Evidence and temporality of the obesity paradox in coronary bypass surgery: an analysis of cause-specific mortality

Evidence and temporality of the obesity paradox in coronary bypass surgery: an analysis of... Abstract OBJECTIVES We evaluated the presence of an ‘obesity paradox’ in coronary artery bypass grafting (CABG) patients, determined its time course and ascertained whether it is associated with improved cardiovascular (CV) survival versus non-CV survival. METHODS A retrospective analysis of 3 prospectively collected databases was conducted. A fifteen-year Kaplan–Meier analysis in 7091 CABG patients was performed and repeated in 5 body mass index [BMI (kg/m2)] cohorts [Normal (18.5–24.99 kg/m2), Overweight (25–29.99 kg/m2), Obese I (30–34.99 kg/m2), Obese II (35–39.99 kg/m2) and Obese III (≥40 kg/m2)]. Mortality hazard ratios {HR [95% confidence interval (CI)]} were derived using comprehensive multivariable competing risk Cox regression, accounting for BMI categories for overall (0–15), Early (0–1), Intermediate (1–8) and Late (8–15) postoperative years, to relax the proportional hazards assumption. The regression was repeated using BMI as a continuous variable. Mortality was classified into any, CV and non-CV. RESULTS Obese patients were younger with more comorbidities. Fifteen-year survival was improved in the Overweight and Obese I groups (P < 0.001). Adjusted 15-year mortality was reduced in the Overweight [HR (95% CI) = 0.88 (0.79–0.98)] and Obese I [HR = 0.88 (0.78–0.99)] groups driven by improved CV and non-CV survival. This trend was noted in the early (Overweight) and intermediate postoperative periods (Overweight and Obese I) with no significance in the late period. Higher mortality in the Obese III [HR = 1.28 (1.06–1.55)] group was driven by a decreased CV survival. Using BMI as a continuous variable, a BMI of 29 kg/m2 was associated with optimal survival. CONCLUSIONS We identified a protective partial obesity paradox in the early and intermediate postoperative periods among Overweight and mildly obese (Obese I) patients with improved CV and non-CV survival. The morbidly obese (the Obese III group) had higher early and late CV mortality. Obesity, Coronary artery bypass grafting, Obesity paradox INTRODUCTION Obesity is a global health challenge of epidemic proportions. The World Health Organization (WHO) estimates that the rate of obesity has doubled worldwide within the last 3 decades, and currently, over 70% of American adults are either overweight or obese compared to fewer than 25% 40 years ago [1–3]. Obesity is a risk factor for cardiovascular (CV) disease including atherosclerosis, diabetes, hypertension, heart failure and atrial fibrillation [4, 5]. The estimated direct annual cost of treating obesity is $51.6 billion and another $30 billion is spent on weight reduction programmes [6, 7]. Obesity is an independent predictor of long-term morbidity and mortality due in part to its association with the cardiometabolic syndrome and likely through yet unrecognized mechanisms [8]. Although obesity is a risk factor for CV disease, it has been suggested, paradoxically, that obese patients with established CV disease may have a more favourable prognosis than the non-obese patients. This ‘obesity paradox’ has been identified in patients with renal and pulmonary disease [9, 10], patients with peripheral arterial disease [11] and patients undergoing coronary revascularization [12]. There are conflicting reports on whether an obesity paradox exists in cardiac surgery [13–22] with some studies documenting improved, equivalent or worse outcomes in the obese compared to their non-obese counterparts. A number of plausible mechanisms behind the obesity paradox have been suggested [4, 23], and although some have postulated that the obesity paradox may be explained by the observation that obese patients undergoing coronary artery bypass grafting (CABG) may have less complex coronary artery disease [24, 25], we are unaware of any studies specifically assessing whether the reported improved survival of obese CABG patients is attributable to diminished CV mortality versus non-CV mortality. Moreover, only few studies [14] have investigated the possible survival benefit of obesity after long-term follow-up (>8–10 years). Within this conceptual framework, the aim of this study is to ascertain whether an obesity paradox exists in CABG patients and, if present, determine its time course and whether it is attributable to improved CV and/or non-CV survival. METHODS Patients This investigation is a retrospective analysis of prospectively collected registries from 3 Ohio cardiac surgical centres. The centres collected data in accordance with the Society of Thoracic Surgeons (STS) Cardiac Surgery Database. The study was approved by the respective institutional review board, and informed consent was waived. No patient contact or additional review of patient records was needed. Primary CABG patients with multivessel disease receiving 2 or more bypass grafts including a left internal thoracic artery (LITA) graft were studied (1996–2012). Patients were excluded in case of a single graft, no LITA, concomitant cardiac or aortic surgery, salvage surgery, preoperative renal failure, reoperation and BMI <18.5 kg/m2 or those without a valid Social Security number. Patients undergoing concurrent carotid endarterectomy or atrial fibrillation ablation surgery were included. Patients were divided into body habitus cohorts per WHO criteria [26] based on body mass index [BMI (kg/m2)]: Normal (18.5–24.99 kg/m2), Overweight (25–29.99 kg/m2), Obese I (30–34.99 kg/m2), Obese II (35–39.99 kg/m2) and Obese III (≥40 kg/m2). The operative techniques and perioperative management have been previously described [27]. Briefly, on-pump techniques were utilized in 95% of patients. Forty-one percent of patients received multiarterial grafting using the radial artery in addition to the LITA. Aortocoronary grafting was used in more than 95% of patients. End points and follow-up The primary outcome was all-cause mortality with secondary analysis of cause-specific mortality (CV and non-CV) after competing risk adjustment across the BMI groups. Outcomes were categorized into 3 time intervals (years) [Early (0–1 year), Intermediate (1–8 years) and Late (8–15 years)] defined by inspection of the 15-year hazard rate plots. Mortality data were secured from institutional follow-up and verified from recurrent queries of the US Social Security Death Index (http://ssdi.genealogy.rootsweb.com—last checked November 2011, following which this platform was no longer a valid research tool). Death data were also obtained from the 1994–2013 Ohio Death Index for verification for determining death status after November 2011 and for derivation of the primary cause of death. Causes of death were allocated to CV category: {Cerebrovascular-[ICD10-(I60–I69), ICD9-(430–434, 436–438)]}; {Cardiovascular-[ICD10-(I00–I09, I11, I13, I20–I51), ICD9-(390–398, 402, 404, 410–429)]} and {Other Vascular-[ICD10-(I10, I12, I15, I70, I71–78, I80–I99), ICD9-(401, 403, 404, 441–448, 451–459)]} and non-CV category (all other ICD9/ICD10 codes). Patients with an unknown cause of death were analysed within all-cause mortality. Statistical analysis Continuous variables for the BMI-based cohorts are expressed as mean ± standard deviation and compared pairwise versus the Normal group (reference) using the independent t-test or the Mann–Whitney U-test based on normality. Categorical variables are expressed as counts (percentages) and compared using the χ2 test. Time-to-event analyses were truncated after 15 years due to the small number of patients who were followed up beyond 15 