- Email: [email protected]
Reduction in Predicted Coronary Heart Disease Risk After Substantial Weight Reduction After Bariatric Surgery
Reduction in Predicted Coronary Heart Disease Risk After Substantial Weight Reduction After Bariatric Surgery Jody A. Vogel, MDa, Barry A. Franklin, PhDa, Kerstyn C. Zalesin, MDa, Justin E. Trivax, MDa, Kevin R. Krause, MDb, David L. Chengelis, MDb, and Peter A. McCullough, MD, MPHa,* In recent years, bariatric surgery has become an increasingly used therapeutic option for morbid obesity. The effect of weight loss after bariatric surgery on the predicted risk of coronary heart disease (CHD) has not previously been studied. We evaluated baseline (preoperative) and follow-up (postoperative) body mass index, CHD risk factors, and Framingham risk scores (FRSs) for 109 consecutive patients with morbid obesity who lost weight after laparoscopic Roux-en-Y gastric bypass surgery. Charts were abstracted using a case-report form by a reviewer blinded to the FRS results. The study included 82 women (75%) and 27 men (25%) (mean age 46 ⴞ 10 years). Mean body mass index values at baseline and follow-up were 49 ⴞ 8 and 36 ⴞ 8 kg/m2, respectively (p <0.0001). During an average follow-up of 17 months, diabetes, hypertension, and dyslipidemia resolved or improved after weight loss. Thus, the risks of CHD as predicted by FRS decreased by 39% in men and 25% in women. The predicted 10-year CHD risks at baseline and follow-up were 6 ⴞ 5% and 4 ⴞ 3%, respectively (p <0.0001). For those without CHD, men compared favorably with the age-matched general population, with a final 10-year risk of 5 ⴞ 4% versus an expected risk of 11 ⴞ 6% (p <0.0001). Likewise, women achieved a level below the age-adjusted expected 10-year risk of the general population, with a final risk of 3 ⴞ 3% versus 6 ⴞ 4% (p <0.0001). In conclusion, weight loss results in a significant decrease in FRS 10-year predicted CHD risk. Bariatric surgery decreases CHD risk to rates lower than the age- and gender-adjusted estimates for the general population. These data suggest substantial and sustained weight loss after bariatric surgery may be a powerful intervention to decrease future rates of myocardial infarction and death in the morbidly obese. © 2007 Elsevier Inc. All rights reserved. (Am J Cardiol 2007;99:222–226)
Bariatric surgery has become an increasingly used therapeutic option for those with morbid obesity, which is a substantial health risk.1–13 Surgery for obesity results in sustained weight loss and improvement or often resolution of co-morbid conditions, including type 2 diabetes mellitus, hypertension, sleep apnea, lower extremity venous stasis, gastroesophageal reflux disease, degenerative joint disease, and nonalcoholic steatohepatitis.14,15 In 2003, ⬎100,000 bariatric operations were performed with ⬎80% of cases being gastric bypass surgery.16 The reported in-hospital mortality rate for this procedure is 0.1% to 0.2%.16 Rates of adverse events including reoperations and cardiopulmonary complications are 6% to 9% and 4% to 7%, respectively.16 The Framingham 10-year coronary heart disease (CHD) risk calculation is a global representation of CHD risk and, although not validated specifically in obese patients who have clustering of risk factors, it is a tool that can summarize changes in multiple risk factors within the same patient over time. Because some cardiovascular risk factors improve or perhaps resolve with substantial
a Department of Medicine, Divisions of Cardiology, Nutrition and Preventive Medicine, and bDepartment of Surgery, William Beaumont Hospital, Royal Oak, Michigan. Manuscript received June 7, 2006; revised manuscript received and accepted August 7, 2006. *Corresponding author: Tel: 248-655-5948; fax: 248-655-5901. E-mail address: [email protected] (P.A. McCullough).
0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2006.08.017
weight loss, we hypothesized that the predicted Framingham 10-year CHD risk might be favorably decreased after bariatric surgery. Methods The study sample included 109 consecutive, morbidly obese patients who were evaluated at the William Beaumont Hospital Weight Control Center (Royal Oak, Michigan) from November 2001 to December 2003, and underwent laparoscopic Roux-en-Y gastric bypass surgery. Because the study was a retrospective chart review, it was considered exempt from the hospital’s human investigations committee review. All patients underwent a comprehensive medical evaluation with a complete history and physical examination by a cardiologist. A 12-lead electrocardiogram at rest and blood chemistry testing were completed. All data recorded for study purposes were cross checked with the medical record to verify accuracy. Patients were weighed without shoes using a large-capacity scale (Scaletronix 6702, Scaletronix, Wheaton, Illinois), and a wall tape measure was used to measure height. Body mass index (BMI) was calculated as weight (kilograms) divided by height (meters) squared. Coronary artery disease (CAD) was defined as self-reported myocardial infarction or previous coronary revascularization. Hypertension was defined as a blood pressure ⱖ140/90 mm Hg and/or use of antihypertensive medications. Smoking status was www.AJConline.org
Preventive Cardiology/Weight Loss Reduces CHD Risk
