates of severe obesity (defined as a BMI ≥40) have increased faster than rates of more moderate obesity. Data from the Behavioral Risk Factor Surveillance System indicate that the prevalence of BMI ≥40 quadrupled between 1986 and 2000 (1). Although the reasons for the increase in severe obesity are not understood completely, current rates are consistent with general increases in the prevalence of obesity and a corresponding rightward shift in the overall distribution of population BMIs. Severe obesity is associated with high levels of medical (2) and psychiatric (3, 4) comorbidity as well as increased mortality and increased medical expenditures (5). Current nonsurgical treatments are regarded as inadequate for severe obesity, as they are usually associated with only moderate weight losses followed by weight regain.

In light of the limited success of medically supervised dietary interventions, behavior modification, and medication in the management of severe obesity, a National Institutes of Health (NIH) Consensus Panel (6) recommended bariatric surgery for well-informed, motivated obese patients with BMIs ≥40 or BMIs of 35–40 with significant obesity-related comorbidity who have acceptable risks for operation. Since that time the number of surgeries has proliferated (7). Growing evidence of the benefits, the development of new surgical procedures, and reductions in postoperative complications as a result of use of laparoscopic methods all have promoted the appeal of surgery.

Although numerous bariatric surgery procedures are in use, they can be divided into those that reduce food intake by gastric restriction only (e.g., laparoscopic adjustable gastric banding, vertical banded gastroplasty, and sleeve gastrectomy), those that bypass varying amounts of the upper intestine in order to reduce uptake of food in the digestive tract through malabsorption (e.g., biliopancreatic diversion and biliopancreatic diversion with duodenal switch), and combined restriction and bypass procedures (e.g., Roux-en-Y gastric bypass) (8). Postoperative course differs as a function of the type of surgical procedure. In particular, weight change patterns vary across surgical procedures. For example, while weight often stabilizes by about 18 months after Roux-en-Y surgery, weight loss after laparoscopic adjustable gastric banding is more gradual and may continue over a period of 2–3 years.

The benefits of surgery are well documented. A meta-analysis that included 22,094 patients who underwent a variety of different bariatric procedures (9) documented an overall percentage of excess weight loss of 61.2%, as well as resolution or improvement of diabetes, hyperlipidemia, hypertension, and obstructive sleep apnea. Similarly, a large prospective investigation of more than 4,000 obese persons that compared outcomes between those undergoing surgery and those undergoing conventional nonsurgical intervention (the Swedish Obese Subjects Study) found that bariatric surgery is associated with long-term weight loss, decreased all-cause mortality (10), and improved health-related quality of life (11).

Although bariatric surgery has unambiguous benefits for the majority of patients, there are risks. Operative mortality in experienced clinical centers is approximately 0.5% for gastric bypass and 0.1% for laparoscopic adjustable gastric banding (5). Short- and long-term surgical morbidity also varies as a function of type of operation. Perioperative morbidity occurs in about 5% of patients, and longer-term complications (e.g., stomal stenosis after gastric bypass and access port problems after laparoscopic gastric banding) are not uncommon (5).

A significant number of patients may have ongoing difficulties in losing weight or maintaining weight loss. If successful bariatric surgery outcome is defined as having lost at least 50% of excess weight, as many as 20% of patients appear to have a less than optimal outcome after surgery, including inadequate initial weight loss or significant weight regain after reaching a weight loss plateau (12). Long-term follow-up data are still limited. In the Swedish Obese Subjects Study (11), which includes many patients who underwent procedures that are no longer in use in the United States, a significant proportion of patients had experienced considerable weight regain at 10-year follow-up (11). Approximately 12% of patients maintained weight losses ≥30% of their baseline body weight, but 9% gained weight. In a recent report of longer-term follow-up (up to 5 years) of an Australian cohort (N=600) of patients who received an adjustable gastric band, the average percentage of excess weight loss was about 60% (N=228) at 3-year follow-up, but about 10% of patients lost less than 25% of excess weight (13).

