CLINICAL RESEARCH STUDY

CLINICAL RESEARCH STUDY Incremental Weight Loss Improves Cardiometabolic Risk in Extremely Obese Adults William D. Johnson, PhD, Meghan M. Brashear, MPH, Alok K. Gupta, MD, Jennifer C. Rood, PhD, Donna H. Ryan, MD Pennington Biomedical Research Center, Baton Rouge, La. ABSTRACT OBJECTIVE: Excessively obese adults often acquire many metabolic disorders that put them at high risk for developing type 2 diabetes mellitus and cardiovascular disease. We investigated the hypothesis that cardiometabolic risk in a primary care cohort of 208 excessively obese adults (body mass index 40-60 kg/m 2, 48 with type 2 diabetes mellitus) would deteriorate with additional weight gain and improve incrementally beginning with 5% weight reduction. METHODS: Further analysis of the Louisiana Obese Subjects Study of excessively obese patients enrolled and followed during 2005-2008 is reported. RESULTS: Weight loss correlated significantly with improvements in fasting plasma glucose, triglycerides, high- and low-density lipoprotein cholesterol, uric acid, alanine aminotransferase, lactate dehydrogenase, and high-sensitivity C-reactive protein. Most parameters deteriorated with weight gain and progressively improved with 5% or more weight loss. Except for low-density lipoprotein cholesterol, all risk factors significantly improved with 20% loss of body weight. Among patients who had not been diagnosed with type 2 diabetes mellitus and had normoglycemia at baseline, median fasting plasma glucose increased significantly (13%) with stable or gained weight at 1 year, but did not change significantly with reduced weight. Although glucose levels did not change significantly in patients with type 2 diabetes mellitus who gained weight, a decline beginning after 5% weight reduction culminated in 25% glucose reduction with 20% weight loss. Resting blood pressure declined independently of weight change. CONCLUSION: Very obese adults can improve their cardiometabolic risk under primary care weight management. Incremental success may help motivate further therapeutic weight reduction. 2011 Elsevier Inc. All rights reserved. The American Journal of Medicine (2011) 124, 931-938 KEYWORDS: Cardiometabolic benefits of weight reduction; Diabetes; Pragmatic weight loss for class III obesity SEE RELATED EDITORIAL p. 891 This research was presented at the 11th International Congress on Obesity, July 11-15, 2010, Stockholm, Sweden. Funding: The Louisiana Obese Subjects Study was funded by the Office of Group Benefits, the health insurance provider for Louisiana state employees. Some of the sibutramine used in the Louisiana Obese Subjects Study was donated by Abbott Laboratories (Columbus, OH). Conflict of Interest: None of the authors have any conflicts of interest associated with the work presented in this manuscript. Authorship: All authors had access to the data and played a role in writing this manuscript. Trial registration: ClinicalTrials.gov NCT00115063. Requests for reprints should be addressed to William D. Johnson, PhD, Pennington Biomedical Research Center, 6400 Perkins Road, Baton Rouge, LA 70808-4124. E-mail address: William.Johnson@pbrc.edu Adults in the United States are usually classified as obese if their body mass index (BMI) is 30 kg/m 2 or greater, where BMI is calculated as weight in kilograms divided by height in meters squared. Moreover, obesity is classified as class I if 30 BMI 35, class II if 35 BMI 40, and class III if BMI 40. In 2007 and 2008, an estimated 33.9% of US adults aged 20 years were obese and 5.7% were class III or extremely obese (95% confidence interval, 4.9-6.5). The prevalence of extreme obesity was higher in women (7.9%) than in men (4.2%) and higher in non-hispanic blacks (11.1%) than in Mexican Americans (6.0%) and non-hispanic whites (5.2%). 1 The concern regarding the increasing prevalence of extreme obesity is for the greatly increased risk it imposes for developing morbidities, such as type 2 diabetes, 2,3 and for its impact on mortality. Indeed, it is informative to consider 0002-9343/$ -see front matter 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2011.04.033

932 The American Journal of Medicine, Vol 124, No 10, October 2011 the impact of extreme obesity on years of life lost. Fontaine et al 4 estimated on average that 13 years of life in white men and 8 years of life in white women were lost as the result of having a BMI 45 at ages 20 to 30 years. Although African Americans did not demonstrate reduction in life expectancy from moderate obesity at younger ages, an estimated 20 years of life CLINICAL SIGNIFICANCE Precipitous weight loss in class III obese patients can be successfully managed with pragmatic primary care. Triglyceride and HDL-cholesterol levels demonstrate progressive incremental improvement after 5% and 10% weight reduction, respectively. Uric acid, alanine aminotransferase, and lactate dehydrogenase levels all decline with weight loss. Fasting glucose levels decline progressively in severely obese diabetic patients with 5% to 10%, 10% to 20%, and 20% weight reduction. in men and 5 years in women were lost as a result of having a BMI 45. If obesity has implications for increased health risks, does weight reduction benefit patients? Weight reduction achieved through surgery has been linked with reversal of hypertension and diabetes 5 and with mortality reduction. 