years. Time zero for longitudinal follow-up was defined as the time of surgery. Cumulative incidence of mortality based on all-cause, CV and non-CV deaths were estimated using the Kaplan–Meier analysis and compared for the different body habitus study groups. Hazard ratios were calculated [±95% confidence intervals (CIs)] using the competing risk Cox regression models which were applied to all patients without adjustment (unadjusted HR) and then in all patients with comprehensive risk adjustment (covariate adjusted). Demographic and risk factors were significantly different between the comparison cohorts (Table 1). Thus, covariate adjustment was done by building a non-parsimonious competing risk Cox regression model that included patient demographics and risk factors and other surgical factors included in Table 1, except for highly correlated variables (e.g. body surface area which correlates with BMI category—see the legend of Table 1). Interaction terms between BMI categorical variables and a number of patient demographic and risk factors were attempted but none were found to be significant. Revascularization index was defined as the number of grafts minus the number of diseased principal coronary arterial systems. The regression model was utilized across the BMI groups for the overall follow-up duration and then separately for each of the 3 time periods. For validation of our results, the association between mortality and BMI was also analysed using BMI as a continuous variable. On the basis of visual interpretation of the results of the primary analysis with BMI categories, we established that the relationship between BMI and mortality is non-linear as in previous studies [28]. The best fit for the relationship between overall (0–15 years) all-cause mortality and CV mortality and continuous BMI after covariate adjustment was a cubic function model. A higher 4th-order polynomial was attempted but was not justified. The relationship was plotted using beta covariates of BMI (linear, quadratic and cubic orders) after the non-parsimonious competing risk Cox regression. This model was used to study the association between BMI and mortality for the overall follow-up duration (15 years) and for the aforementioned time periods using time segmentation to relax to proportional hazard assumption. Table 1: Patient characteristics and operative data for the 5 body habitus (BMI category) groups Normal (BMI 18.5– 24.99 kg/m2) Overweight (BMI 25– 29.99 kg/m2) Obese I (BMI 30– 34.99 kg/m2) Obese II (BMI 35– 39.99 kg/m2) Obese III (BMI ≥40 kg/m2) Categorical variables, counts (%)  Number of patientsa 1312 (18.5) 2786 (39.3) 1866 (26.3) 729 (10.3) 398 (5.6)  Female 462 (35.2) 672 (24.1)* 531 (28.5)* 280 (38.4) 188 (47.2)*  Smoking 836 (63.7) 1722 (61.8) 1133 (60.7)*** 445 (61) 237 (59.5)  Diabetes 315 (24) 839 (30.1)* 785 (42.1)* 359 (49.2)* 249 (62.6)*   Insulin dependence 99 (7.5) 260 (9.3)*** 247 (13.2)* 132 (18.1)* 97 (24.4)*  Hypercholesterolaemia 816 (62.2) 1994 (71.6)* 1365 (73.2)* 565 (77.5)* 294 (73.9)*  Hypertension 977 (74.5) 2196 (78.8)* 1572 (84.2)* 615 (84.4)* 360 (90.5)*  Peripheral vascular disease 261 (19.9) 442 (15.9)* 280 (15)* 91 (12.5)* 65 (16.3)  Cerebrovascular disease 357 (27.2) 641 (23)* 412 (22.1)* 129 (17.7)* 74 (18.6)*  Chronic lung disease 339 (25.8) 491 (17.6)* 325 (17.4)* 153 (21)** 92 (23.1)  Myocardial infarction 703 (53.6) 1544 (55.4) 1018 (54.6) 367 (50.3) 218 (54.8)  Congestive heart failure 143 (10.9) 265 (9.5) 203 (10.9) 97 (13.3) 70 (17.6)*  2-Vessel disease 290 (22.1) 580 (20.8) 415 (22.2) 169 (23.2) 86 (21.6)  3-Vessel disease 1022 (77.9) 2206 (79.2) 1451 (77.8) 560 (76.8) 312 (78.4)  Left main disease 336 (25.6) 658 (23.6) 403 (21.6)* 140 (19.2)* 79 (19.8)**  Percutaneous coronary intervention 203 (15.5) 517 (18.6)** 365 (19.6)* 157 (21.5)* 66 (16.6)  Emergency 76 (5.8) 155 (5.6) 100 (5.4) 35 (4.8) 22 (5.5)  CABG + other surgeries 132 (10.1) 256 (9.2) 176 (9.4) 61 (8.4) 34 (8.5)   Other non-cardiac surgeries 101 (7.7) 165 (5.9)** 109 (5.8)** 158 (21.5)* 21 (5.3)   Other cardiac surgeries 38 (2.9) 100 (3.6) 77 (4.1)*** 40 (5.5)* 14 (3.5)  Off-pump 85 (6.5) 158 (5.7) 91 (4.9)*** 41 (5.6) 28 (7)  Surgery erab ** * *   1995–2000 562 (42.8) 1114 (40) 717 (38.4) 224 (30.7) 104 (26.1)   2000–2005 619 (47.2) 1345 (48.3) 918 (49.2) 397 (54.5) 224 (56.3)   2005–2012 131 (10) 327 (11.7) 231 (12.4) 108 (14.8) 70 (17.6)  Stroke 104 (7.9) 226 (8.1) 148 (7.9) 48 (6.6) 22 (5.5)  All arterial grafts 77 (5.9) 210 (7.5)** 172 (9.2)* 65 (8.9)** 48 (12.1)*  Bilateral internal mammary graft 31 (2.4) 87 (3.1) 55 (2.9) 12 (1.6) 2 (0.5)**  Radial artery graft 387 (29.5) 1110 (39.8)* 823 (44.1)* 369 (50.6)* 228 (57.3)*  Ejection fraction <35%b 119 (9.1) 247 (8.9) 164 (8.8) 65 (8.9) 49 (12.3)*** Continuous variables, mean ± SD  Age (years) 67.1 ± 10.8 65.2 ± 10.6* 62.9 ± 10.0* 60.4 ± 9.9* 58.9 ± 9.3*  BSA (m2)b 1.78 ± 0.17 1.98 ± 0.17* 2.13 ± 0.19* 2.25 ± 0.21* 2.41 ± 0.23*  Ejection fraction (%) 49.2 ± 11.2 49.1 ± 11.2 49.2 ± 11.1 49.2 ± 10.9 47.8 ± 11.8**  Perfusion time (min) 79 ± 34 82 ± 35* 83 ± 34* 85 ± 35* 84 ± 43**  Cross-clamp time (min)b 49 ± 23 52 ± 25* 53 ± 25* 54 ±  25* 54 ± 29*  Revascularization index 0.59 ± 0.77 0.67 ± 0.79* 0.67 ± 0.81** 0.74 ± 0.84* 0.63 ± 0.78  Number of grafts 3.37 ± 0.84 3.46 ± 0.86* 3.44 ± 0.90** 3.50 ± 0.90* 3.42 ± 0.88   Arterialb 1.39 ± 0.64 1.55 ± 0.71* 1.60 ± 0.72* 1.69 ± 0.78* 1.72 ± 0.72*   Veinb 1.98 ± 0.94 1.92 ± 0.97 ** 1.85 ± 0.99* 1.82 ± 0.95* 1.70 ± 0.99*  Year of surgeryc 7.31 ± 4.07 7.7 ± 4.13* 7.81 ± 4.10* 8.58 ± 4.10* 9.04 ± 3.99* Normal (BMI 18.5– 24.99 kg/m2) Overweight (BMI 25– 29.99 kg/m2) Obese I (BMI 30– 34.99 kg/m2) Obese II (BMI 35– 39.99 kg/m2) Obese III (BMI ≥40 kg/m2) Categorical variables, counts (%)  Number of patientsa 1312 (18.5) 2786 (39.3) 1866 (26.3) 729 (10.3) 398 (5.6)  Female 462 (35.2) 672 (24.1)* 531 (28.5)* 280 (38.4) 188 (47.2)*  Smoking 836 (63.7) 1722 (61.8) 1133 (60.7)*** 445 (61) 237 (59.5)  Diabetes 315 (24) 839 (30.1)* 785 (42.1)* 359 (49.2)* 249 (62.6)*   Insulin dependence 99 (7.5) 260 (9.3)*** 247 (13.2)* 132 (18.1)* 97 (24.4)*  Hypercholesterolaemia 816 (62.2) 1994 (71.6)* 1365 (73.2)* 565 (77.5)* 294 (73.9)*  Hypertension 977 (74.5) 2196 (78.8)* 1572 (84.2)* 615 (84.4)* 360 (90.5)*  Peripheral vascular disease 261 (19.9) 442 (15.9)* 280 (15)* 91 (12.5)* 65 (16.3)  Cerebrovascular disease 357 (27.2) 641 (23)* 412 (22.1)* 129 (17.7)* 74 (18.6)*  Chronic lung disease 339 (25.8) 491 (17.6)* 325 (17.4)* 153 (21)** 92 (23.1)  Myocardial infarction 703 (53.6) 1544 (55.4) 1018 (54.6) 367 (50.3) 218 (54.8)  Congestive heart failure 143 (10.9) 265 (9.5) 203 (10.9) 97 (13.3) 70 (17.6)*  2-Vessel disease 290 (22.1) 580 (20.8) 415 (22.2) 169 (23.2) 86 (21.6)  3-Vessel disease 1022 (77.9) 2206 (79.2) 1451 (77.8) 560 (76.8) 312 (78.4)  Left main disease 336 (25.6) 658 (23.6) 403 (21.6)* 140 (19.2)* 79 (19.8)**  Percutaneous coronary intervention 203 (15.5) 517 (18.6)** 365 (19.6)* 157 (21.5)* 66 (16.6)  Emergency 76 (5.8) 155 (5.6) 100 (5.4) 35 (4.8) 22 (5.5)  CABG + other surgeries 132 (10.1) 256 (9.2) 176 (9.4) 61 (8.4) 34 (8.5)   Other non-cardiac surgeries 101 (7.7) 165 (5.9)** 109 (5.8)** 158 (21.5)* 21 (5.3)   Other cardiac surgeries 38 (2.9) 100 (3.6) 77 (4.1)*** 40 (5.5)* 14 (3.5)  Off-pump 85 (6.5) 158 (5.7) 91 (4.9)*** 41 (5.6) 28 (7)  Surgery erab ** * *   1995–2000 562 (42.8) 1114 (40) 717 (38.4) 224 (30.7) 104 (26.1)   2000–2005 619 (47.2) 1345 (48.3) 918 (49.2) 397 (54.5) 224 (56.3)   2005–2012 131 (10) 327 (11.7) 231 (12.4) 108 (14.8) 70 (17.6)  Stroke 104 (7.9) 226 (8.1) 148 (7.9) 48 (6.6) 22 (5.5)  All arterial grafts 77 (5.9) 210 (7.5)** 172 (9.2)* 65 (8.9)** 48 (12.1)*  Bilateral internal mammary