ESPEN Congress Florence 2008 Severe obesity - Session organised in conjunction with ASPEN The SOS study Setting the Scene A. Thorell (Sweden)
The SOS study Setting the Scene Anders Thorell MD, PhD Associate Professor of Surgery Karolinska Institutet, CLINTEC, Division of Surgery & Ersta hospital, Stockholm, Sweden ESPEN, Florence 2008
SOS Effects of surgically induced weight loss on diabetes, stroke, myocardial infarction and overall mortality
SOS Effects of surgically induced weight loss on diabetes, stroke, myocardial infarction and overall mortality Scientific background Study design and aims Weight loss
SOS Effects of surgically induced weight loss on diabetes, stroke, myocardial infarction and overall mortality Scientific background Study design and aims Weight loss Effects of surgically induced, intentional weight loss on: Incidence of and recovery from diabetes Incidence of stroke and MI Overall mortality
SOS Effects of surgically induced weight loss on diabetes, stroke, myocardial infarction and overall mortality Scientific background Study design and aims Weight loss Effects of surgically induced, intentional weight loss on: Incidence of and recovery from diabetes Incidence of stroke and MI Overall mortality
Scientific background Mortality ratio Baseline BMI Lew EA et al. J Chronic Dis 1979; 32: 563-76.
The crucial question Mortality ratio Baseline BMI
Most observational epidemiological studies Mortality ratio Baseline BMI
Weight loss is associated with increased mortality Lean ME Diabet Med1990;7:228-33 Wannamethee G Eur J Clin Nutr 1990;44:133-142 Lissner L N Engl J Med 1991;324:1839-1844 Lee IM JAM A 1992;268:2045-2049 Pamuk ER Am J Epidemiol 1992;136:686-97 Walker M Int J Epidemiol1995;24:694-703 Suadicani P J Cardiovasc Risk 1997;4:25-32 Yaari S Am J Epidemiol 1998;148:546-55 Nilsson PM J Intern Med 2002;252:70-78 Gaesser GA Med Sci Sports Exerc 1999:31:1118 Mikkelsen KL Epidemiol 1999;10:671-678 Williamsson DF Am J Epidemiol 1999;149:491-503 Wannamethee SG Am J Epidemiol 2000;151:667-675
Risk factor changes, % Risk factor changes by weight change over two years in 842 obese SOS subjects 5 89 82 133 71 66 67 127 121 86 No. 40 89 82 133 71 66 67 127 121 No. 86 0 0-5 SBP -40 TG DBP INS -10-80 50 15 HDL 25 CHOL 0 0-15 URIC GLU -25-30 -100-40 -30-20 -15-10 -5 0 5 40-100 -40-30 -20-15 -10-5 0 5 Body weight changes, kg 40
Body weight loss and risk change Risk Mortality Risk factors Body weight
Body weight loss and risk change Mortality Risk Condition Body weight
Body weight loss and risk change Mortality Risk Condition? Body weight
The SOS intervention study Primary aim Effects of intentional weight loss on total mortality
The SOS intervention study Primary aim Effects of intentional weight loss on total mortality Secondary aims Effects of intentional weight loss on incidence of diabetes, MI, stroke and cancer
The SOS intervention study Power SOS was powered (80 % power, p<0.05) to detect a 23% reduction in total mortality in the surgery group as compared to a conventionally treated contemporaneously matched control group. This power required 2000 surgically treated subjects and 2000 controls followed for 10 years (40,000 person years).
SOS Swedish Obese Subjects Participating sites: 25 surgical departments 480 primary health care centers
SOS Swedish Obese Subjects Application forms n=11,953 Obese eligible applicants n=8,966 Registry study n=6,905 Intervention study
Postoperative mortality in the 1970ies and 1980ies: 1 5 % Brolin, RE; Gastroenterol Clin North Am 1987;16:317
Contemp. Matched SOS Swedish Obese Subjects Application forms n=11,953 Obese eligible applicants n=8,966 Registry study n=6,905 Intervention study Surgical group n=2,010 Non-surgical controls n=2,037
SOS recruitment Study Inclusion n Follow up period years SOS 1987-2001 4,047 20
SOS Swedish Obese Subjects Inclusion and exclusion criteria Inclusion criteria: Exclusion criteria: Age 37 60 years BMI men 34 women 38 minimal and aimed at ensuring that subjects in the surgery group could tolerate the operation Identical inclusion and exclusion criteria in both study groups.