6 Weight reduction achieved through lifestyle change has been shown to improve many cardiovascular risk factors and quality of life, 7 and reduce the incidence of type II diabetes mellitus in those with impaired glucose tolerance. 8 Severely obese adults may be candidates for bariatric surgery, and although surgery is effective, 5,6,9,10 it is expensive and has obvious risks for complications; thus, there is need for effective interventions in primary care practice. Severely obese adults have had moderate success with weight loss programs aimed at lifestyle modifications that target education, diet, and physical activity. 11-13 Only recently, 12 a lifestyle intervention was found to be effective in individuals with BMI 35 kg/m 2, both in terms of achieving weight loss ( 8%) and associated improvements in risk factors. However, little other evidence exists to draw conclusions about effectiveness of specific weight loss increments for enhancing health. The Louisiana Obese Subjects Study (LOSS) 14 demonstrated that with brief training, primary care practices can manage effective intensive medical weight loss programs for patients who are extremely obese. The LOSS recommended evidence- and guidelines-based and US Food and Drug Administration-approved treatments, and mimicked real practice, where physicians and patients could negotiate treatment choices; included relatively unselected patients with class III obesity; and used diverse primary care practice sites and practitioners. As reported previously, 14 intensive medical weight loss intervention provided significantly greater weight reduction compared with usual care intervention. Most metabolic health outcomes improved differentially with intervention after 1 year. This current additional report investigated the hypothesis that graduated increases in weight gain or loss among severely obese patients, irrespective of type or efficacy of intervention, translate into parallel incremental deterioration or improvement in blood pressure, fasting plasma glucose, highdensity lipoprotein (HDL) and low-density lipoprotein (LDL) cholesterol, triglycerides, uric acid, alanine aminotransferase, lactate dehydrogenase, and high-sensitivity C-reactive protein at 1-year follow-up. MATERIALS AND METHODS Source of Data The LOSS was a 2-year randomized, controlled trial conducted in 7 primary care practices and 1 research clinic to compare efficacy with intensive medical weight loss intervention against usual care for 390 extremely obese (40 BMI 60 kg/m 2 ) men and women. The 208 adults in the combined intervention groups who had 1-year blood work assessments provided the data source for investigating the impact of weight change (no change, categorized percent gain or loss) on percent change in blood pressure, fasting plasma glucose, HDL and LDL cholesterol, triglycerides, uric acid, alanine aminotransferase, lactate dehydrogenase, and high-sensitivity C-reactive protein. The Office of Group Benefits, insurer for more than 100,000 State of Louisiana employees and their dependents, approached Pennington Biomedical Research Center to develop treatments for class III obesity that would be deliverable in primary care physicians offices. The Office of Group Benefits selected 8 sites (Alexandria, Baton Rouge, Hammond, Lafayette, Lake Charles, Monroe, New Orleans, and Shreveport), and Pennington Biomedical Research Center identified physicians. The institutional review boards at each site approved protocol and consent forms. A data safety monitoring board approved the protocol and monitored serious adverse events. Weight Loss Interventions As previously described, 14 participants randomized to usual care were instructed in use of the Mayo Clinic Weight Management Web site (http://mayoclinic.com/health/ weightloss/my00432). Those randomized to intensive medical weight loss intervention followed guidelines for general treatment, whereby deviations were allowed for participant requests and physician discretion. The intensive medical weight loss intervention recommended the following. In phase 1, participants began with a low-calorie liquid diet plus 10 g of added fat (choice of 2 teaspoons of vegetable oil or ten 1-g fish oil capsules). The study dispensed, at no cost, powered HealthOne formula (Health and Nutrition Technology Inc, Carmel, CA) and recommended consumption of 3 shakes per day, providing 890 kcal/d, 75 g protein,