based on patient self-report. Because fasting glucose level was not available in all subjects, diabetes mellitus was defined as use of antidiabetic medication and/or a glycosylated hemoglobin ⬎6%. In addition, because patients who undergo bariatric surgery in general cannot tolerate substantial quantities of oral glucose postoperatively because of the risk of dumping syndrome, the definition of diabetes remained the same, and oral glucose tolerance testing was not performed. CHD risk was estimated in all patients using the 1998 gender-specific Framingham risk scores (FRSs).18 The FRS was calculated using a patient’s age, calculated low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, systolic blood pressure, diabetes mellitus status, and smoking status. Low-density lipoprotein cholesterol concentration was unavailable in 6 patients, so the total cholesterol version of the FRS was used instead of low-density lipoprotein cholesterol for these patients.17 To further isolate the effect of weight loss on modifiable components of the FRS, preoperative age was also used for calculation of postoperative risk of CHD. The predicted FRS CHD risk was compared with the expected risk of the general population using published tables of CHD rates by age strata.17 These data were derived by Wilson et al17 from 2,489 men and 2,856 women who were 30 to 74 years old at the time of their Framingham Heart Study examination in 1971 to 1974, with CHD ascertained at the 12-year follow-up.18 Because of their low frequency and greater expected 10-year predicted CHD risk, we elected to include those with a baseline history of CAD. All patients underwent planned, laparoscopic Roux-en-Y gastric bypass surgery. After receiving general endotracheal anesthesia, the abdomen was prepped, draped, and accessed using laparoscopic techniques. Gastric bypass using the Roux-en-Y anastomosis restricting the size of the stomach by division of 90% of the lower stomach from the residual gastric pouch was performed. The jejunum was anastomosed to the gastric pouch. The abdomen and viscera were checked for bleeding and leaks. Thereafter, the abdomen was evacuated of air, and subcutaneous drains were placed only for excess serosanguineous drainage. Patients underwent planned extubation in the postoperative recovery room. Average length of stay was 3 ⫾ 2 days. All patients underwent a postoperative comprehensive follow-up evaluation similar to the baseline assessment with repeat physical examination and blood testing. When patients had multiple follow-up visits in the chart, the evaluation most remote from surgery was recorded for study purposes. Patients with a history of CAD were not excluded from the study. Separate analyses of the FRS-predicted CHD risk for patients with and without CAD were planned a priori. Charts were abstracted using a case-report form by a trained reviewer blinded to the FRS results. A computerized laboratory information system was accessed to review assessments performed at the William Beaumont Hospital Reference Laboratory. Blood test results from outside laboratories were used if they were the most recently available in the chart. Baseline characteristics were expressed as mean ⫾ SD or counts with proportions, as appropriate. Univariate compar-
223
Table 1 Baseline characteristics of the study sample Characteristic Age (yrs) Caucasian African-American Other race BMI Diabetes mellitus Hypertension CAD* Smoking Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Total cholesterol (mg/dl) Low-density lipoprotein cholesterol (mg/dl) High-density lipoprotein cholesterol (mg/dl) Triglycerides (mg/dl)
Women (n ⫽ 82)
Men (n ⫽ 27)
p Value
46 ⫾ 10 74 (90) 5 (6) 3 (4) 49 ⫾ 8 25 (31) 44 (54) 4 (5) 9 (11) 131 ⫾ 15
47 ⫾ 11 22 (82) 4 (15) 1 (4) 51 ⫾ 7 15 (56) 20 (74) 7 (26) 4 (15) 138 ⫾ 15
0.59 0.30 0.22 1.00 0.09 0.02 0.06 0.005 0.73 0.04
79 ⫾ 10
82 ⫾ 11
0.24
205 ⫾ 37 115 ⫾ 32
184 ⫾ 30 104 ⫾ 29
0.009 0.14
52 ⫾ 13
42 ⫾ 7
⬍0.0001
199 ⫾ 106
217 ⫾ 98
0.43
Counts and proportions are listed in parentheses. Means ⫾ SDs are given for continuous variables. * CAD was defined as self-reported myocardial infarction or previous coronary revascularization.
isons were made with paired Student’s t test, Fisher’s exact test, chi-square test, and chi-square test for trend, as appropriate. Pearson’s correlation was used to evaluate relations between the actual calculated 10-year CHD risk and that expected in the general population based on age and gender. All hypothesis testing was 2-tailed. A p value ⬍0.05 was considered statistically significant. An analytic file was created and analysis was performed with conventional statistical software (SPSS 14.0, SPSS, Inc., Chicago, Illinois). Results The study population was a mean age of 47 ⫾ 10 years with 82 women (75%) and 27 men (25%; Table 1). Of 109 patients, 96 were Caucasian (88%), 9 were African-American (8%), and 4 were of another race (4%). Baseline diabetes mellitus was noted in 25 of 82 men (31%) and 15 of 27 women (56%, p ⫽ 0.02). Overall, baseline hypertension was present in 20 of 27 men (74%) and 44 of 82 women (54%, p ⫽ 0.06). Mean preoperative BMI for all patients was 49 ⫾ 8 kg/m2 (range 37 to 90). Average baseline BMI values for men and women were 51 ⫾ 7 and 49 ⫾ 8 kg/m2, respectively (p ⫽ 0 .09). Table 1 lists additional baseline characteristics of the study sample. Mean duration between baseline studies and surgery was 4 ⫾ 2 months, and time from surgery to follow-up was 13 ⫾ 10 months. Accordingly, the overall mean duration of follow-up was 17 ⫾ 10 months. Mean weight loss and decrease in BMI were 36 ⫾ 16 kg (79 ⫾ 36 lb) and 12 ⫾ 6 kg/m2, respectively (Table 2). Mean follow-up BMI values for men and women were 40 ⫾ 6.3 and 35 ⫾ 9 kg/m2, respectively (p ⫽ 0.04).