Data documenting variability in outcome among bariatric surgery patients over time highlight the importance of patient selection and education, as well as the potential importance of pre- and postsurgical intervention. Below we review evidence bearing on the psychiatric evaluation and follow-up of bariatric surgery patients.

The NIH Consensus Panel (6) recommended careful selection of surgical candidates by a multidisciplinary team with access to psychiatric expertise. In a more recent consensus conference statement (5), a panel concluded that psychiatric evaluation was not needed routinely but should be available if indicated. Nevertheless, all major insurers in the United States currently require a comprehensive presurgical psychological/psychiatric evaluation as part of a mandatory workup before approving surgery.

There are compelling reasons for a comprehensive presurgical evaluation. A growing body of research evidence documents high rates of psychiatric disorders among candidates for surgery. In a recent investigation by our research group (3) using structured psychiatric interviews conducted by clinicians independent of the surgery approval process, we found that approximately 66% of surgery candidates had a lifetime history of at least one axis I diagnosis, 38% met criteria for a current axis I disorder, and 29% met criteria for an axis II disorder. Mood (15.6%), anxiety (24%), and binge eating (16.3%) disorders were the most common current disorders at the time of the presurgical evaluation. Current substance use problems were less common (1.7%), but lifetime substance use disorders were prevalent (32.6%). The most common axis II disorder was avoidant personality disorder (17%); avoidant behavior was frequently associated with a desire to evade social judgments and the stigma associated with severe obesity. Other investigators have documented high rates of night eating syndrome (14), body image disturbance (15), and childhood maltreatment (16).

Unsurprisingly, psychiatric treatment and use of psychotropic drugs among individuals seeking surgery also are common (17–19). For example, Friedman and colleagues (18) reported that 16% of a series of 837 patients were seeing a mental health professional at the time of evaluation for bariatric surgery, and 41% were taking psychotropic medications. Sarwer and colleagues (19) reported that 34.4% of 90 surgical candidates were taking psychiatric medication. The use of antidepressant medication was most common (30%), followed by anxiolytic drugs (6.6%) and antipsychotic medication (3.3%); most medications were prescribed by primary care physicians.

Despite the prevalence of psychiatric disorders and psychosocial problems among surgical candidates, evidence relating to the impact of these difficulties on postsurgical outcome has been equivocal and at times contradictory. Difficulties in interpreting the research evidence are due in part to methodological limitations of existing studies and the fact that there have been few well-designed longitudinal investigations. Overall, however, there is no consistent evidence that psychosocial variables predict postsurgical weight loss or mental health (20).

Considering available evidence and current clinical practice, we offer the following recommendations. First, there is a general consensus that a psychiatric disorder per se should not be an exclusion criterion for bariatric surgery. Nevertheless, there may be psychiatric reasons to delay or deny surgery. A recent survey of mental health clinicians found that surgery was delayed or denied for approximately 15% of surgery candidates for psychological reasons, primarily for untreated depression or a lack of understanding about the risks of and requirements for surgery (21). In the absence of robust predictors of poor surgical outcomes or alternative treatments for severe obesity, decisions to defer or deny surgery should be considered carefully. Wadden and Sarwer (22) reported recommending adjunctive counseling of some kind prior to surgery for 33% of a series of 90 patients but concluded that only 3% had clear contraindications for surgery.

Second, surgery candidates must be able to tolerate the risks of surgery and comply with a postoperative regimen that requires long-term medical surveillance and lifelong changes in diet and activity. Although surgery is a potent tool, weight loss will not be sustained over the long term if patients do not comply with recommendations to modify eating and exercise. There is increasing recognition that since surgery does not affect the behavioral or psychological factors that promote or maintain obesity, patients undergoing bariatric surgery will benefit from a thorough discussion of issues that may affect longer-term success and the need to commit to monitoring eating and exercise over time to optimize outcome.

Third, in light of the demands of the postsurgical regimen and the life changes consequent to major weight loss, discussion of patient expectations and consideration of the individual goals and interpersonal milieu are indicated. Patients may have unrealistic expectations regarding weight loss, and although body image concerns usually improve after surgery, some patients report dissatisfaction relating to excess skin or flabbiness. Psychosocial problems and marital conflict may not improve and may worsen after surgery (15). Thus, presurgical consultation provides an opportunity to clarify expectations regarding surgery and to identify psychosocial issues that may require attention over time.

