Surgery for obesity: Panacea or pandora's box?

Rahul Pannala, Mark Kidd, Irvin M. Modlin

Research output: Contribution to journalReview articlepeer-review

11 Scopus citations


Background: Given the increasing prevalence of obesity, healthcare providers should be cognizant of various non-operative (diet, exercise, behavior therapy, and medications) and operative obesity treatments. This review critically evaluates these treatments, especially bariatric surgeries. Methods: Medline analyses using a combination of the following terms: obesity, bariatric surgery, and outcomes were performed with particular emphasis on prospective studies and randomized trials. Results: Non-operative treatments result in modest sustained weight loss (5-8%) at one year. Surgery is recommended for those with BMI >40 or >35 with comorbidities. Laparoscopic adjustable gastric banding, a restrictive procedure, causes 35-54% excess weight loss (EWL) at 1 year. Malabsorptive procedures (biliopancreatic diversions with and without duodenal switch) induce 72-77% EWL but are only performed at few centers. Roux-en-Y gastric bypass, acting through a combination of restriction and malabsorption, results in 69% EWL at 1 year and 49% at 14 years. Each procedure is associated with unique anatomic and nutritional complications. Overall, operative treatment improves comorbidities and may improve all-cause mortality. Conclusions: Surgery is an effective long-term treatment for selected obese patients who have failed other treatments. Further research is needed on prospective comparisons of procedures, evaluation of long-term outcomes, especially between centers and increasingly unrealistic patient expectations.

Original languageEnglish (US)
Pages (from-to)1-11
Number of pages11
JournalDigestive Surgery
Issue number1-2
StatePublished - Aug 1 2006


  • Bariatric surgery
  • Obesity, surgery

ASJC Scopus subject areas

  • Surgery
  • Gastroenterology


Dive into the research topics of 'Surgery for obesity: Panacea or pandora's box?'. Together they form a unique fingerprint.

Cite this