224
The American Journal of Cardiology (www.AJConline.org)
Table 2 Comprehensive clinical and laboratory value changes in men and women Characteristic
Weight (kg) [lbs] BMI (kg/m2) High-density lipoprotein cholesterol (mg/dl) Low-density lipoprotein cholesterol (mg/dl) Total cholesterol (mg/dl) Triglycerides (mg/dl) Glucose (mg/dl) Glycohemoglobin (%) C-peptide (ng/ml) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) FRS 10-yr FRS CHD risk estimate (%)
No. of Patients
Women
Men
All
Baseline
Follow-up
p Value
Baseline
Follow-up
p Value
Baseline
Follow-up
p Value
93 ⫾ 23 [205 ⫾ 49] 35 ⫾ 9 56 ⫾ 17
⬍0.0001
156 ⫾ 20 [344 ⫾ 43] 51 ⫾ 7 42 ⫾ 8
120 ⫾ 21 [264 ⫾ 46] 39 ⫾ 6 46 ⫾ 14
⬍0.0001 ⬍0.0001 0.118
135 ⫾ 25 [298 ⫾ 55] 49 ⫾ 8 50 ⫾ 12
100 ⫾ 25 [219 ⫾ 55] 36 ⫾ 8 54 ⫾ 17
⬍0.0001
109 108
129 ⫾ 23 [283 ⫾ 51] 49 ⫾ 8 52 ⫾ 13
⬍0.0001 0.012
103
116 ⫾ 31
103 ⫾ 37
0.0006
102 ⫾ 30
89 ⫾ 28
0.043
113 ⫾ 31
100 ⫾ 35
0.001
108 108 107 107 107 109
207 ⫾ 37 200 ⫾ 108 105 ⫾ 41 6⫾1 4⫾2 131 ⫾ 15
181 ⫾ 33 132 ⫾ 60 92 ⫾ 34 5⫾1 3⫾1 113 ⫾ 14
⬍0.0001 ⬍0.0001 0.003 ⬍0.0001 ⬍0.0001 ⬍0.0001
184 ⫾ 31 224 ⫾ 100 119 ⫾ 50 7⫾1 5⫾3 141 ⫾ 16
159 ⫾ 32 133 ⫾ 69 99 ⫾ 32 6⫾1 4⫾1 125 ⫾ 21
⬍0.0001 ⬍0.0001 0.026 0.002 0.006 0.002
201 ⫾ 37 205 ⫾ 105 108 ⫾ 44 6⫾1 4⫾2 133 ⫾ 16
176 ⫾ 34 132 ⫾ 61 94 ⫾ 34 6⫾1 3⫾1 116 ⫾ 17
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
109
79 ⫾ 9
70 ⫾ 9
⬍0.0001
82 ⫾ 12
75 ⫾ 13
0.021
80 ⫾ 10
71 ⫾ 10
⬍0.0001
107 107
3⫾6 5⫾4
⫺1 ⫾ 8 3⫾3
⬍0.0001 ⬍0.0001
5⫾3 12 ⫾ 7
2⫾3 6⫾4
⬍0.0001 0.002
3⫾6 6⫾5
109
⬍0.0001 0.04
⫺1 ⫾ 7 4⫾3
⬍0.0001 ⬍0.0001
Figure 1. Change in mean lipid values for men and women. Chol ⫽ total cholesterol; HDL-C ⫽ high-density lipoprotein cholesterol; LDL-C ⫽ low-density lipoprotein cholesterol; TG ⫽ triglycerides.
Changes in lipoprotein values after weight loss are presented in Table 2 and Figure 1. Men and women demonstrated significant absolute and percent decreases in total cholesterol (⫺26 ⫾ 35 mg/dl, ⫺26 ⫾ 38%), low-density lipoprotein cholesterol (⫺14 ⫾ 39 mg/dl, ⫺7 ⫾ 38%), and triglycerides (⫺73 ⫾ 96 mg/dl, ⫺26 ⫾ 37%) and a significant increase in high-density lipoprotein cholesterol (⫹4 ⫾ 16 mg/dl,⫹10 ⫾ 30%). Changes in glycosylated hemoglobin, glucose, and Cpeptide levels are listed in Table 2. All 3 variables demonstrated statistically significant decreases after weight loss, as did the prevalence of diabetes mellitus. In men, 15 of 27 (56%) had diabetes mellitus before surgery, which after
Figure 2. Changes in diabetes mellitus (DM) and smoking status in men and women at baseline (black bars) and follow-up (white bars).
weight loss decreased to 6 of 27 (22%, p ⫽ 0.02). Likewise in women, diabetes mellitus decreased from 25 (31%) to 7 (9%) of 82 (p ⫽ 0.003; Figure 2). Figure 3 shows a significant decrease in systolic and diastolic blood pressures at rest of patients after weight loss. In men, rates of hypertension decreased from 16 of 27 (60%) to 4 of 21 (19%, p ⫽ 0.004). Likewise in women, hypertension decreased from 28 of 82 (34%) to 5 of 71 (7%, p ⱕ0.0001). Blood pressure data were not available in 6 men and 9 women.
Preventive Cardiology/Weight Loss Reduces CHD Risk
Figure 3. Changes in mean systolic (black bars) and diastolic (white bars) blood pressures in men and women (p values indicate differences from baseline values).
225
Figure 5. Average 10-year CHD risk change in those with and without CAD at baseline (black bars) and follow-up (white bars).
baseline and final CHD risks for men and women without baseline CAD were 12 ⫾ 7% and 6 ⫾ 4% (p ⫽ 0.001) and 5 ⫾ 3% and 3 ⫾ 3% (p ⬍0.0001), respectively. Relative decreases in predicted 10-year CHD risk in patients without baseline CAD were 48 ⫾ 25% for men, 24 ⫾ 46% for women, and 28 ⫾ 44% for the entire cohort. For those without CAD, men compared favorably with the general population (Framingham) of age-adjusted men with a final 10-year risk of 6 ⫾ 4% (vs an expected 11 ⫾ 6%).17 Likewise, women achieved levels below the ageadjusted 10-year risk of the general population, with a final risk of 3 ⫾ 3% versus 6 ⫾ 4% (p ⱕ0.0001).17 Overall, the final 10-year risk predictions in our patients were similar to the published, age-adjusted “low”-risk rates of 5.0 ⫾ 2.3% for men and 3.6 ⫾ 2.4% for women in the general population.18 Correlations (r values) between the calculated FRS and the expected, age-adjusted low-risk rates in the general population were 0.5 for men (p ⫽ 0.02), r ⫽ 0.7 for women (p ⬍0.0001), and 0.7 for the entire cohort (p ⬍0.0001). Discussion Figure 4. Ten-year predicted CHD risk before (black bars) and after (white bars) bariatric surgery for men and women.