In summary, although a comprehensive presurgical evaluation provides an opportunity to identify the small number of patients for whom surgery is contraindicated, for most individuals careful presurgical evaluation should serve a planning and education function rather than a gatekeeping function (23). The consultation provides an opportunity to review motivations for and expectations of surgery and to provide education. Individuals with psychiatric problems may benefit from treatment prior to surgery and from establishment of a plan for postsurgical monitoring and intervention.

Bariatric surgery requires a dramatic alteration in eating behavior. Gastric restriction leads to a drastic reduction in the amount of food eaten at any one time, and this restriction is the major mechanism of weight loss. Patients are directed to eat small meals, to chew all foods well, and to stop eating as soon as they feel full. Vomiting in response to feelings of fullness or food lodged in the upper digestive tract (“plugging”) is common during the first 6 months after surgery. The vomiting may be involuntary or self-induced, but self-induced vomiting for most patients serves to relieve discomfort rather than to promote weight loss.

Procedures that combine gastric restriction with malabsorption, such as gastric bypass, increase the risk of additional consequences, including nutrient deficiencies, “dumping syndrome,” and diarrhea. Dumping syndrome occurs in conjunction with rapid emptying of the gastric pouch after ingestion of foods high in refined sugar or carbohydrates. Early dumping occurs 30–60 minutes after eating and may cause sweating, light-headedness, palpitations, nausea, diarrhea, or cramping. Late dumping occurs 1–3 hours after eating as a result of reactive hypoglycemia. Dumping symptoms occur in up to 85% of patients (24). For many, these symptoms help them make appropriate food choices. For others, dumping symptoms are more persistent and aversive; management emphasizes nutrition counseling geared to helping patients make appropriate food choices. A small subset of postbypass patients may develop severe postprandial hypoglycemic symptoms months or years after surgery that are refractory to nutritional and medical management. In these cases, the rare possibility of insulinoma or nesidioblastosis (abnormal pancreatic islet morphology) must be considered (25).

Many patients will have problems with dietary adherence. Some patients adopt eating patterns such as frequent ingestion of high-calorie liquids or grazing (a pattern of continual eating of small amounts of food throughout the day) that are associated with less than optimal outcomes. For others, caloric intake increases gradually over time and contributes to the weight gain seen after a weight nadir is achieved (4). Ongoing nutrition counseling or behavioral weight management may be indicated for individuals who report potentially maladaptive eating behavior.

There have been reports of anorexia nervosa (26, 27) and bulimia nervosa (28) after surgery, but the onset of frank eating disorders appears to be uncommon (29). Some patients present with the onset of eating problems that do not meet DSM-IV criteria for an eating disorder but are associated with impairment and distress. Segal and colleagues (30) have proposed criteria for a surgery-related disorder, “postsurgical eating avoidance disorder.” The syndrome is characterized by anxiety and disturbed eating (purging, restriction, rapid weight loss, and body image dissatisfaction). Although there is no current empirical support for the syndrome, and it is unclear that it is in fact distinct from other eating disorders, it is important to recognize that surgery is associated with the onset of clinically significant eating problems for some patients. These individuals should be referred to clinicians with expertise in the management of aberrant eating.

Most interest has focused on the impact of binge eating on postsurgical outcome, perhaps because of the robust association between binge eating and obesity and the deleterious effect of uncontrolled eating on weight management after surgery. Studies reporting rates of binge eating disorder among surgical candidates have yielded prevalence rates ranging from 2% to 49%, but overall, available evidence indicates high rates of binge eating disorder (31). For example, in the Kalarchian et al. study (3), which used a structured clinical interview to assess binge eating disorder, 27% of surgery candidates reported a lifetime history of the disorder. However, investigations that have examined the impact of binge eating or binge eating disorder on postsurgical outcome have yielded inconsistent results (32). Existing studies have used different assessment methods and examined individuals receiving different surgical procedures, and the timing and duration of follow-up have varied widely.