graft 31 (2.4) 87 (3.1) 55 (2.9) 12 (1.6) 2 (0.5)**  Radial artery graft 387 (29.5) 1110 (39.8)* 823 (44.1)* 369 (50.6)* 228 (57.3)*  Ejection fraction <35%b 119 (9.1) 247 (8.9) 164 (8.8) 65 (8.9) 49 (12.3)*** Continuous variables, mean ± SD  Age (years) 67.1 ± 10.8 65.2 ± 10.6* 62.9 ± 10.0* 60.4 ± 9.9* 58.9 ± 9.3*  BSA (m2)b 1.78 ± 0.17 1.98 ± 0.17* 2.13 ± 0.19* 2.25 ± 0.21* 2.41 ± 0.23*  Ejection fraction (%) 49.2 ± 11.2 49.1 ± 11.2 49.2 ± 11.1 49.2 ± 10.9 47.8 ± 11.8**  Perfusion time (min) 79 ± 34 82 ± 35* 83 ± 34* 85 ± 35* 84 ± 43**  Cross-clamp time (min)b 49 ± 23 52 ± 25* 53 ± 25* 54 ±  25* 54 ± 29*  Revascularization index 0.59 ± 0.77 0.67 ± 0.79* 0.67 ± 0.81** 0.74 ± 0.84* 0.63 ± 0.78  Number of grafts 3.37 ± 0.84 3.46 ± 0.86* 3.44 ± 0.90** 3.50 ± 0.90* 3.42 ± 0.88   Arterialb 1.39 ± 0.64 1.55 ± 0.71* 1.60 ± 0.72* 1.69 ± 0.78* 1.72 ± 0.72*   Veinb 1.98 ± 0.94 1.92 ± 0.97 ** 1.85 ± 0.99* 1.82 ± 0.95* 1.70 ± 0.99*  Year of surgeryc 7.31 ± 4.07 7.7 ± 4.13* 7.81 ± 4.10* 8.58 ± 4.10* 9.04 ± 3.99* * P-value <0.01. ** P-value = 0.01–0.05. *** P-value = 0.05–0.10. a Reflect percentage of the overall study population (%). b Variables were not used in multivariable risk adjustment as they are highly correlated with other model variables. c Year of surgery: 1 = 1994, 2 = 1996, … , 19 = 2012. BMI: body mass index; BSA: body surface area; CABG: coronary artery bypass grafting; SD: standard deviation. Table 1: Patient characteristics and operative data for the 5 body habitus (BMI category) groups Normal (BMI 18.5– 24.99 kg/m2) Overweight (BMI 25– 29.99 kg/m2) Obese I (BMI 30– 34.99 kg/m2) Obese II (BMI 35– 39.99 kg/m2) Obese III (BMI ≥40 kg/m2) Categorical variables, counts (%)  Number of patientsa 1312 (18.5) 2786 (39.3) 1866 (26.3) 729 (10.3) 398 (5.6)  Female 462 (35.2) 672 (24.1)* 531 (28.5)* 280 (38.4) 188 (47.2)*  Smoking 836 (63.7) 1722 (61.8) 1133 (60.7)*** 445 (61) 237 (59.5)  Diabetes 315 (24) 839 (30.1)* 785 (42.1)* 359 (49.2)* 249 (62.6)*   Insulin dependence 99 (7.5) 260 (9.3)*** 247 (13.2)* 132 (18.1)* 97 (24.4)*  Hypercholesterolaemia 816 (62.2) 1994 (71.6)* 1365 (73.2)* 565 (77.5)* 294 (73.9)*  Hypertension 977 (74.5) 2196 (78.8)* 1572 (84.2)* 615 (84.4)* 360 (90.5)*  Peripheral vascular disease 261 (19.9) 442 (15.9)* 280 (15)* 91 (12.5)* 65 (16.3)  Cerebrovascular disease 357 (27.2) 641 (23)* 412 (22.1)* 129 (17.7)* 74 (18.6)*  Chronic lung disease 339 (25.8) 491 (17.6)* 325 (17.4)* 153 (21)** 92 (23.1)  Myocardial infarction 703 (53.6) 1544 (55.4) 1018 (54.6) 367 (50.3) 218 (54.8)  Congestive heart failure 143 (10.9) 265 (9.5) 203 (10.9) 97 (13.3) 70 (17.6)*  2-Vessel disease 290 (22.1) 580 (20.8) 415 (22.2) 169 (23.2) 86 (21.6)  3-Vessel disease 1022 (77.9) 2206 (79.2) 1451 (77.8) 560 (76.8) 312 (78.4)  Left main disease 336 (25.6) 658 (23.6) 403 (21.6)* 140 (19.2)* 79 (19.8)**  Percutaneous coronary intervention 203 (15.5) 517 (18.6)** 365 (19.6)* 157 (21.5)* 66 (16.6)  Emergency 76 (5.8) 155 (5.6) 100 (5.4) 35 (4.8) 22 (5.5)  CABG + other surgeries 132 (10.1) 256 (9.2) 176 (9.4) 61 (8.4) 34 (8.5)   Other non-cardiac surgeries 101 (7.7) 165 (5.9)** 109 (5.8)** 158 (21.5)* 21 (5.3)   Other cardiac surgeries 38 (2.9) 100 (3.6) 77 (4.1)*** 40 (5.5)* 14 (3.5)  Off-pump 85 (6.5) 158 (5.7) 91 (4.9)*** 41 (5.6) 28 (7)  Surgery erab ** * *   1995–2000 562 (42.8) 1114 (40) 717 (38.4) 224 (30.7) 104 (26.1)   2000–2005 619 (47.2) 1345 (48.3) 918 (49.2) 397 (54.5) 224 (56.3)   2005–2012 131 (10) 327 (11.7) 231 (12.4) 108 (14.8) 70 (17.6)  Stroke 104 (7.9) 226 (8.1) 148 (7.9) 48 (6.6) 22 (5.5)  All arterial grafts 77 (5.9) 210 (7.5)** 172 (9.2)* 65 (8.9)** 48 (12.1)*  Bilateral internal mammary graft 31 (2.4) 87 (3.1) 55 (2.9) 12 (1.6) 2 (0.5)**  Radial artery graft 387 (29.5) 1110 (39.8)* 823 (44.1)* 369 (50.6)* 228 (57.3)*  Ejection fraction <35%b 119 (9.1) 247 (8.9) 164 (8.8) 65 (8.9) 49 (12.3)*** Continuous variables, mean ± SD  Age (years) 67.1 ± 10.8 65.2 ± 10.6* 62.9 ± 10.0* 60.4 ± 9.9* 58.9 ± 9.3*  BSA (m2)b 1.78 ± 0.17 1.98 ± 0.17* 2.13 ± 0.19* 2.25 ± 0.21* 2.41 ± 0.23*  Ejection fraction (%) 49.2 ± 11.2 49.1 ± 11.2 49.2 ± 11.1 49.2 ± 10.9 47.8 ± 11.8**  Perfusion time (min) 79 ± 34 82 ± 35* 83 ± 34* 85 ± 35* 84 ± 43**  Cross-clamp time (min)b 49 ± 23 52 ± 25* 53 ± 25* 54 ±  25* 54 ± 29*  Revascularization index 0.59 ± 0.77 0.67 ± 0.79* 0.67 ± 0.81** 0.74 ± 0.84* 0.63 ± 0.78  Number of grafts 3.37 ± 0.84 3.46 ± 0.86* 3.44 ± 0.90** 3.50 ± 0.90* 3.42 ± 0.88   Arterialb 1.39 ± 0.64 1.55 ± 0.71* 1.60 ± 0.72* 1.69 ± 0.78* 1.72 ± 0.72*   Veinb 1.98 ± 0.94 1.92 ± 0.97 ** 1.85 ± 0.99* 1.82 ± 0.95* 1.70 ± 0.99*  Year of surgeryc 7.31 ± 4.07 7.7 ± 4.13* 7.81 ± 4.10* 8.58 ± 4.10* 9.04 ± 3.99* Normal (BMI 18.5– 24.99 kg/m2) Overweight (BMI 25– 29.99 kg/m2) Obese I (BMI 30– 34.99 kg/m2) Obese II (BMI 35– 39.99 kg/m2) Obese III (BMI ≥40 kg/m2) Categorical variables, counts (%)  Number of patientsa 1312 (18.5) 2786 (39.3) 1866 (26.3) 729 (10.3) 398 (5.6)  Female 462 (35.2) 672 (24.1)* 531 (28.5)* 280 (38.4) 188 (47.2)*  Smoking 836 (63.7) 1722 (61.8) 1133 (60.7)*** 445 (61) 237 (59.5)  Diabetes 315 (24) 839 (30.1)* 785 (42.1)* 359 (49.2)* 249 (62.6)*   Insulin dependence 99 (7.5) 260 (9.3)*** 247 (13.2)* 132 (18.1)* 97 (24.4)*  Hypercholesterolaemia 816 (62.2) 1994 (71.6)* 1365 (73.2)* 565 (77.5)* 294 (73.9)*  Hypertension 977 (74.5) 2196 (78.8)* 1572 (84.2)* 615 (84.4)* 360 (90.5)*  Peripheral vascular disease 261 (19.9) 442 (15.9)* 280 (15)* 91 (12.5)* 65 (16.3)  Cerebrovascular disease 357 (27.2) 641 (23)* 412 (22.1)* 129 (17.7)* 74 (18.6)*  Chronic lung disease 339 (25.8) 491 (17.6)* 325 (17.4)* 153 (21)** 92 (23.1)  Myocardial infarction 703 (53.6) 1544 (55.4) 1018 (54.6) 367 (50.3) 218 (54.8)  Congestive heart failure 143 (10.9) 265 (9.5) 203 (10.9) 97 (13.3) 70 (17.6)*  2-Vessel disease 290 (22.1) 580 (20.8) 415 (22.2) 169 (23.2) 86 (21.6)  3-Vessel disease 1022 (77.9) 2206 (79.2) 1451 (77.8) 560 (76.8) 312 (78.4)  Left main disease 336 (25.6) 658 (23.6) 403 (21.6)* 140 (19.2)* 79 (19.8)**  Percutaneous coronary intervention 203 (15.5) 517 (18.6)** 365 (19.6)* 157 (21.5)* 66 (16.6)  Emergency 76 (5.8) 155 (5.6) 100 (5.4) 35 (4.8) 22 (5.5)  CABG + other surgeries 132 (10.1) 256 (9.2) 176 (9.4) 61 (8.4) 34 (8.5)   Other non-cardiac surgeries 101 (7.7) 165 (5.9)** 109 (5.8)** 158 (21.5)* 21 (5.3)   Other cardiac surgeries 38 (2.9) 100 (3.6) 77 (4.1)*** 40 (5.5)* 14 (3.5)  Off-pump 85 (6.5) 158 (5.7) 91 (4.9)*** 41 (5.6) 28 (7)  Surgery erab ** * *   1995–2000 562 (42.8) 1114 (40) 717 (38.4) 224 (30.7) 104 (26.1)   2000–2005 619 (47.2) 1345 (48.3) 918 (49.2) 397 (54.5) 224 (56.3)   2005–2012 131 (10) 327 (11.7) 231 (12.4) 108 (14.8) 70 (17.6)  Stroke 104 (7.9) 226 (8.1) 148 (7.9) 48 (6.6) 22 (5.5)  All arterial grafts 77 (5.9) 210 (7.5)** 172 (9.2)* 65 (8.9)** 48 (12.1)*  Bilateral internal mammary graft 31 (2.4) 87 (3.1) 55 (2.9) 12 (1.6) 2 (0.5)**  Radial artery graft 387 (29.5) 1110 (39.8)* 823 (44.1)* 369 (50.6)* 228 (57.3)*  Ejection fraction <35%b 119 (9.1) 247 (8.9) 164 (8.8) 65 (8.9) 49 (12.3)*** Continuous variables, mean ± SD  Age (years) 67.1 ± 10.8 65.2 ± 10.6* 62.9 ± 10.0* 60.4 ± 9.9* 58.9 ± 9.3*  BSA (m2)b 1.78 ± 0.17 1.98 ± 0.17* 2.13 ± 0.19* 2.25 ± 0.21* 2.41 ± 0.23*  Ejection fraction (%) 49.2 ± 11.2 49.1 ± 11.2 49.2 ± 11.1 49.2 ± 10.9 47.8 ± 11.8**  Perfusion time (min) 79 ± 34 82 ± 35* 83 ± 34* 85 ± 35* 84 ± 43**  Cross-clamp time (min)b 49 ± 23 52 ± 25* 53 ± 25* 54 ±  25* 54 ± 29*  Revascularization index 0.59 ± 0.77 0.67 ± 0.79* 0.67 ± 0.81** 0.74 ± 0.84* 0.63 ± 0.78  Number of grafts 3.37 ± 0.84 3.46 ± 0.86* 3.44 ± 0.90** 3.50 ± 0.90* 3.42 ± 0.88   Arterialb 1.39 ± 0.64 1.55 ± 0.71* 1.60 ± 0.72* 1.69 ± 0.78* 1.72 ± 0.72*   Veinb 1.98 ± 0.94 1.92 ± 0.97 ** 1.85 ± 0.99* 1.82 ± 0.95* 1.70 ± 0.99*  Year of surgeryc 7.31 ± 4.07 7.7 ± 4.13* 7.81 ± 4.10* 8.58 ± 4.10* 9.04 ± 3.99* * P-value <0.01. ** P-value = 0.01–0.05. *** P-value = 0.05–0.10. a Reflect percentage of the overall study population (%). b Variables were not used in multivariable risk adjustment as they are highly correlated with other model variables. c Year of surgery: 1 = 1994, 2 = 1996, … , 19 = 2012. BMI: body mass index; BSA: body surface area; CABG: coronary artery bypass grafting; SD: standard deviation. A 2-sided P-value <0.05 was used uniformly to indicate significance. Statistical analysis was conducted using the IBM SPSS Statistics for Windows, Version 23.0 (Armonk, NY: IBM Corp. and StataCorp. 