Mean baseline and final 10-year FRS CHD risks for men and women are shown in Figure 4. A statistically significant decrease in predicted CHD risk was achieved after weight loss. Mean predicted 10-year CHD risks decreased by 39 ⫾ 53% in men, 25 ⫾ 46% in women, and 27 ⫾ 47% in all patients. On average, men moved from the moderate (10% to 20%) to the low (⬍10%) 10-year CHD risk category after weight loss. Patients with CAD at baseline had a nonsignificant decrease in predicted risk (Figure 5). In those without CAD, average baseline and follow-up 10-year CHD risks were 6 ⫾ 5% and 3 ⫾ 3%, respectively (p ⬍0.0001). Mean
Our study is the first to quantify a global decrease in CHD risk after substantial and sustained weight loss secondary to bariatric surgery. We found that weight loss was effective in decreasing the FRS 10-year predicted CHD risk in men and women; however, the decrease in CHD risk was greater in men. Moreover, men and women without CAD achieved a lower CHD risk compared with the gender-specific population of a comparable age.17 Accordingly, this is the first study to suggest that a surgical intervention can elicit CHD risk levels that are below those expected for the population. The lower CHD risk, as projected by the FRS, was driven by significant improvement in multiple risk factors after weight loss induced by bariatric surgery. This form of surgery causes gastric restriction of caloric intake and a
226
The American Journal of Cardiology (www.AJConline.org)
mild degree of malabsorption. During the rapid weight-loss phase, patients consume on average ⬃700 to 800 kcal/day and are gradually able to increase to a stable baseline intake of ⬃1,500 kcal/day. Previous studies have suggested that caloric restriction alone can have a significant effect on glycemic control, lipids, blood pressure, and inflammatory factors.18 –20 It is currently apparent that dyslipidemia, hypertension, and diabetes mellitus substantially improve or resolve after bariatric surgery.16,21 We have demonstrated that when all of these risk factors are considered in an integrative way using the FRS, the projected CHD risk is markedly lower after bariatric surgery. This is important, because recent data have suggested that obesity in addition to exacerbating traditional CHD risk factors seem to increase the acute rate of cardiovascular events above what is predicted by the FRS.22 This consideration is clinically relevant because our patients, despite substantial weight loss, had final BMI values in the class II obesity range. Thus, future studies evaluating traditional and novel atherosclerotic risk factors, weight loss, and cardiovascular outcomes are warranted in the obese. Our findings are consistent with a recent, preliminary report demonstrating a decrease in CHD risk after bariatric surgery. Batsis et al23 categorized the decrease in CHD risk using risk models and reported improvements in CHD risk factors from the National Health and Nutrition Examination Survey (NHANES) I and the NHANES I Epidemiologic Follow-up Study, which investigated CHD mortality and hospitalization outcomes.24 The predicted 10-year risk was decreased from 37% to 18% from these estimating equations. The investigators projected that 4 deaths would be prevented for every 100 patients who underwent bariatric surgery. In contrast, our study used the FRS, which estimated total CHD risk (risk of developing angina pectoris, myocardial infarction, or coronary disease death) over the course of 10 years and found concordant results.17 We recognize several limitations in our study. There was variation in the postsurgical time points at which follow-up measurements were obtained and not all patients had completed the expected weight-loss phase after bariatric surgery. By using the final measurement available for each patient, we were able to estimate the longer term effects of weight loss on 10-year CHD risk. Many patients had limited short-term follow-up, and the full benefit of weight loss was probably not reflected in their calculated 10-year CHD risk estimate. Long-term follow-up is indicated to ascertain the effects on 10-year CHD risk estimates because reversal or stabilization of the decrease in cardiovascular risk may occur, particularly if individuals regain weight. Our study population was largely homogenous, i.e., primarily Caucasian and women. It would be beneficial to evaluate the effects of weight loss on CHD risk in a more diverse population over a longer period. We had only a small group with established CAD, so larger studies are needed to evaluate the effect of weight loss on future CHD risk and associated rates of myocardial infarction and death. We acknowledge that some changes in CHD risk may not be a direct result of bariatric surgery (e.g., decrease in smoking) and may reflect the effect of patient education, self-improvement, and overall better medical care than that of the general population.
1. Flegal KM, Carroll MD, Ogden C, Johnson CL. Prevalence and trends in obesity among US adults, 1999 –2000. JAMA 2002;288:1723–1727. 2. McCullough PA, Sandberg KR, Miller WM, Odom JS, Sloan KC, de Jong AT, Nori KE, Irving SD, Krause KR, Franklin BA. Substantial weight gain during adulthood: the road to bariatric surgery. Prev Cardiol 2005;8:155–159. 3. Fontaine KR, Redden DT, Wang C, Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA 2003;289:187–193. 4. Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in the United States, 2000. JAMA 2004;291:1238 –1245. 5. Flegal KM, Graubard BI, Williamson DF, Gail MH. Excess deaths associated with underweight, overweight, and obesity. JAMA 2005; 293:1861–1867. 6. Eckel RH, Krauss RM. American Heart Association call to action: obesity as a major risk factor for coronary heart disease. Circulation 1998;97:2099 –2100. 7. Chobanian AV, Bakris GL, Black HR, for the National Heart, Lung, and Blood Institute Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: National High Blood Pressure Education Program Coordinating Committee. The seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003;289:2560 –2572. 8. Wilson PW, D’Agostino RB, Sullivan L, Parise H, Kannel WB. Overweight and obesity as determinants of cardiovascular risk. The Framingham experience. Arch Intern Med 2002;162:1867–1872. 9. Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales VS, Marks JS. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA 2003;289:76 –79. 10. Calle EE, Thun MJ, Petrelli JM, Rodriguez C, Heath CW. Body-mass index and mortality in a prospective cohort of US adults. N Engl J Med 1999;341:1097–1105. 11. Krauss RM, Winston M. Obesity. Impact on cardiovascular disease. Circulation 1998;98:1472–1476. 12. National Task Force on Prevention and Treatment of Obesity. Overweight, obesity and health risk. Arch Intern Med 2000;160:898 –904. 13. Mitka M. Surgery for obesity: demand soars amid scientific, ethical questions. JAMA 2003;289:1761–1762. 14. Buchwald H, Avidor Y, Braunwald E, Jensen MD, Pories W, Fahrbach K, Schoelles K. Bariatric surgery: a systematic review and metaanalysis. JAMA 2004;292:1724 –1737. 15. Lara MD, Kothari SN, Sugerman HJ. Surgical management of obesity: a review of the evidence relating to the health benefits and risks. Treat Endocrinol 2005;4:55– 64. 16. Santry HP, Gillen DL, Lauderdale DS. Trends in bariatric surgical procedures. JAMA 2005;294:1909 –1917. 17. Wilson PWF, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation 1998;97:1837–1847. 18. Harder H, Dinesen B, Astrup A. The effect of a rapid weight loss on lipid profile and glycemic control in obese type 2 diabetic patients. Int J Obes Relat Metab Disord 2004;28:180 –182. 19. Lofgren I, Zern T, Herron K, West K, Sharman MJ, Volek JS, Shachter NS, Koo SI, Fernandez ML. Weight loss associated with reduced intake of carbohydrate reduces the atherogenicity of LDL in premenopausal women. Metabolism 2005;54:1133–1141. 20. Pereira MA, Swain J, Goldfine AB, Rifai N, Ludwig DS. Effects of a low-glycemic load diet on resting energy expenditure and heart disease risk factors during weight loss. JAMA 2004;292:2482–2490. 21. Sjostrom L, Lindroos AK, Peltonen M, Torgerson J, Bouchard C, Carlsson B, Dahlgren S, Larsson B, Narbro K, Sjostrom CD, Sullivan M, Wedel H. Lifestyle, diabetes, and cardiovascular risk factors 10 years after bariatric surgery. N Engl J Med 2004;351:2683–2693. 22. Mora S, Yanek LR, Moy TF, Fallin MD, Becker LC, Becker DM. Interaction of body mass index and Framingham risk score in predicting incident coronary disease in families. Circulation 2005;111:1871– 1876. 23. Batsis JA, Romero-Corral A, Collazo-Clavell ML, Sarr MG, Somers V, Brekke L, Lopez-Jimenez F. Effect of bariatric surgery on cardiovascular risk factors and predicted effect on cardiovascular events and mortality in class II–III obesity (abstract). J Am Coll Cardiol 2006; 42:358A–359A.