Johnson et al Weight Loss Improves Risk 933 13 g fat, and 110 g carbohydrates. Electrocardiograms and electrolytes were obtained biweekly during the liquid diet. Phase 1 could continue for 12 weeks, but if the liquid diet could not be tolerated, participants could proceed to phase 2. During the 4 months of phase 2, a highly structured diet and medication were recommended along with behavioral counseling in group sessions held weekly for 1 month and then biweekly for 3 months. Physician visits occurred monthly. The recommended diet consisted of 2 meal replacements (HealthOne, Slim fast [Unilever, Englewood Cliffs, NJ], Glucerna [Abbott Nutrition, Abbott Laboratories, Columbus, Ohio], Boost [Nestle Health Care, Nutrition, Inc, Minneapolis, Minn], or other commercial product) along with 2 portion-controlled snacks and 1 structured meal per day. The diet was approximately 1200 to 1600 kcal/d. Sibutramine hydrochloride and orlistat were dispensed to aid weight loss and maintenance, but for some patients diethylpropion hydrochloride was dispensed for intermittent use. An estimated 60% of patients in intensive medical weight loss intervention used weight loss medications. Sibutramine was recommended preferentially as the first-line therapy and was the most commonly dispensed drug. Metformin was recommended as the oral medication for type 2 diabetes mellitus, and venlafaxine hydrochloride and bupropion hydrochloride were recommended for depression because these drugs do not promote weight gain. For behavioral intervention, participants received education in weight management, physical activity, self-monitoring, stimulus control, social support, contingency management problem solving, and relapse prevention. Beginning in phase 2, group sessions were conducted for walking, water exercise, and weight training. Phase 3 included activities recommended for months 8 to 24. Weight loss medications, 1 daily meal replacement, and monthly group sessions were continued. Medications were used as needed, and repeated low-caloric liquid diets were allowed in 4- to 12-week episodes. Participants in the intensive medical weight loss intervention group lost an average of 13.1% of their baseline weight, whereas those in the usual care group lost an average of 0.9%. Measured Assessments Weight was measured twice per assessment using a calibrated office scale with a digital display. Participants were weighed post-voiding and wearing light clothing without shoes. Blood pressure was obtained with an appropriately sized cuff using a standard mercury sphygmomanometer or electronic blood pressure monitor. Fasting plasma glucose, HDL cholesterol, triglycerides, uric acid, alanine aminotransferase, and lactate dehydrogenase were analyzed on a Beckman-Coulter DXC 600 Pro (Beckman-Coulter Inc, Brea, Calif) using standard spectrophotometric assays. LDL cholesterol was calculated using the Friedewald equation. High-sensitivity C-reactive protein was analyzed on a Siemens 2000 (New York, NY). Categories of Weight Change Used to Assess Health Benefits Health benefits were assessed at year 1 for different weightchange categories: (1) gain ( 2%, n 30); (2) stable ( 2%, n 37); (3) minimal loss (2%-4.9%, n 24); (4) modest loss (5%-9.9%, n 41); (5) substantial loss (10.0%- 19.9%, n 44); and (6) big loss ( 20%, n 32). Statistical Analysis Statistical analyses were conducted using SAS version 9.1 (SAS Institute, Inc, Cary, NC). Spearman s correlations were calculated to assess associations between change in weight and change in each risk factor. Significance was declared if P.05 and positive or negative direction of correlations identified factors beneficially associated with larger percentages of weight loss. Median change in each risk factor at 1 year was plotted against categorized percent change in weight. Patients with diabetes mellitus included those who had been diagnosed by a physician or had baseline fasting plasma glucose level greater than 125 mg/dl. RESULTS Participants were predominantly white women, and median (interquartile range) was 49.5 (11.0) years for age and 46.1 (8.0) kg/m 2 for BMI (Table 1). Weight gain was observed in 14.9% of patients, weight stability was observed in 17.8% of patients, minimum loss was observed in 11.5% of patients, modest loss was observed in 19.7% of patients, substantial loss was observed in 20.7% of patients, and big loss was observed in 15.4% of patients. Table 2 shows that median weight decreased by 8 kg or 5.9% (P.0001) during the 1-year period even though the analytic sample was a mixture of weight-loss groups. Significant improvements were observed for all listed cardiometabolic risk factors except LDL cholesterol and fasting plasma glucose. Although the median for uric acid decreased significantly, the decrease was not clinically important. Alanine transaminase, lactate dehydrogenase, and high-sensitivity C-reactive protein all demonstrated significant improvements. As Table 1 Characteristics of the Sample N 208 Usual care (N) 42.8% (89) Intensive care (N) 57.2% (119) Type 2 diabetes mellitus (N) 23% (48) Male (N) 14.4% (30) White (N) 78.9% (164) African American (N) 20.7% (43) Hispanic (N) 0.5% (1) Age, y (median [IQR]) 49.5 (11.0) Baseline BMI, kg/m 2 (median [IQR]) 46.1 (8.0) Baseline weight, kg (median [IQR]) 126.6 (22.5) IQR interquartile range.