2017) and the Stata Statistical Software: Release 15 (College Station, TX: StataCorp LLC) for competing risk analysis. RESULTS The study population consisted of 7091 patients including 1312 patients in the Normal group (18.6%), 2786 patients in the Overweight group (39.3%), 1866 patients in the Obese I group (26.3%), 729 patients in the Obese II group (10.3%) and 398 patients in the Obese III group (5.6%). Prevalence of diabetes, hypertension and heart failure [Normal (10.9%) vs Obese III (17.6%)] was progressively higher with increasing BMI despite age being systematically lower [Normal (67.1 ± 10.8) vs Obese III (58.9 ± 9.3); years] (Table 1). Multiarterial grafts were frequently used in this series, and the rate increased with BMI given their younger age [Normal (29.5%) vs Obese III (57.3%)]. Mean follow-up was 9.5 ± 4.6 years. A total of 4063 (57.3%) deaths were documented overall, of which CV deaths accounted for 20.3%, non-CV deaths for 28.4% and 8.6% of deaths were of unknown cause. Unadjusted 30-day mortality was 1.38% and was similar across BMI categories [Normal (1.68%), Overweight (1.29%), Obese I (1.23%), Obese II (1.23%) and Obese III (1.37%)]. Overall and cause-specific mortality cumulative incidence functions Figure 1 shows the cumulative incidence functions for all-cause mortality partitioned to its CV, non-CV and unknown cause components for the entire population and then for the Normal, Overweight, Obese I, Obese II and Obese III subgroups. Compared to Normal (61.0%), the estimated 15-year unadjusted all-cause mortality was significantly lower for the Overweight [(56.6%), P < 0.001], Obese I [(54.7%), P < 0.001] and Obese II [(57.2%), P = 0.007] groups, whereas it was similar for the Obese III [(62.8%), P = 0.91] group. Figure 1: View largeDownload slide Fifteen-year cumulative incidence functions showing all-cause mortality and cause-specific mortality for all patients (n = 7091) and for all the 5 body habitus sub-cohorts. CABG: coronary artery bypass grafting; CV: cardiovascular; non-CV: non-cardiovascular; Unk: unknown cause of death. Figure 1: View largeDownload slide Fifteen-year cumulative incidence functions showing all-cause mortality and cause-specific mortality for all patients (n = 7091) and for all the 5 body habitus sub-cohorts. CABG: coronary artery bypass grafting; CV: cardiovascular; non-CV: non-cardiovascular; Unk: unknown cause of death. The unadjusted and comprehensively covariate-adjusted hazard ratios across BMI categories are summarized in Fig. 2—with ‘Normal’ designated as the reference group and accounting for competing risks in case of cause-specific (CV and non-CV) mortality calculations. Figure 2 also shows results assuming proportional hazard over the entire 15-year follow-up (overall) and after relaxing this assumption by time segmenting the analysis to Early mortality, Intermediate mortality and Late mortality. Generally, the unadjusted all-cause, CV and non-CV death data showed a significant and appreciable partial obesity paradox [Fig. 2 (left panels)] whereby Overweight and Obese I and Obese II groups showed superior outcomes (mortality hazard ratios <1), whereas the Obese III group had unadjusted hazard ratios similar to the Normal group. This paradox was (i) the largest in case of early outcomes, (ii) present but relatively reduced in the intermediate term (1–8 years) and (iii) absent in the late term. Figure 2: View largeDownload slide Comparison of mortality risk stratified by the body habitus groups for all-cause, CV and non-CV deaths shown for entire follow-up [overall (0–15 years) and for time-segmented follow-up (Early: 0–1 year, Intermediate: 1–8 years and Late: 8–15 years)]. Open symbols indicate unadjusted hazard ratios with or without accounting for competing risk of death, closed symbols indicate comprehensive risk-adjusted hazard ratios with or without accounting for competing risk of death and error bars represent 95% CIs. CI: confidence interval; CV: cardiovascular; HR: hazard ratio; Ob: obese; OW: overweight. Figure 2: View largeDownload slide Comparison of mortality risk stratified by the body habitus groups for all-cause, CV and non-CV deaths shown for entire follow-up [overall (0–15 years) and for time-segmented follow-up (Early: 0–1 year, Intermediate: 1–8 years and Late: 8–15 years)]. Open symbols indicate unadjusted hazard ratios with or without accounting for competing risk of death, closed symbols indicate comprehensive risk-adjusted hazard ratios with or without accounting for competing risk of death and error bars represent 95% CIs. CI: confidence interval; CV: cardiovascular; HR: hazard ratio; Ob: obese; OW: overweight. Risk-adjustment modified the mortality outcome—BMI relationships such that the magnitude of the paradoxical (improved) outcomes at higher BMI were generally reduced in magnitude and limited to Overweight and Obese I patients, whereas severely obese patients (Obese III) showed significantly worse outcomes than Normal BMI. The complementary time-segmented and cause-specific analysis suggested that a higher BMI paradox was mostly due to superior intermediate survival. The overall 15-year mortality demonstrated the presence of the J-shaped protective partial ‘obesity paradox’ in the Overweight and Obese I showing lower mortality as evidenced by the following adjusted hazard ratios (AHR): [AHR (95% CI) = 0.88 (0.79–0.98), P = 0.017 and AHR = 0.88 (0.78–0.99), P = 0.032, respectively]. The morbidly obese patients (Obese III) had a higher mortality [AHR = 1.28 (1.06–1.55), P = 0.012]. The Obese II group showed an increased trend for overall mortality, but this did not reach statistical significance [AHR = 1.13 (0.97–1.33)]. The protective nature of the Overweight and Obese I category was driven by decreased CV mortality and non-CV mortality although this trend did not reach significance in the Obese I patients (Fig. 2). The detrimental impact on survival in the Obese III category was driven by increased CV mortality only [AHR = 1.47 (1.08–2.00), P = 0.015] (Table 2; Fig. 2). Table 2: Overall and time-segmented risk-adjusted hazard ratios for all-cause and cause-specific mortality across the BMI groups PH regression, HR (95% CI) Time-segmented analysis, HR (95% CI) Overall (0–15 years) Early (0–1 year) Intermediate (1–8 years) Late (8–15 years) All deaths  N (ref) 1 1 1 1  Overweight 0.88 (0.79–0.98) 0.75 (0.56–1.00) 0.82 (0.71–0.96) 0.99 (0.83–1.18)  Obese I 0.88 (0.78–0.99) 0.84 (0.6–1.19) 0.80 (0.67–0.95) 0.97 (0.80–1.18)  Obese II 1.13 (0.97–1.33) 0.74 (0.44–1.22) 1.16 (0.94–1.44) 1.21 (0.93–1.57)  Obese III 1.28 (1.06–1.55) 1.91 (1.19–3.06) 1.24 (0.95–1.62) 1.08 (0.77–1.52) CV deaths  N (ref) 1 1 1 1  Overweight 0.83 (0.70–1.00) 0.76 (0.53–1.09) 0.61 (0.46–0.81) 1.19 (0.87–1.62)  Obese I 0.90 (0.73–1.10) 0.87 (0.57–1.34) 0.73 (0.53–1.00) 1.09 (0.77–1.55)  Obese II 1.11 (0.85–1.45) 0.99 (0.56–1.76) 0.95 (0.64–1.42) 1.43 (0.90–2.27)  Obese III 1.47 (1.08–2.00) 2.31 (1.31–4.07) 1.20 (0.77–1.86) 1.20 (0.64–2.24) Non-CV deaths  N (ref) 1 1 1 1  Overweight 0.86 (0.74–1.00) 0.57 (0.31–1.07) 0.85 (0.69–1.04) 0.82 (0.64–1.04)  Obese I 0.85 (0.72–1.01) 0.60 (0.30–1.23) 0.77 (0.61–0.98) 0.91 (0.69–1.19)  Obese II 1.02 (0.81–1.28) 0.24 (0.05–1.08) 1.22 (0.91–1.64) 0.90 (0.62–1.30)  Obese III 0.88 (0.65–1.18) 1.01 (0.34–3.04) 1.07 (0.72–1.58) 0.74 (0.45–1.22) PH regression, HR (95% CI) Time-segmented analysis, HR (95% CI) Overall (0–15 years) Early (0–1 year) Intermediate (1–8 years) Late (8–15 years) All deaths  N (ref) 1 1 1 1  Overweight 0.88 (0.79–0.98) 0.75 (0.56–1.00) 0.82 (0.71–0.96) 0.99 (0.83–1.18)  Obese I 0.88 (0.78–0.99) 0.84 (0.6–1.19) 0.80 (0.67–0.95) 0.97 (0.80–1.18)  Obese II 1.13 (0.97–1.33) 0.74 (0.44–1.22) 1.16 (0.94–1.44) 1.21 (0.93–1.57)  Obese III 1.28 (1.06–1.55) 1.91 (1.19–3.06) 1.24 (0.95–1.62) 1.08 (0.77–1.52) CV deaths  N (ref) 1 1 1 1  Overweight 0.83 (0.70–1.00) 0.76 (0.53–1.09) 0.61 (0.46–0.81) 1.19 (0.87–1.62)  Obese I 0.90 (0.73–1.10) 0.87 (0.57–1.34) 0.73 (0.53–1.00) 1.09 (0.77–1.55)  Obese II 1.11 (0.85–1.45) 0.99 (0.56–1.76) 0.95 (0.64–1.42) 1.43 (0.90–2.27)  Obese III 1.47 (1.08–2.00) 2.31 (1.31–4.07) 1.20 (0.77–1.86) 1.20 (0.64–2.24) Non-CV deaths  N (ref) 1 1 1 1  Overweight 0.86 (0.74–1.00) 0.57 (0.31–1.07) 0.85 (0.69–1.04) 0.82 (0.64–1.04)  Obese I 0.85 (0.72–1.01) 0.60 (0.30–1.23) 0.77 (0.61–0.98) 0.91 (0.69–1.19)  Obese II 1.02 (0.81–1.28) 0.24 (0.05–1.08) 1.22 (0.91–1.64) 0.90 (0.62–1.30)  Obese III 0.88 (0.65–1.18) 1.01 (0.34–3.04) 1.07 (0.72–1.58) 0.74 (0.45–1.22) CI: confidence interval; CV: cardiovascular; HR: hazard ratio; PH: proportional hazard. Bold values indicate significance at P-value = 0.05. Table 2: Overall and time-segmented risk-adjusted hazard ratios for all-cause and cause-specific mortality across the BMI groups PH regression, HR (95% CI) Time-segmented analysis, HR (95% CI) Overall (0–15 years) Early (0–1 year) Intermediate (1–8 years) Late (8–15 years) All deaths  N (ref) 1 1 1 1  Overweight 0.88 (0.79–0.98) 0.75 (0.56–1.00) 0.82 (0.71–0.96) 0.99 (0.83–1.18)  Obese I 0.88 (0.78–0.99) 0.84 (0.6–1.19) 0.80 (0.67–0.95) 0.97 (0.80–1.18)  Obese II 1.13 (0.97–1.33) 0.74 (0.44–1.22) 1.16 (0.94–1.44) 1.21 (0.93–1.57)  Obese III 1.28 (1.06–1.55) 1.91 (1.19–3.06) 1.24 (0.95–1.62) 1.08 (0.77–1.52) CV deaths  N (ref) 1 1 1 1  Overweight 0.83 (0.70–1.00) 0.76 (0.53–1.09) 0.61 (0.46–0.81) 1.19 (0.87–1.62)  Obese I 0.90 (0.73–1.10) 0.87 (0.57–1.34) 0.73 (0.53–1.00) 1.09 (0.77–1.55)  Obese II 1.11 (0.85–1.45) 0.99 (0.56–1.76) 0.95 (0.64–1.42) 1.43 (0.90–2.27)  Obese III 1.47 (1.08–2.00) 2.31 (1.31–4.07) 1.20 (0.77–1.86) 1.20 (0.64–2.24) Non-CV deaths  N (ref) 1 1 1 1  Overweight 0.86 (0.74–1.00) 0.57 (0.31–1.07) 0.85 (0.69–1.04) 0.82 (0.64–1.04)  Obese I 0.85 (0.72–1.01) 0.60 (0.30–1.23) 0.77 (0.61–0.98) 0.91 (0.69–1.19)  Obese II 1.02 (0.81–1.28) 0.24 (0.05–1.08) 1.22 (0.91–1.64) 0.90 (0.62–1.30)  Obese III 0.88 (0.65–1.18) 1.01 (0.34–3.04) 1.07 (0.72–1.58) 0.74 (0.45–1.22) PH regression, HR (95% CI) Time-segmented analysis, HR (95% CI) Overall (0–15 years) Early (0–1 year) Intermediate (1–8 years) Late (8–15 years) All deaths  N (ref) 1 1 1 1  Overweight 0.88 (0.79–0.98) 0.75 (0.56–1.00) 0.82 (0.71–0.96) 0.99 (0.83–1.18)  Obese I 0.88 (0.78–0.99) 0.84 (0.6–1.19) 0.80 (0.67–0.95) 0.97 (0.80–1.18)  Obese II 1.13 (0.97–1.33) 0.74 (0.44–1.22) 1.16 (0.94–1.44) 1.21 (0.93–1.57)  Obese III 1.28 (1.06–1.55) 1.91 (1.19–3.06) 1.24 (0.95–1.62) 1.08 (0.77–1.52) CV deaths  N (ref) 1 1 1 1  Overweight 0.83 (0.70–1.00) 0.76 (0.53–1.09) 0.61 (0.46–0.81) 1.19 (0.87–1.62)  Obese I 0.90 (0.73–1.10) 0.87 (0.57–1.34) 0.73 (0.53–1.00) 1.09 (0.77–1.55)  Obese II 1.11 (0.85–1.45) 0.99 (0.56–1.76) 0.95 (0.64–1.42) 1.43 (0.90–2.27)  Obese III 1.47 (1.08–2.00) 2.31 (1.31–4.07) 1.20 (0.77–1.86) 1.20 (0.64–2.24) Non-CV deaths  N (ref) 1 1 1 1  Overweight 0.86 (0.74–1.00) 0.57 (0.31–1.07) 0.85 (0.69–1.04) 0.82 (0.64–1.04)  Obese I 0.85 (0.72–1.01) 0.60 (0.30–1.23) 0.77 (0.61–0.98) 0.91 (0.69–1.19)  Obese II 1.02 (0.81–1.28) 0.24 (0.05–1.08) 1.22 (0.91–1.64) 0.90 (0.62–1.30)  Obese III 0.88 (0.65–1.18) 1.01 (0.34–3.04) 1.07 (0.72–1.58) 0.74 (0.45–1.22) CI: confidence interval; CV: cardiovascular; HR: hazard ratio; PH: proportional hazard. Bold values indicate significance at P-value = 0.05. Comprehensive multivariable Cox regression model analyses using BMI as a continuous non-linear variable corroborated the 5 BMI category analyses (Fig. 3). A cubic function was found to be the best fit for the overall follow-up analysis; however, a quadratic function was the best fit for some of the time-segmented and cause-specific death analyses. The late follow-up time period showed no significant correlation with BMI as continuous function regardless of the function used (linear, quadratic and cubic). In the overall (0–15 years), early (0–1 year) and intermediate (1–8 years) analyses, there was evidence of the consistent obesity paradox with the adjusted hazard ratios at slightly above normal BMI [25–30], being less than 1. Also, with the exception of early non-CV mortality, it appears that a higher BMI (>37 kg/m2) is associated with worse outcome (Fig. 3). Based on the continuous model, a BMI between 28.5 and 29 kg/m2 is associated with the best outcomes, mostly due to less early mortality and intermediate-term mortality. Figure 3: View largeDownload slide Association between BMI and mortality (all-cause, CV and non-CV) across the BMI spectrum (continuous variable with cubic outcome association) for overall and time-segmented follow-up periods following multivariable competing risk regression. Shaded area represents 95% CI. Symbol with error bars represent HR with 95% CIs for BMI when categorized into the body habitus groups (same as risk-adjusted data in Fig. 2). Dagger represents results of model fit with a forced BMI cubic function (a1*BMI ± a2*BMI2 ± a3*BMI3), where the model coefficient a3 (for BMI3) was not significant, whereas a2 (for BMI2) was significant and double dagger indicates that the BMI outcome association is not significant for any order. BMI: body mass index; CI: confidence interval; CV: cardiovascular; HR: hazard ratio. Figure 3: View largeDownload slide Association between BMI and mortality (all-cause, CV and non-CV) across the BMI spectrum (continuous variable with cubic outcome association) for overall and time-segmented follow-up periods following multivariable competing risk regression. Shaded area represents 95% CI. Symbol with error bars represent HR with 95% CIs for BMI when categorized into the body habitus groups (same as risk-adjusted data in Fig. 2). Dagger represents results of model fit with a forced BMI cubic function (a1*BMI ± a2*BMI2 ± a3*BMI3), where the model coefficient a3 (for BMI3) was not significant, whereas a2 (for BMI2) was significant and double dagger indicates that the BMI outcome association is not significant for any order. BMI: body mass index; CI: confidence interval; CV: cardiovascular; HR: hazard ratio. DISCUSSION Our study supports the existence of a long-term (15 years) protective partial obesity paradox in CABG patients as evidenced by improved survival in the Overweight and Obese I cohorts, which appeared in the Early postoperative period in Overweight patients and in the Intermediate postoperative period in both the Overweight and Obese I patients. Using the competing risk analysis, the obesity paradox was associated with both CV and non-CV mortality risk reduction during the entire 0–15-year follow-up period in Overweight patients and during the Intermediate postoperative time period in Obese I patients. The morbidly obese patients (Obese III) were found to have increased mortality risk primarily due to worse CV survival in the early