Bariatric surgery has become an increasingly used therapeutic option for those with morbid obesity, which is a substantial health risk.1–13 Surgery for obesity results in sustained weight loss and improvement or often resolution of co-morbid conditions, including type 2 diabetes mellitus, hypertension, sleep apnea, lower extremity venous stasis, gastroesophageal reflux disease, degenerative joint disease, and nonalcoholic steatohepatitis.14,15 In 2003, ⬎100,000 bariatric operations were performed with ⬎80% of cases being gastric bypass surgery.16 The reported in-hospital mortality rate for this procedure is 0.1% to 0.2%.16 Rates of adverse events including reoperations and cardiopulmonary complications are 6% to 9% and 4% to 7%, respectively.16 The Framingham 10-year coronary heart disease (CHD) risk calculation is a global representation of CHD risk and, although not validated specifically in obese patients who have clustering of risk factors, it is a tool that can summarize changes in multiple risk factors within the same patient over time. Because some cardiovascular risk factors improve or perhaps resolve with substantial
a Department of Medicine, Divisions of Cardiology, Nutrition and Preventive Medicine, and bDepartment of Surgery, William Beaumont Hospital, Royal Oak, Michigan. Manuscript received June 7, 2006; revised manuscript received and accepted August 7, 2006. *Corresponding author: Tel: 248-655-5948; fax: 248-655-5901. E-mail address: [email protected] (P.A. McCullough).
0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2006.08.017
weight loss, we hypothesized that the predicted Framingham 10-year CHD risk might be favorably decreased after bariatric surgery. Methods The study sample included 109 consecutive, morbidly obese patients who were evaluated at the William Beaumont Hospital Weight Control Center (Royal Oak, Michigan) from November 2001 to December 2003, and underwent laparoscopic Roux-en-Y gastric bypass surgery. Because the study was a retrospective chart review, it was considered exempt from the hospital’s human investigations committee review. All patients underwent a comprehensive medical evaluation with a complete history and physical examination by a cardiologist. A 12-lead electrocardiogram at rest and blood chemistry testing were completed. All data recorded for study purposes were cross checked with the medical record to verify accuracy. Patients were weighed without shoes using a large-capacity scale (Scaletronix 6702, Scaletronix, Wheaton, Illinois), and a wall tape measure was used to measure height. Body mass index (BMI) was calculated as weight (kilograms) divided by height (meters) squared. Coronary artery disease (CAD) was defined as self-reported myocardial infarction or previous coronary revascularization. Hypertension was defined as a blood pressure ⱖ140/90 mm Hg and/or use of antihypertensive medications. Smoking status was www.AJConline.org
Preventive Cardiology/Weight Loss Reduces CHD Risk
based on patient self-report. Because fasting glucose level was not available in all subjects, diabetes mellitus was defined as use of antidiabetic medication and/or a glycosylated hemoglobin ⬎6%. In addition, because patients who undergo bariatric surgery in general cannot tolerate substantial quantities of oral glucose postoperatively because of the risk of dumping syndrome, the definition of diabetes remained the same, and oral glucose tolerance testing was not performed. CHD risk was estimated in all patients using the 1998 gender-specific Framingham risk scores (FRSs).18 The FRS was calculated using a patient’s age, calculated low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, systolic blood pressure, diabetes mellitus status, and smoking status. Low-density lipoprotein cholesterol concentration was unavailable in 6 patients, so the total cholesterol version of the FRS was used instead of low-density lipoprotein cholesterol for these patients.17 To further isolate the effect of weight loss on modifiable components of the FRS, preoperative age was also used for calculation of postoperative risk of CHD. The predicted FRS CHD risk was compared with the expected risk of the general population using published tables of CHD rates by age strata.17 These data were derived by Wilson et al17 from 2,489 men and 2,856 women who were 30 to 74 years old at the time of their Framingham Heart Study examination in 1971 to 1974, with CHD ascertained at the 12-year follow-up.18 Because of their low frequency and greater expected 10-year predicted CHD risk, we elected to include those with a baseline history of CAD. All patients underwent planned, laparoscopic Roux-en-Y gastric bypass surgery. After receiving general endotracheal anesthesia, the abdomen was prepped, draped, and accessed using laparoscopic techniques. Gastric bypass using the Roux-en-Y anastomosis restricting the size of the stomach by division of 90% of the lower stomach from the residual gastric pouch was performed. The jejunum was anastomosed to the gastric pouch. The abdomen and viscera were checked for bleeding and leaks. Thereafter, the abdomen was evacuated of air, and subcutaneous drains were placed only for excess serosanguineous drainage. Patients underwent planned extubation in the postoperative recovery room. Average length of stay was 3 ⫾ 2 days. All patients underwent a postoperative comprehensive follow-up evaluation similar to the baseline assessment with repeat physical examination and blood testing. When patients had multiple follow-up visits in the chart, the evaluation most remote from surgery was recorded for study purposes. Patients with a history of CAD were not excluded from the study. Separate analyses of the FRS-predicted CHD risk for patients with and without CAD were planned a priori. Charts were abstracted using a case-report form by a trained reviewer blinded to the FRS results. A computerized laboratory information system was accessed to review assessments performed at the William Beaumont Hospital Reference Laboratory. Blood test results from outside laboratories were used if they were the most recently available in the chart. Baseline characteristics were expressed as mean ⫾ SD or counts with proportions, as appropriate. Univariate compar-
223
Table 1 Baseline characteristics of the study sample Characteristic Age (yrs) Caucasian African-American Other race BMI Diabetes mellitus Hypertension CAD* Smoking Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Total cholesterol (mg/dl) Low-density lipoprotein cholesterol (mg/dl) High-density lipoprotein cholesterol (mg/dl) Triglycerides (mg/dl)
Women (n ⫽ 82)
Men (n ⫽ 27)
p Value
46 ⫾ 10 74 (90) 5 (6) 3 (4) 49 ⫾ 8 25 (31) 44 (54) 4 (5) 9 (11) 131 ⫾ 15
47 ⫾ 11 22 (82) 4 (15) 1 (4) 51 ⫾ 7 15 (56) 20 (74) 7 (26) 4 (15) 138 ⫾ 15
0.59 0.30 0.22 1.00 0.09 0.02 0.06 0.005 0.73 0.04
79 ⫾ 10
82 ⫾ 11
0.24
205 ⫾ 37 115 ⫾ 32
184 ⫾ 30 104 ⫾ 29
0.009 0.14
52 ⫾ 13
42 ⫾ 7
⬍0.0001
199 ⫾ 106
217 ⫾ 98
0.43
Counts and proportions are listed in parentheses. Means ⫾ SDs are given for continuous variables. * CAD was defined as self-reported myocardial infarction or previous coronary revascularization.