934 The American Journal of Medicine, Vol 124, No 10, October 2011 Table 2 Changes in Cardiometabolic Risk Factors (Median [Interquartile Range]) Measured Assessment Sample Size Baseline Year 1 Change* P Value Weight, kg 208 126.6 (22.5) 114.2 (26.1) 8.0 (19.1).0001 Type 2 DM 48 128.2 (21.1) 119.4 (25.4) 7.6 (10.1).0001 Non type 2 DM 160 125.8 (22.0) 113.6 (24.0) 8.1 (20.7).0001 Weight loss, % 208 5.9 (14.4).0001 Type 2 DM 48 5.6 (8.5).0001 Non type 2 DM 160 6.1 (17.1).0001 Systolic blood pressure, mm Hg 208 132.0 (20.5) 123.0 (21.5) 7.0 (24.5).0001 Diastolic blood pressure, mm Hg 208 80.0 (15.0) 76.5 (13.5) 3.5 (15.5).0001 Fasting glucose, mg/dl Type 2 DM 37 132.0 (52.0) 117.5 (44.0) 3.0 (35.0).2018 Non type 2 DM 150 94.0 (15.0) 93.0 (16.0) 0 (14.0).5846 HDL cholesterol, mg/dl 193 49.5 (16.4) 51.0 (13.9) 0.85 (9.7).0417 LDL cholesterol, mg/dl 190 115.1 (41.5) 115.5 (42.4) 0.6 (23.5).8997 Triglycerides, mg/dl 193 137.0 (86.0) 113.0 (86.0) 21.0 (61.0).0001 Uric acid, mg/dl 194 5.7 (1.9) 5.6 (2.0) 0.1 (1.0).0201 Alanine transaminase, U/L 194 22.0 (10.0) 20.0 (9.0) 2.0 (10.0).0001 Lactate dehydrogenase, U/L 194 159.0 (39.0) 155.0 (36.0) 5.0 (28.0).0007 High-sensitivity C-reactive protein 165 7.5 (11.6) 6.7 (9.1) 1.6 (4.7).0001 DM diabetes mellitus. *Median of the change from baseline to year 1. Not all participants submitted to blood tests at year 1 visit. Testing hypothesis that change is equal to zero. shown in Table 3, except for blood pressure, improvements in all responses were significantly correlated with weight loss. The correlation was especially strong in favor of C- reactive protein improvement as weight declined (r S 0.53, P.0001). Figure 1A indicates change in resting systolic and diastolic blood pressures decreased inconsistently across weight loss categories. Figure 1B portrays a precipitous decrease in median percent change in fasting plasma glucose levels among patients with type 2 diabetes mellitus who achieved at least modest weight loss. Although median Table 3 Spearman Correlation Coefficients Percent Change in Measured Assessment Sample Size Percent Weight Loss P Value Systolic BP 208 0.13.0642 Diastolic BP 208 0.01.9014 Fasting plasma glucose Type 2 DM 37 0.43.0078 Non type 2 DM 150 0.41.0001 HDL 193 0.31.0001 LDL 190 0.14.0493 Triglycerides 193 0.33.0001 Uric acid 194 0.43.0001 Alanine transaminase 194 0.22.0021 Lactate dehydrogenase 194 0.31.0001 High-sensitivity C-reactive protein 165 0.53.0001 BP blood pressure; DM diabetes mellitus. percent change in fasting plasma glucose increased in patients with normal fasting plasma glucose who gained or maintained stable weight, it declined as much as 6.2% only among those who realized at least 20% weight loss. Figure 2A reveals median percent change in HDL cholesterol increased 6.9% among those who lost substantial weight (5%- 9.9%) and 18.2% in those who lost at least 20%. Percent change in LDL cholesterol was significantly correlated with percent weight loss (r S 0.14, P.0001), but as seen in Figure 2B, change was inconsistent across weight loss categories. Figure 3A shows median percent change in triglyceride concentrations declined approximately 20% with 5% to 20% weight loss, and approximately 38% with 20% or more weight loss. Three patients (1.5%) had elevated uric acid levels at study entry, and 3 patients (1.5%) had high levels at 1 year (normal 3.5-8.5 mg/dl). Conversely, 4 patients (1.9%) initially had below-normal levels at study entry versus 7 (3.6%, range 2.8-3.4 mg/dl) at 1 year. Figure 3B demonstrates that median percent change in uric acid decreased 5.7% with substantial weight loss (5%-9.9%) and 16.6% with 20% or more weight loss. Fourteen patients (6.7%) had elevated alanine aminotransferase levels at study entry, whereas only 5 patients (2.6%) had high levels at year 1 (normal range 45 U/L). The median percent change in alanine aminotransferase declined by 10.5% with modest weight loss and decreased even further with 10% or more weight loss. Improvement in lactate dehydrogenase was realized by patients who accomplished at least modest weight loss.