postoperative time interval. These results agree with previous reports documenting an obesity paradox after CABG in higher BMI patients [15, 16, 18, 19, 28] and worse survival in the morbidly obese patients [13, 18]. Stamou et al. [15] identified an obesity paradox among cardiac surgery patients noting a decreased short-term mortality and intermediate-term mortality among overweight patients. van Straten et al. [18] noted a long-term survival advantage in overweight CABG patients. In a Veterans Administration CABG population, Wagner et al. [28] noted optimal acute perioperative survival in patients with a BMI of 30 kg/m2. The identified partial obesity paradox characterized by a ‘J’-shaped relationship between our 5 BMI groups and long-term survival conflicts with other reports [12, 14, 17, 21] that found no such protective effect of increased BMI on survival. The discrepant findings between those studies and ours could be due to multiple factors including small sample sizes and short follow-up periods in the other studies [17], as well as using different BMI categorization [14] and not taking into account the continuous but non-linear relationship between outcome and BMI. Gurm et al. [12] analysed the impact of obesity on acute mortality and 5-year mortality in patients undergoing CABG and percutaneous intervention in the Bypass Angioplasty Revascularization Investigation (BARI) trial. In contrast to our observation, they identified that among overweight and mildly obese patients, CV mortality risk was increased, whereas all-cause mortality was unaffected by increasing BMI, suggesting, in fact, that non-CV mortality was reduced. Importantly, the BARI cohort did not reflect contemporary treatment modalities (low LITA use rate of 75–86%) as it included patients from the 1990s, and operative deaths were not included in the analysis. Le-Bert et al. [17], despite claims of identifying an obesity paradox in a small cohort of elderly CABG patients, did not find an association between BMI and acute in-hospital mortality. Prabhakar et al. [21], using the Society of Thoracic Surgeons Database-and relying only on a small number of covariates for a logistic regression analysis, found an increased perioperative mortality risk among mildly obese patients [HR = 1.21 (1.45–1.73)]. Long-term mortality was not analysed, and their patient population was characterized by low LITA utilization. As our study population included only LITA-based CABG with an appreciable rate of multiarterial grafting (41.6%), the 2 study populations are not comparable. Importantly, our outcomes were adjusted for the number of arterial grafts and the revascularization index, both of which were generally higher with rising BMI. Not doing so would have further exaggerated the protective effects of the obesity paradox. Benedetto et al. [14] found no effect of obesity on acute perioperative mortality, an observation that agrees with our findings. In propensity-matched groups, an overweight BMI was not protective against long-term mortality [HR = 1.05 (0.90–1.08)], and in contrast to our study, mildly obese patients (our Obese I cohort) had an increased risk of mortality [HR = 1.22 (1.07–2.66)] during a mean follow-up period of 7.6 years. The covariates in the propensity matching algorithm used by Benedetto et al. were limited to those included in the EuroSCORE and did not include specific grafting strategies (single arterial or multiarterial) or the revascularization index. They were unable to corroborate their outcomes in a subsequent Cox regression analysis, noting that overweight patients had decreased risk of long-term mortality [HR = 0.88 (0.80–0.97)], whereas mild obesity and morbid obesity did not influence long-term survival. They only all-cause peri-operative and long-term mortality without details on specific causes of death. The physiological basis of the obesity paradox remains undefined and is beyond the scope of this study, but a number of possible hypotheses have been proposed. Among these are the development of earlier but less complex coronary artery disease [24] (less left main coronary disease with increasing BMI in our series), altered cytokine and neuroendocrine profiles among obese patients with a favourable CV profile, increased adipose-tissue-derived soluble tumour necrosis factor (TNF) receptor production with a resultant decrease in the pathogenic effects of TNF, lower levels of atrial naturetic peptides, attenuated sympathetic and renin–angiotensin activity, decreased systemic vascular resistance and enhanced metabolic reserves that maybe beneficial during stress-induced catabolic states such as the postoperative period [23]. The fact that weight-loss programmes have been associated with increased adverse outcomes maybe considered as indirect evidence for the existence of an obesity paradox [29]. Our findings of increased mortality in the morbidly obese (Obese III) CABG patients confirms those of multiple other reports [12–14, 18, 21, 30], and this was principally driven by increased CV mortality within the first postoperative year. The pathophysiology behind this observation is also undefined and beyond the scope of this study but is unlikely related to postoperative complications, as the acute perioperative outcomes were equivalent and uniformly excellent across all BMI categories. When survival was analysed as a function of BMI as a continuous variable, we identified a BMI of 29 kg/m2 corresponding to the lowest overall mortality rate (Fig. 3). In such an analysis, rather than a ‘J’-shaped relationship between survival and the 5 distinct BMI groups, the survival curves showed a U-shaped relationship between BMI and mortality with increased risk at both extremes, a finding that we have reported on earlier [13]. Benedetto et al. [14] similarly identified a BMI range of 27–29 kg/m2 as being maximally protective for late mortality. Wagner et al. [28] noted a similar U-shaped relationship between acute perioperative mortality and BMI with a nadir mortality point at a BMI of 30 kg/m2. A question that remains unanswered is whether delaying surgery to affect weight reduction may ameliorate the increased mortality risk among the morbidly obese patients. Limitations Our study has a number of limitations. As this is a retrospective analysis, only associations can be established, and we are unable to assess the specific physiological mechanisms that may account for our findings. All observational studies are subject to well-recognized biases including patient selection bias, and conclusions might be altered by exclusion of covariates that may be pertinent to the analysis but, nevertheless, are not included in the study data set. In addition, although BMI is a convenient and widely used parameter to assess obesity, other parameters of obesity such as adiposity, body fatness, waist circumference and weight-to-height ratio, all of which have been shown to impact mortality [23], were not used in our analysis, and it is not known how, or if at all, the inclusion of these measures of obesity may have impacted our conclusions. Furthermore, it is important to note that body mass index is not a static patient characteristic and is subject to change over time. Any changes in BMI over the study period were unavailable to us, and only the values at the index operation were utilized in the analysis. Also, we are unable to assess the patients’ long-term medication use and risk reduction strategies on long-term survival. The impact of any repeat revascularization strategies on long-term survival was not included in our study. Finally, BMI does not provide a distinction between all obese patients and metabolically healthy obese patients