isons were made with paired Student’s t test, Fisher’s exact test, chi-square test, and chi-square test for trend, as appropriate. Pearson’s correlation was used to evaluate relations between the actual calculated 10-year CHD risk and that expected in the general population based on age and gender. All hypothesis testing was 2-tailed. A p value ⬍0.05 was considered statistically significant. An analytic file was created and analysis was performed with conventional statistical software (SPSS 14.0, SPSS, Inc., Chicago, Illinois). Results The study population was a mean age of 47 ⫾ 10 years with 82 women (75%) and 27 men (25%; Table 1). Of 109 patients, 96 were Caucasian (88%), 9 were African-American (8%), and 4 were of another race (4%). Baseline diabetes mellitus was noted in 25 of 82 men (31%) and 15 of 27 women (56%, p ⫽ 0.02). Overall, baseline hypertension was present in 20 of 27 men (74%) and 44 of 82 women (54%, p ⫽ 0.06). Mean preoperative BMI for all patients was 49 ⫾ 8 kg/m2 (range 37 to 90). Average baseline BMI values for men and women were 51 ⫾ 7 and 49 ⫾ 8 kg/m2, respectively (p ⫽ 0 .09). Table 1 lists additional baseline characteristics of the study sample. Mean duration between baseline studies and surgery was 4 ⫾ 2 months, and time from surgery to follow-up was 13 ⫾ 10 months. Accordingly, the overall mean duration of follow-up was 17 ⫾ 10 months. Mean weight loss and decrease in BMI were 36 ⫾ 16 kg (79 ⫾ 36 lb) and 12 ⫾ 6 kg/m2, respectively (Table 2). Mean follow-up BMI values for men and women were 40 ⫾ 6.3 and 35 ⫾ 9 kg/m2, respectively (p ⫽ 0.04).
224
The American Journal of Cardiology (www.AJConline.org)
Table 2 Comprehensive clinical and laboratory value changes in men and women Characteristic
Weight (kg) [lbs] BMI (kg/m2) High-density lipoprotein cholesterol (mg/dl) Low-density lipoprotein cholesterol (mg/dl) Total cholesterol (mg/dl) Triglycerides (mg/dl) Glucose (mg/dl) Glycohemoglobin (%) C-peptide (ng/ml) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) FRS 10-yr FRS CHD risk estimate (%)
No. of Patients
Women
Men
All
Baseline
Follow-up
p Value
Baseline
Follow-up
p Value
Baseline
Follow-up
p Value
93 ⫾ 23 [205 ⫾ 49] 35 ⫾ 9 56 ⫾ 17
⬍0.0001
156 ⫾ 20 [344 ⫾ 43] 51 ⫾ 7 42 ⫾ 8
120 ⫾ 21 [264 ⫾ 46] 39 ⫾ 6 46 ⫾ 14
⬍0.0001 ⬍0.0001 0.118
135 ⫾ 25 [298 ⫾ 55] 49 ⫾ 8 50 ⫾ 12
100 ⫾ 25 [219 ⫾ 55] 36 ⫾ 8 54 ⫾ 17
⬍0.0001
109 108
129 ⫾ 23 [283 ⫾ 51] 49 ⫾ 8 52 ⫾ 13
⬍0.0001 0.012
103
116 ⫾ 31
103 ⫾ 37
0.0006
102 ⫾ 30
89 ⫾ 28
0.043
113 ⫾ 31
100 ⫾ 35
0.001
108 108 107 107 107 109
207 ⫾ 37 200 ⫾ 108 105 ⫾ 41 6⫾1 4⫾2 131 ⫾ 15
181 ⫾ 33 132 ⫾ 60 92 ⫾ 34 5⫾1 3⫾1 113 ⫾ 14
⬍0.0001 ⬍0.0001 0.003 ⬍0.0001 ⬍0.0001 ⬍0.0001
184 ⫾ 31 224 ⫾ 100 119 ⫾ 50 7⫾1 5⫾3 141 ⫾ 16
159 ⫾ 32 133 ⫾ 69 99 ⫾ 32 6⫾1 4⫾1 125 ⫾ 21
⬍0.0001 ⬍0.0001 0.026 0.002 0.006 0.002
201 ⫾ 37 205 ⫾ 105 108 ⫾ 44 6⫾1 4⫾2 133 ⫾ 16
176 ⫾ 34 132 ⫾ 61 94 ⫾ 34 6⫾1 3⫾1 116 ⫾ 17
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
109
79 ⫾ 9
70 ⫾ 9
⬍0.0001
82 ⫾ 12
75 ⫾ 13
0.021
80 ⫾ 10
71 ⫾ 10
⬍0.0001
107 107
3⫾6 5⫾4
⫺1 ⫾ 8 3⫾3
⬍0.0001 ⬍0.0001
5⫾3 12 ⫾ 7
2⫾3 6⫾4
⬍0.0001 0.002
3⫾6 6⫾5
109
⬍0.0001 0.04
⫺1 ⫾ 7 4⫾3
⬍0.0001 ⬍0.0001
Figure 1. Change in mean lipid values for men and women. Chol ⫽ total cholesterol; HDL-C ⫽ high-density lipoprotein cholesterol; LDL-C ⫽ low-density lipoprotein cholesterol; TG ⫽ triglycerides.