Johnson et al Weight Loss Improves Risk 935 Figure 1 A, Median percent (%) change in blood pressure versus categories of percent change in weight at 1 year. B, Median percent (%) change in fasting serum glucose versus categories of percent change in weight at 1 year. BP blood pressure. DISCUSSION This ancillary report evaluates the efficacy of differing degrees of weight loss in 208 class III obese patients for improving fasting plasma glucose, triglycerides, HDL and LDL cholesterol, uric acid, alanine aminotransferase, lactate dehydrogenase, high-sensitivity C-reactive protein concentrations, and blood pressure. Results support the value of modest weight loss, because our study showed severely obese patients with type 2 diabetes mellitus may achieve 10.5% improvement in fasting plasma glucose levels with modest weight loss. Overall, with modest weight loss, patients may achieve 22% improvement in triglyceride levels, 10.5% improvement in alanine aminotransferase, and 5.7% improvement in lactate dehydrogenase; however, substantial weight loss may be necessary to realize improvement in uric acid and HDL-cholesterol levels. Blood pressures and LDL-cholesterol levels showed a proclivity to improve with weight loss, but there were confounding inconsistencies across incremental categories for weight change. Many patients were taking at least 1 medication for metabolic disorders, and the inconsistencies may be related. However, the lack of LDL response in those who lost more than 20% of their weight suggests the need for confirmatory studies in other extremely obese populations and supporting mechanistic explanations. Our results are in general agreement with previous findings of beneficial effects of weight loss in severely obese individuals. Sjostrom et al 5 found 2-year incidences of hypertension, diabetes mellitus, hyperinsulinemia, hypertriglyceridemia, and low HDL cholesterol were all improved significantly in severely obese patients who underwent surgical treatment for obesity compared with those randomized to a control group. Obese women who underwent Rouxen-Y gastric bypass surgery demonstrated improvements after 1-year in glucose, insulin, HDL and LDL cholesterol, and triglyceride levels. 15 Busetto et al 10 observed modest but significant correlations for improvement in fasting plasma glucose, triglycerides, and HDL-cholesterol levels as percentage weight loss increased, after laparoscopic gastric banding. Analogous significance was not found for resting systolic or diastolic blood pressure. Samaha et al 13 compared low-carbohydrate and low-fat diets for weight

936 The American Journal of Medicine, Vol 124, No 10, October 2011 Figure 2 A, Median percent (%) change in HDL cholesterol versus categories of percent change in weight at 1 year. B, Median percent (%) change in LDL cholesterol versus categories of percent change in weight at 1 year. LDL lowdensity lipoprotein; HDL high-density lipoprotein. loss with severe obesity and found that both diets produced beneficial effects on triglycerides, insulin, and insulin sensitivity; glucose levels improved in patients with diabetes mellitus, but improvement was not observed among those in whom diabetes had not been diagnosed. Goodpaster et al 12 investigated lifestyle dietary interventions combined with initial or delayed physical activity and found both significantly reduced blood pressure, fasting insulin, and homeostasis model assessment of insulin resistance. It is generally accepted that overweight and obese adults who lose more than 5% tend to experience improvement in various cardiometabolic parameters, 16 but the favorable evidence for modest weight loss in extremely obese patients was not as compelling for some metabolic parameters in LOSS and other studies reviewed in this article. A novel aspect of our findings is our assessment of change in cardiometabolic parameters across a broad spectrum of specific categories of percent change in weight from baseline to 1 year. This approach allowed us to identify levels of weight loss where metabolic improvements occurred. A limitation of LOSS is that only 53% of randomized patients completed assessments at 1 year. The issue of poor retention in weight loss studies is well known, 8 and it takes special efforts to achieve excellent adherence to follow-up protocol. The lesson for primary care physicians is that only half of very obese patients who enter weight loss programs may stay in the program through its completion. Although improvements in various cardiometabolic parameters were observed among 1-year protocol completers, investigators were unable to ascertain data for the noncompleters. This suggests that severely obese patients may struggle to comply with both usual care and intensive medical intervention for weight reduction. Nevertheless, these results also provide evidence that weight loss can be achieved by most patients who remain compliant for 1 year, and, moreover, those who do achieve weight reduction realize improvements in their measures for cardiometabolic risk. It seems intuitive that severely obese adults should seek substantial weight reduction and manage their weight loss under primary care supervision.