who may have a distinctly different prognosis for any given BMI [23]. CONCLUSION In conclusion, our study demonstrates the presence of a partial obesity paradox up to 15 years postoperatively in CABG patients. Patients who are overweight and mildly obese show improved survival, with a decreased risk of CV mortality and non-CV mortality in the entire 0–15-year postoperative period and in the 1–8-year postoperative period, respectively. Conversely, morbidly obese patients had significantly worse survival driven by increased CV mortality within a year of the index operation. Conflict of interest: none declared. REFERENCES 1 Flegal KM , Carroll MD , Ogden CL , Johnson CL. Prevalence and trends in obesity among US adults, 1999-2000 . JAMA 2002 ; 288 : 1723 – 7 . Google Scholar CrossRef Search ADS PubMed 2 Ogden CL , Carroll MD , Kit BK , Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011-2012 . JAMA 2014 ; 311 : 806 – 14 . Google Scholar CrossRef Search ADS PubMed 3 Manson JE , Bassuk SS. Obesity in the United States: a fresh look at its high toll . JAMA 2003 ; 289 : 229 – 30 . Google Scholar CrossRef Search ADS PubMed 4 Lavie CJ , Milani RV , Ventura HO. Obesity and cardiovascular disease: risk factor, paradox, and impact of weight loss . J Am Coll Cardiol 2009 ; 53 : 1925 – 32 . Google Scholar CrossRef Search ADS PubMed 5 5, Zacharias A , Schwann TA , Riordan CJ , Durham SJ , Shah AS , Habib RH. Obesity and risk of new-onset atrial fibrillation after cardiac surgery . Circulation 2005 ; 112 : 3247 – 55 . Google Scholar CrossRef Search ADS PubMed 6 National Institutes of Diabetes and Digestive and Kidney Diseases. Statistics Related to Overweight and Obesity. NIH Publ No.96-4158. Update 2000 . 7 Rosenbaum M , Leibel RL , Hirsch J. Obesity . N Engl J Med 1997 ; 337 : 396 – 407 . Google Scholar CrossRef Search ADS PubMed 8 Poirier P , Giles TD , Bray GA , Hong Y , Stern JS , Pi SFX. Obesity and cardiovascular disease: pathophysiology, evaluation, and effect of weight loss: an update of the 1997 American Heart Association Scientific Statement on Obesity and Heart Disease from the Obesity Committee of the Council on Nutrition, Physical Activity, and Metabolism . Circulation 2006 ; 113 : 898 – 918 . Google Scholar CrossRef Search ADS PubMed 9 Park J , Ahmadi S-F , Streja E , Molnar MZ , Flegal KM , Gillen D et al. . Obesity paradox in end-stage kidney disease patients . Prog Cardiovasc Dis 2014 ; 56 : 415 – 25 . Google Scholar CrossRef Search ADS PubMed 10 Chittal P , Babu AS , Lavie CJ. Obesity paradox: does fat alter outcomes in chronic obstructive pulmonary disease? COPD 2015 ; 12 : 14 – 18 . Google Scholar CrossRef Search ADS PubMed 11 Galal W , van Gestel YRBM , Hoeks SE , Sin DD , Winkel TA , Bax JJ et al. . The obesity paradox in patients with peripheral arterial disease: the influence of chronic obstructive pulmonary disease . Chest 2008 ; 134 : 925 – 30 . Google Scholar CrossRef Search ADS PubMed 12 Gurm HS , Whitlow PL , Kip KE ; BARI Investigators . The impact of body mass index on short and long-term outcomes in patients undergoing coronary revascularization . J Am Coll Cardiol 2002 ; 39 : 834 – 40 . Google Scholar CrossRef Search ADS PubMed 13 Habib RH , Zacharias A , Schwann TA , Riordan CJ , Durham SJ , Shah A. Effects of obesity and small body size on operative and long-term outcomes of coronary artery bypass surgery: a propensity-matched analysis . Ann Thorac Surg 2005 ; 79 : 1976 – 86 . Google Scholar CrossRef Search ADS PubMed 14 Benedetto U , Danese C , Codispoti M. Obesity paradox in coronary artery bypass grafting: myth or reality? J Thorac Cardiovasc Surg 2014 ; 147 : 1517 – 23 . Google Scholar CrossRef Search ADS PubMed 15 Stamou SC , Nussbaum M , Stiegel RM , Reames MK , Skipper ER , Robicsek F et al. . Effect of body mass index on outcomes after cardiac surgery: is there an obesity paradox? Ann Thorac Surg 2011 ; 91 : 42 – 7 . Google Scholar CrossRef Search ADS PubMed 16 Sung S-H , Wu T-C , Huang C-H , Lin S-J , Chen J-W. Prognostic impact of body mass index in patients undergoing coronary artery bypass surgery . Heart 2011 ; 97 : 648 – 54 . Google Scholar CrossRef Search ADS PubMed 17 Le-Bert G , Santana O , Pineda AM , Zamora C , Lamas GA , Lamelas J. The obesity paradox in elderly obese patients undergoing coronary artery bypass surgery . Interact CardioVasc Thorac Surg 2011 ; 13 : 124 – 7 . Google Scholar CrossRef Search ADS PubMed 18 van Straten AHM , Bramer S , Soliman Hamad MA , van Zundert AAJ , Martens EJ , Schönberger JPAM et al. . Effect of body mass index on early and late mortality after coronary artery bypass grafting . Ann Thorac Surg 2010 ; 89 : 30 – 7 . Google Scholar CrossRef Search ADS PubMed 19 Gruberg L , Mercado N , Milo S , Boersma E , Disco C , van Es G-A et al. . Impact of body mass index on the outcome of patients with multivessel disease randomized to either coronary artery bypass grafting or stenting in the ARTS trial: the obesity paradox II? Am J Cardiol 2005 ; 95 : 439 – 44 . Google Scholar CrossRef Search ADS PubMed 20 Alam M , Siddiqui S , Lee V-V , Elayda MA , Nambi V , Yang EY et al. . Isolated coronary artery bypass grafting in obese individuals: a propensity matched analysis of outcomes . Circ J 2011 ; 75 : 1378 – 85 . Google Scholar CrossRef Search ADS PubMed 21 Prabhakar G , Haan CK , Peterson ED , Coombs LP , Cruzzavala JL , Murray GF. The risks of moderate and extreme obesity for coronary artery bypass grafting outcomes: a study from the Society of Thoracic Surgeons Database . Ann Thorac Surg 2002 ; 74 : 1125 – 31 . Google Scholar CrossRef Search ADS PubMed 22 Hartrumpf M , Kuehnel RU , Albes JM. The obesity paradox is still there: a risk analysis of over 15 000 cardiosurgical patients based on body mass index . Interact CardioVasc Thorac Surg 2017 ; 25 : 18 – 24 . Google Scholar CrossRef Search ADS PubMed 23 Lavie CJ , McAuley PA , Church TS , Milani RV , Blair SN. Obesity and cardiovascular diseases: implications regarding fitness, fatness, and severity in the obesity paradox . J Am Coll Cardiol 2014 ; 63 : 1345 – 54 . Google Scholar CrossRef Search ADS PubMed 24 Sturm R , Wells KB. Does obesity contribute as much to morbidity as poverty or smoking? Public Health 2001 ; 115 : 229 – 35 . Google Scholar CrossRef Search ADS PubMed 25 Lavie CJ , Messerli FH. Cardiovascular adaption to obesity and hypertension . Chest 1986 ; 90 : 275 – 9 . Google Scholar CrossRef Search ADS PubMed 26 Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser., 2000 ; 894: i–xii, 1–253. 27 Schwann TA , Habib RH , Zacharias A , Parenteau GL , Riordan CJ , Durham SJ et al. . Effects of body size on operative, intermediate, and long-term outcomes after coronary artery bypass operation . Ann Thorac Surg 2001 ; 71 : 521 – 30 . Google Scholar CrossRef Search ADS PubMed 28 Wagner BD , Grunwald GK , Rumsfeld JS , Hill JO , Ho PM , Wyatt HR et al. . Relationship of body mass index with outcomes after coronary artery bypass graft surgery . Ann Thorac Surg 2007 ; 84 : 10 – 16 . Google Scholar CrossRef Search ADS PubMed 29 Fonarow GC , Horwich TB , Hamilton MA , MacLellan WR , Tillisch JH. Obesity, weight reduction, and survival in heart failure: reply . J Am Coll Cardiol 2002 ; 39 : 1563 – 4 . Google Scholar CrossRef Search ADS PubMed 30 Sun X , Hill PC , Bafi AS , Garcia JM , Haile E , Corso PJ et al. . Is cardiac surgery safe in extremely obese patients (body mass index 50 or greater)? Ann Thorac Surg 2009 ; 87 : 540 – 7 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Cardio-Thoracic Surgery Oxford University Press

Evidence and temporality of the obesity paradox in coronary bypass surgery: an analysis of cause-specific mortality

Loading next page...