Changes in lipoprotein values after weight loss are presented in Table 2 and Figure 1. Men and women demonstrated significant absolute and percent decreases in total cholesterol (⫺26 ⫾ 35 mg/dl, ⫺26 ⫾ 38%), low-density lipoprotein cholesterol (⫺14 ⫾ 39 mg/dl, ⫺7 ⫾ 38%), and triglycerides (⫺73 ⫾ 96 mg/dl, ⫺26 ⫾ 37%) and a significant increase in high-density lipoprotein cholesterol (⫹4 ⫾ 16 mg/dl,⫹10 ⫾ 30%). Changes in glycosylated hemoglobin, glucose, and Cpeptide levels are listed in Table 2. All 3 variables demonstrated statistically significant decreases after weight loss, as did the prevalence of diabetes mellitus. In men, 15 of 27 (56%) had diabetes mellitus before surgery, which after
Figure 2. Changes in diabetes mellitus (DM) and smoking status in men and women at baseline (black bars) and follow-up (white bars).
weight loss decreased to 6 of 27 (22%, p ⫽ 0.02). Likewise in women, diabetes mellitus decreased from 25 (31%) to 7 (9%) of 82 (p ⫽ 0.003; Figure 2). Figure 3 shows a significant decrease in systolic and diastolic blood pressures at rest of patients after weight loss. In men, rates of hypertension decreased from 16 of 27 (60%) to 4 of 21 (19%, p ⫽ 0.004). Likewise in women, hypertension decreased from 28 of 82 (34%) to 5 of 71 (7%, p ⱕ0.0001). Blood pressure data were not available in 6 men and 9 women.
Preventive Cardiology/Weight Loss Reduces CHD Risk
Figure 3. Changes in mean systolic (black bars) and diastolic (white bars) blood pressures in men and women (p values indicate differences from baseline values).
225
Figure 5. Average 10-year CHD risk change in those with and without CAD at baseline (black bars) and follow-up (white bars).
baseline and final CHD risks for men and women without baseline CAD were 12 ⫾ 7% and 6 ⫾ 4% (p ⫽ 0.001) and 5 ⫾ 3% and 3 ⫾ 3% (p ⬍0.0001), respectively. Relative decreases in predicted 10-year CHD risk in patients without baseline CAD were 48 ⫾ 25% for men, 24 ⫾ 46% for women, and 28 ⫾ 44% for the entire cohort. For those without CAD, men compared favorably with the general population (Framingham) of age-adjusted men with a final 10-year risk of 6 ⫾ 4% (vs an expected 11 ⫾ 6%).17 Likewise, women achieved levels below the ageadjusted 10-year risk of the general population, with a final risk of 3 ⫾ 3% versus 6 ⫾ 4% (p ⱕ0.0001).17 Overall, the final 10-year risk predictions in our patients were similar to the published, age-adjusted “low”-risk rates of 5.0 ⫾ 2.3% for men and 3.6 ⫾ 2.4% for women in the general population.18 Correlations (r values) between the calculated FRS and the expected, age-adjusted low-risk rates in the general population were 0.5 for men (p ⫽ 0.02), r ⫽ 0.7 for women (p ⬍0.0001), and 0.7 for the entire cohort (p ⬍0.0001). Discussion Figure 4. Ten-year predicted CHD risk before (black bars) and after (white bars) bariatric surgery for men and women.
Mean baseline and final 10-year FRS CHD risks for men and women are shown in Figure 4. A statistically significant decrease in predicted CHD risk was achieved after weight loss. Mean predicted 10-year CHD risks decreased by 39 ⫾ 53% in men, 25 ⫾ 46% in women, and 27 ⫾ 47% in all patients. On average, men moved from the moderate (10% to 20%) to the low (⬍10%) 10-year CHD risk category after weight loss. Patients with CAD at baseline had a nonsignificant decrease in predicted risk (Figure 5). In those without CAD, average baseline and follow-up 10-year CHD risks were 6 ⫾ 5% and 3 ⫾ 3%, respectively (p ⬍0.0001). Mean
Our study is the first to quantify a global decrease in CHD risk after substantial and sustained weight loss secondary to bariatric surgery. We found that weight loss was effective in decreasing the FRS 10-year predicted CHD risk in men and women; however, the decrease in CHD risk was greater in men. Moreover, men and women without CAD achieved a lower CHD risk compared with the gender-specific population of a comparable age.17 Accordingly, this is the first study to suggest that a surgical intervention can elicit CHD risk levels that are below those expected for the population. The lower CHD risk, as projected by the FRS, was driven by significant improvement in multiple risk factors after weight loss induced by bariatric surgery. This form of surgery causes gastric restriction of caloric intake and a
226
The American Journal of Cardiology (www.AJConline.org)
mild degree of malabsorption. During the rapid weight-loss phase, patients consume on average ⬃700 to 800 kcal/day and are gradually able to increase to a stable baseline intake of ⬃1,500 kcal/day. Previous studies have suggested that caloric restriction alone can have a significant effect on glycemic control, lipids, blood pressure, and inflammatory factors.18 –20 It is currently apparent that dyslipidemia, hypertension, and diabetes mellitus substantially improve or resolve after bariatric surgery.16,21 We have demonstrated that when all of these risk factors are considered in an integrative way using the FRS, the projected CHD risk is markedly lower after bariatric surgery. This is important, because recent data have suggested that obesity in addition to exacerbating traditional CHD risk factors seem to increase the acute rate of cardiovascular events above what is predicted by the FRS.22 This consideration is clinically relevant because our patients, despite substantial weight loss, had final BMI values in the class II obesity range. Thus, future studies evaluating traditional and novel atherosclerotic risk factors, weight loss, and cardiovascular outcomes are warranted in the obese. Our findings are consistent with a recent, preliminary report demonstrating a decrease in CHD risk after bariatric surgery. Batsis et al23 categorized the decrease in CHD risk using risk models and reported improvements in CHD risk factors from the National Health and Nutrition Examination Survey (NHANES) I