Johnson et al Weight Loss Improves Risk 937 Figure 3 A, Median percent (%) change in triglycerides versus categories of percent change in weight at 1 year. B, Median percent (%) change in uric acid versus categories of percent change in weight at 1 year. CONCLUSIONS Although some patients achieved substantial and big weight loss, our data suggest that modest weight loss (5%-9.9%) across a 1-year period is an appropriate short-term goal for men and women who are severely obese. The benefits of achieving modest weight loss include an expectation for improved triglycerides and HDL-cholesterol levels. Severely obese patients with diabetes mellitus should expect improved fasting plasma glucose levels after modest weight loss with continued improvement with additional weight loss. Long-term trials for severe obesity are needed to advance the pragmatic nonsurgical strategies for achieving significant weight reduction that would be sustainable throughout the remaining life span. We showed that weight loss and positive changes in several measures of cardiometabolic risk can be achieved in severely obese individuals in a primary care setting. These data further demonstrate that health care providers should be encouraging lifestyle interventions to bring about health changes similar to those seen in this study. References 1. Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999-2008. JAMA. 2010;303:235-241. 2. Chan JM, Rimm EB, Colditz GA, Stampfer MJ, Willett WC. Obesity, fat distribution, and weight gain as risk factors for clinical diabetes in men. Diabetes Care. 1994;17:961-969. 3. Colditz GA, Willett WC, Rotnitzky A, Manson JE. Weight gain as a risk factor for clinical diabetes mellitus in women. Ann Intern Med. 1995;122:481-486. 4. Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA. 2003;289:187-193. 5. Sjostrom CD, Lissner L, Wedel H, Sjostrom L. Reduction in incidence of diabetes, hypertension and lipid disturbances after intentional weight loss induced by bariatric surgery: the SOS Intervention Study. Obes Res. 1999;7:477-484. 6. Sjostrom L, Narbro K, Sjostrom CD, et al. Effects of bariatric surgery on mortality in Swedish obese subjects. N Engl J Med. 2007;357:741-752. 7. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: executive summary. Expert Panel on the Identification, Evaluation, and Treatment of Overweight in Adults. Am J Clin Nutr. 1998;68:899-917. 8. The Diabetes Prevention Program (DPP): description of lifestyle intervention. Diabetes Care. 2002;25:2165-2171.

938 The American Journal of Medicine, Vol 124, No 10, October 2011 9. Gastrointestinal surgery for severe obesity. Consens Statement. 1991;9:1-20. 10. Busetto L, Sergi G, Enzi G, et al. Short-term effects of weight loss on the cardiovascular risk factors in morbidly obese patients. Obes Res. 2004;12:1256-1263. 11. Anderson JW, Brinkman VL, Hamilton CC. Weight loss and 2-y follow-up for 80 morbidly obese patients treated with intensive very-low-calorie diet and an education program. Am J Clin Nutr. 1992;56(1 Suppl):244S-246S. 12. Goodpaster BH, Delany JP, Otto AD, et al. Effects of diet and physical activity interventions on weight loss and cardiometabolic risk factors in severely obese adults: a randomized trial. JAMA. 2010;304:1795-1802. 13. Samaha FF, Iqbal N, Seshadri P, et al. A low-carbohydrate as compared with a low-fat diet in severe obesity. N Engl J Med. 2003;348: 2074-2081. 14. Ryan DH, Johnson WD, Myers VH, et al. Nonsurgical weight loss for extreme obesity in primary care settings: results of the Louisiana Obese Subjects Study. Arch Intern Med. 2010;170:146-154. 15. Asztalos BF, Swarbrick MM, Schaefer EJ, et al. Effects of weight loss, induced by gastric bypass surgery, on HDL remodeling in obese women. J Lipid Res. 2010;51:2405-2412. 16. Douketis JD, Sharma AM. Obesity and cardiovascular disease: pathogenic mechanisms and potential benefits of weight reduction. Semin Vasc Med. 2005;5:25-33.