One issue in the assessment of postsurgical binge eating behavior has to do with the fact that the DSM-IV definition of binge eating requires the ingestion of an “objectively large amount of food.” Although it appears that over the long term, at least some individuals resume objective binge eating, in the shorter term, postsurgical gastric restriction precludes the episodic ingestion of large amounts of food. Thus, investigators increasingly have focused on subjective binge eating, that is, the loss of control over eating an amount of food that may not be objectively large.

Despite limitations in the existing literature, accumulating evidence indicates that in many individuals binge eating resolves after bariatric surgery (33). However, binge eating that starts or reemerges after surgery is associated with less weight loss and with weight regain (31). Thus, postsurgical monitoring of binge eating problems is recommended, and adjunctive treatment with medication or psychotherapy is indicated for clinically meaningful binge eating problems.

There has been considerable interest in the impact of depression on postsurgical outcome. Most studies have examined the impact of presurgical depressive symptoms on short-term postsurgical outcome. This work also has yielded equivocal findings (34, 35). There is little information about the impact of syndromal depression or other psychiatric disorders on postsurgical course. Consequently, the effect of diagnosable disorders on bariatric surgery outcome is poorly understood. Our research group recently reported on the relationship between preoperative psychiatric disorders and 6-month outcomes after gastric bypass (36). Findings indicated that a lifetime history of mood or anxiety disorders was associated with poorer short-term weight loss. Although absolute differences in weight loss between those who had a history of disorders and those who did not were modest, these data suggest that postsurgical monitoring of depression and anxiety, as well as binge eating, may be important.

Recent media attention has focused on anecdotal reports of alcohol misuse (or other “addictive” behaviors, such as gambling, compulsive shopping, and driven sexual behavior) after bariatric surgery (37), but research evidence is lacking. There are good reasons, however, to monitor use of alcohol after surgery. First, lifetime rates of substance use disorders among candidates for bariatric surgery are substantial, but rates of current substance use disorders prior to surgery are low (3), suggesting that there might be some vulnerability to substance use problems among individuals who undergo surgery. Moreover, gastric bypass alters the metabolism of alcohol, which may enhance its effects, and suggests the need for patients to exercise caution when drinking (38). Thus, although additional research is necessary to understand the relationship between substance use and eating problems, it seems prudent to monitor patient alcohol use after bariatric surgery.

Finally, although bariatric surgery is associated with decreases in long-term mortality from diabetes, heart disease, and cancer, the risk of death from non-disease-related causes, including suicide, is higher among bariatric surgery patients than other severely obese individuals (39). Increased rates of suicide might reflect the high rates of preexisting psychiatric illness in cohorts of surgery patients or be related to the consequences of surgery on the lives of severely obese individuals. Prospective research is needed to document the short- and longer-term effects of psychiatric status on surgery outcome as well as to examine the impact of surgery on the psychiatric status and quality of life of surgery patients.

We recommend that the surgical team routinely monitor all patients after surgery for the onset or recurrence of eating, mood, or substance use problems, as well as other psychiatric symptoms, and refer patients who report significant symptoms for further mental health evaluation and treatment. Careful attention to the psychiatric and psychosocial status of bariatric surgery patients is likely to enhance individual well-being and minimize the potentially negative impact on postoperative weight loss.

Patients seeking bariatric surgery are usually required to participate in an assessment to provide psychological/psychosocial clearance prior to surgery. Conduct of presurgical assessments varies widely (22, 40), and there are few clear-cut contraindications to surgery. There is consensus that the presence of current acute or inadequately managed mental illness, active substance abuse or dependence, inability to participate in informed decision making, and unwillingness to comply with postsurgical protocol indicate a need to defer or deny surgery. Identification of contraindications to surgery is only one function of a thorough assessment, however. The consultation provides the opportunity to work with the patient to clarify expectations, provide education, identify potential barriers to compliance, and plan for postsurgical follow-up.