and the NHANES I Epidemiologic Follow-up Study, which investigated CHD mortality and hospitalization outcomes.24 The predicted 10-year risk was decreased from 37% to 18% from these estimating equations. The investigators projected that 4 deaths would be prevented for every 100 patients who underwent bariatric surgery. In contrast, our study used the FRS, which estimated total CHD risk (risk of developing angina pectoris, myocardial infarction, or coronary disease death) over the course of 10 years and found concordant results.17 We recognize several limitations in our study. There was variation in the postsurgical time points at which follow-up measurements were obtained and not all patients had completed the expected weight-loss phase after bariatric surgery. By using the final measurement available for each patient, we were able to estimate the longer term effects of weight loss on 10-year CHD risk. Many patients had limited short-term follow-up, and the full benefit of weight loss was probably not reflected in their calculated 10-year CHD risk estimate. Long-term follow-up is indicated to ascertain the effects on 10-year CHD risk estimates because reversal or stabilization of the decrease in cardiovascular risk may occur, particularly if individuals regain weight. Our study population was largely homogenous, i.e., primarily Caucasian and women. It would be beneficial to evaluate the effects of weight loss on CHD risk in a more diverse population over a longer period. We had only a small group with established CAD, so larger studies are needed to evaluate the effect of weight loss on future CHD risk and associated rates of myocardial infarction and death. We acknowledge that some changes in CHD risk may not be a direct result of bariatric surgery (e.g., decrease in smoking) and may reflect the effect of patient education, self-improvement, and overall better medical care than that of the general population.
1. Flegal KM, Carroll MD, Ogden C, Johnson CL. Prevalence and trends in obesity among US adults, 1999 –2000. JAMA 2002;288:1723–1727. 2. McCullough PA, Sandberg KR, Miller WM, Odom JS, Sloan KC, de Jong AT, Nori KE, Irving SD, Krause KR, Franklin BA. Substantial weight gain during adulthood: the road to bariatric surgery. Prev Cardiol 2005;8:155–159. 3. Fontaine KR, Redden DT, Wang C, Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA 2003;289:187–193. 4. Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in the United States, 2000. JAMA 2004;291:1238 –1245. 5. Flegal KM, Graubard BI, Williamson DF, Gail MH. Excess deaths associated with underweight, overweight, and obesity. JAMA 2005; 293:1861–1867. 6. Eckel RH, Krauss RM. American Heart Association call to action: obesity as a major risk factor for coronary heart disease. Circulation 1998;97:2099 –2100. 7. Chobanian AV, Bakris GL, Black HR, for the National Heart, Lung, and Blood Institute Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: National High Blood Pressure Education Program Coordinating Committee. The seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003;289:2560 –2572. 8. Wilson PW, D’Agostino RB, Sullivan L, Parise H, Kannel WB. Overweight and obesity as determinants of cardiovascular risk. The Framingham experience. Arch Intern Med 2002;162:1867–1872. 9. Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales VS, Marks JS. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA 2003;289:76 –79. 10. Calle EE, Thun MJ, Petrelli JM, Rodriguez C, Heath CW. Body-mass index and mortality in a prospective cohort of US adults. N Engl J Med 1999;341:1097–1105. 11. Krauss RM, Winston M. Obesity. Impact on cardiovascular disease. Circulation 1998;98:1472–1476. 12. National Task Force on Prevention and Treatment of Obesity. Overweight, obesity and health risk. Arch Intern Med 2000;160:898 –904. 13. Mitka M. Surgery for obesity: demand soars amid scientific, ethical questions. JAMA 2003;289:1761–1762. 14. Buchwald H, Avidor Y, Braunwald E, Jensen MD, Pories W, Fahrbach K, Schoelles K. Bariatric surgery: a systematic review and metaanalysis. JAMA 2004;292:1724 –1737. 15. Lara MD, Kothari SN, Sugerman HJ. Surgical management of obesity: a review of the evidence relating to the health benefits and risks. Treat Endocrinol 2005;4:55– 64. 16. Santry HP, Gillen DL, Lauderdale DS. Trends in bariatric surgical procedures. JAMA 2005;294:1909 –1917. 17. Wilson PWF, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation 1998;97:1837–1847. 18. Harder H, Dinesen B, Astrup A. The effect of a rapid weight loss on lipid profile and glycemic control in obese type 2 diabetic patients. Int J Obes Relat Metab Disord 2004;28:180 –182. 19. Lofgren I, Zern T, Herron K, West K, Sharman MJ, Volek JS, Shachter NS, Koo SI, Fernandez ML. Weight loss associated with reduced intake of carbohydrate reduces the atherogenicity of LDL in premenopausal women. Metabolism 2005;54:1133–1141. 20. Pereira MA, Swain J, Goldfine AB, Rifai N, Ludwig DS. Effects of a low-glycemic load diet on resting energy expenditure and heart disease risk factors during weight loss. JAMA 2004;292:2482–2490. 21. Sjostrom L, Lindroos AK, Peltonen M, Torgerson J, Bouchard C, Carlsson B, Dahlgren S, Larsson B, Narbro K, Sjostrom CD, Sullivan M, Wedel H. Lifestyle, diabetes, and cardiovascular risk factors 10 years after bariatric surgery. N Engl J Med 2004;351:2683–2693. 22. Mora S, Yanek LR, Moy TF, Fallin MD, Becker LC, Becker DM. Interaction of body mass index and Framingham risk score in predicting incident coronary disease in families. Circulation 2005;111:1871– 1876. 23. Batsis JA, Romero-Corral A, Collazo-Clavell ML, Sarr MG, Somers V, Brekke L, Lopez-Jimenez F. Effect of bariatric surgery on cardiovascular risk factors and predicted effect on cardiovascular events and mortality in class II–III obesity (abstract). J Am Coll Cardiol 2006; 42:358A–359A.