SOS Swedish Obese Subjects Matching variables Sex Age Height Weight Waist Hip Systolic BP S-Cholesterol S-Triglycerides Menopausal status Diabetes Smoking Psychastenia Monotony avoidance Current health Social interaction Attachment Stressful life events SOS: L. Sjöström
SOS Swedish Obese Subjects Key variables at registry examination Control group n = 2037 Surgery group n = 2010 Age, yrs 47 46 Male gender, % 30 30 Smoker, % 20 28 Weight, kg 117 118 Height, cm 169 169 BMI, kg/m 2 40.9 41.3
SOS Swedish Obese Subjects Key variables at registry examination Control group n = 2037 Surgery group n = 2010 Syst BP, mm Hg 140 139 Glucose, mmol/l 5.3 5.3 Insulin, mu/l 20 21 TG, mmol/l 2.2 2.2 Total Chol, mmol/l 5.8 5.8 HDL Chol, mmol/l 1.2 1.2
SOS examinations Examination at sites with weight, blood pressure, anthropometry, patient and site questionnaires. Examination also includes ECG and a centralized laboratory evaluation. 0 5 10 15 20 yrs n: 4047 3058 1963 361 0
Participation rates at SOS examinations 2 yr 10 yr 15 yr Surgery 94% 84% 66% Control 83% 75% 87%
SOS Intervention study Surgical treatment Vertical Banded Gastroplasty Gastric Banding Gastric bypass Conventional treatment of obese controls According to standard at the primary health care center the patient belonged to
SOS Weight change Sjostrom L et al. NEJM 2007;357:741-752
Two and ten year diabetes incidence 30 24 *** 20 10 0 Odds-ratio: 95% CI: 8 *** 2-year 1 10-year 0.16 0.24 0.09-0.27 0.14-0.41 Odds ratio: 0.14 0.25 95% CI: 0.08-0.24 0.17-0.38 N control 1402 539 N surgery 1489 517 7 Control Surgery Adapted from L. Sjöström NEJM 2004;351:2683-2693
Two and ten year recovery from diabetes 100 80 *** 72 60 36 40 21 13 *** * 20 0 2-year 10-year Odds-ratio: 95% CI: 8.26 2.75 5.49-12.4 1.13-6.72 Odds ratio: 8.26 2.75 95% CI: 5.49-12.4 1.13-6.72 Control Surgery Adapted from L. Sjöström NEJM 2004;351:2683-2693
Cumulative incidence, % SOS Myocardial infarction 8 6 Individuals with MI, n Control 87 Surgery 64-28.5% 4 2 HR = 0.715 95% CI: 0.518 to 0.987 P= 0.0411 0 0 2 4 6 8 10 12 14 16 Years of follow up Lars Sjöström, Aug. 2006
SOS Myocardial infarction Percent of subjects by gender and diabetic state 12,00 10,00 % 8,00 8 6,00 6 4,00 4 2,00 2 0,00 0 All Males Females Diab Non-Diab Controls Surgery Lars Sjöström, Aug. 2006
SOS Mortality HR = 0.76 95% CI: 0.59 to 0.99 P= 0.04 Sjostrom L et al. NEJM 2007;357:741-752
SOS Mortality % 16 14 12 10 8 6 4 2 0 All Males Females Diab Non-Diab Controls Surgery Lars Sjöström, Aug. 2006
SOS Cause of death Cardiovascular Surgery Control n=2010 n=2037 ANY EVENT 43 53 Cardiac 35 44 Myocardial infarction 13 25 Heart failure 2 5 Sudden death 20 14 Stroke 6 6 Intracerebral hemorrhage 2 4 Infarction 1 2 Subarachnoid bleeding 3 0 Other 2 3 Aortic aneurysm 1 2 Aortic thrombosis 0 1 Diabetic gangrene 1 0 Noncardiovascular Surgery Control n=2010 n=2037 ANY EVENT 58 76 Tumor 29 48 Cancer 29 47 Meningioma 0 1 Infection 12 3 Thromboembolic disease 5 7 Pulmonary embolism 4 7 Vena caval thrombosis 1 0 Other 12 18 NO. OF DEATHS 101 129 Sjostrom L et al. NEJM 2007;357:741-752
SOS - Conclusions Surgically induced weight loss is associated with: Improved HRQL Improvements in cardiovascular risk factors Recovery from diabetes Reduced incidence in diabetes Reduced risk for myocardial infarction Reduction in overall mortality
Future of obesity surgery Metabolic surgery Diabetic surgery Tailored surgery/medication Studies of mechanisms of action!!
ERAS
Gastric bypass
Laparoscopic gastric bypass
High volume of surgical cases Standardized surgical technique (RYGBP) Access to biological material
Treatment goals PREVENT AND TREAT DISEASES IMPROVE QUALITY OF LIFE