 
/lp/ou_press/evidence-and-temporality-of-the-obesity-paradox-in-coronary-bypass-PRyDpZ6O2t
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ISSN
1010-7940
eISSN
1873-734X
D.O.I.
10.1093/ejcts/ezy207
Publisher site
See Article on Publisher Site

Abstract

Abstract OBJECTIVES We evaluated the presence of an ‘obesity paradox’ in coronary artery bypass grafting (CABG) patients, determined its time course and ascertained whether it is associated with improved cardiovascular (CV) survival versus non-CV survival. METHODS A retrospective analysis of 3 prospectively collected databases was conducted. A fifteen-year Kaplan–Meier analysis in 7091 CABG patients was performed and repeated in 5 body mass index [BMI (kg/m2)] cohorts [Normal (18.5–24.99 kg/m2), Overweight (25–29.99 kg/m2), Obese I (30–34.99 kg/m2), Obese II (35–39.99 kg/m2) and Obese III (≥40 kg/m2)]. Mortality hazard ratios {HR [95% confidence interval (CI)]} were derived using comprehensive multivariable competing risk Cox regression, accounting for BMI categories for overall (0–15), Early (0–1), Intermediate (1–8) and Late (8–15) postoperative years, to relax the proportional hazards assumption. The regression was repeated using BMI as a continuous variable. Mortality was classified into any, CV and non-CV. RESULTS Obese patients were younger with more comorbidities. Fifteen-year survival was improved in the Overweight and Obese I groups (P < 0.001). Adjusted 15-year mortality was reduced in the Overweight [HR (95% CI) = 0.88 (0.79–0.98)] and Obese I [HR = 0.88 (0.78–0.99)] groups driven by improved CV and non-CV survival. This trend was noted in the early (Overweight) and intermediate postoperative periods (Overweight and Obese I) with no significance in the late period. Higher mortality in the Obese III [HR = 1.28 (1.06–1.55)] group was driven by a decreased CV survival. Using BMI as a continuous variable, a BMI of 29 kg/m2 was associated with optimal survival. CONCLUSIONS We identified a protective partial obesity paradox in the early and intermediate postoperative periods among Overweight and mildly obese (Obese I) patients with improved CV and non-CV survival. The morbidly obese (the Obese III group) had higher early and late CV mortality. Obesity, Coronary artery bypass grafting, Obesity paradox INTRODUCTION Obesity is a global health challenge of epidemic proportions. The World Health Organization (WHO) estimates that the rate of obesity has doubled worldwide within the last 3 decades, and currently, over 70% of American adults are either overweight or obese compared to fewer than 25% 40 years ago [1–3]. Obesity is a risk factor for cardiovascular (CV) disease including atherosclerosis, diabetes, hypertension, heart failure and atrial fibrillation [4, 5]. The estimated direct annual cost of treating obesity is $51.6 billion and another $30 billion is spent on weight reduction programmes [6, 7]. Obesity is an independent predictor of long-term morbidity and mortality due in part to its association with the cardiometabolic syndrome and likely through yet unrecognized mechanisms [8]. Although obesity is a risk factor for CV disease, it has been suggested, paradoxically, that obese patients with established CV disease may have a more favourable prognosis than the non-obese patients. This ‘obesity paradox’ has been identified in patients with renal and pulmonary disease [9, 10], patients with peripheral arterial disease [11] and patients undergoing coronary revascularization [12]. There are conflicting reports on whether an obesity paradox exists in cardiac surgery [13–22] with some studies documenting improved, equivalent or worse outcomes in the obese compared to their non-obese counterparts. A number of plausible mechanisms behind the obesity paradox have been suggested [4, 23], and although some have postulated that the obesity paradox may be explained by the observation that obese patients undergoing coronary artery bypass grafting (CABG) may have less complex coronary artery disease [24, 25], we are unaware of any studies specifically assessing whether the reported improved survival of obese CABG patients is attributable to diminished CV mortality versus non-CV mortality. Moreover, only few studies [14] have investigated the possible survival benefit of obesity after long-term follow-up (>8–10 years). Within this conceptual framework, the aim of this study is to ascertain whether an obesity paradox exists in CABG patients and, if present, determine its time course and whether it is attributable to improved CV and/or non-CV survival. METHODS Patients This investigation is a retrospective analysis of prospectively collected registries from 3 Ohio cardiac surgical centres. The centres collected data in accordance with the Society of Thoracic Surgeons (STS) Cardiac Surgery Database. The study was approved by the respective institutional review board, and informed consent was waived. No patient contact or additional review of patient records was needed. Primary CABG patients with multivessel disease receiving 2 or more bypass grafts including a left internal thoracic artery (LITA) graft were studied (1996–2012). Patients were excluded in case of a single graft, no LITA, concomitant cardiac or aortic surgery, salvage surgery, preoperative renal failure, reoperation and BMI <18.5 kg/m2 or those without a valid Social Security number. Patients undergoing concurrent carotid endarterectomy or atrial fibrillation ablation surgery were included. Patients were divided into body habitus cohorts per WHO criteria [26] based on body mass index [BMI (kg/m2)]: Normal (18.5–24.99 kg/m2), Overweight (25–29.99 kg/m2), Obese I (30–34.99 kg/m2), Obese II (35–39.99 kg/m2) and Obese III (≥40 kg/m2). The operative techniques and perioperative management have been previously described [27]. Briefly, on-pump techniques were utilized in 95% of patients. Forty-one percent of patients received multiarterial grafting using the radial artery in addition to the LITA. Aortocoronary grafting was used in more than 95% of patients. End points and follow-up The primary outcome was all-cause mortality with secondary analysis of cause-specific mortality (CV and non-CV) after competing risk adjustment across the BMI groups. Outcomes were categorized into 3 time intervals (years) [Early (0–1 year), Intermediate (1–8 years) and Late (8–15 years)] defined by inspection of the 15-year hazard rate plots. Mortality data were secured from institutional follow-up and verified from recurrent queries of the US Social Security Death Index (http://ssdi.genealogy.rootsweb.com—last checked November 2011, following which this platform was no longer a valid research tool). Death data were also obtained from the 1994–2013 Ohio Death Index for verification for determining death status after November 2011 and for derivation of the primary cause of death. Causes of death were allocated to CV category: {Cerebrovascular-[ICD10-(I60–I69), ICD9-(430–434, 436–438)]}; {Cardiovascular-[ICD10-(I00–I09, I11, I13, I20–I51), ICD9-(390–398, 402, 404, 410–429)]} and {Other Vascular-[ICD10-(I10, I12, I15, I70, I71–78, I80–I99), ICD9-(401, 403, 404, 441–448, 451–459)]} and non-CV category (all other ICD9/ICD10 codes). Patients with an unknown cause of death were analysed within all-cause mortality. Statistical analysis Continuous variables for the BMI-based cohorts are expressed as mean ± standard deviation and compared pairwise versus the Normal group (reference) using the independent t-test or the Mann–Whitney U-test based on normality. Categorical variables are expressed as counts (percentages) and compared using the χ2 test. Time-to-event analyses were truncated after 15 years due to the small number of patients who were followed up beyond 15 years. Time zero for longitudinal follow-up was defined as the time of surgery. Cumulative incidence of mortality based on all-cause, CV and non-CV deaths were estimated using the Kaplan–Meier analysis and compared for the different body habitus study groups. Hazard ratios were calculated [±95% confidence intervals (CIs)] using the competing risk Cox regression models which were applied to all patients without adjustment (unadjusted HR) and then in all patients with comprehensive risk adjustment (covariate adjusted). Demographic and risk factors were significantly different between the comparison cohorts (Table 1). Thus, covariate adjustment was done by building a non-parsimonious competing risk Cox regression model that included patient demographics and risk factors and other surgical factors included in Table 1, except for highly correlated variables (e.g. body surface area which correlates with BMI category—see the legend of Table 1). Interaction terms between BMI categorical variables and a number of patient demographic and risk factors were attempted but none were found to be significant. Revascularization index was defined as the number of grafts minus the number of diseased principal coronary arterial systems. The regression model was utilized across the BMI groups for the overall follow-up duration and then separately for each of the 3 time periods. For validation of our results, the association between mortality and BMI was also analysed using BMI as a continuous variable. On the basis of visual interpretation of the results of the primary analysis with BMI categories, we established that the relationship between BMI and mortality is non-linear as in previous studies [28]. The best fit for the relationship between overall (0–15 years) all-cause mortality and CV mortality and continuous BMI after covariate adjustment was a cubic function model. A higher 4th-order polynomial was attempted but was not justified. The relationship was plotted using beta covariates of BMI (linear, quadratic and cubic orders) after the non-parsimonious competing risk Cox regression. This model was used to study the association between BMI and mortality for the overall follow-up duration (15 years) and for the aforementioned time periods using time segmentation to relax to proportional hazard assumption. Table 1: Patient characteristics and operative data for the 5 body habitus (BMI category) groups Normal (BMI 18.5– 24.99 kg/m2) Overweight (BMI 25– 29.99 kg/m2) Obese I (BMI 30– 34.99 kg/m2) Obese II (BMI 35– 39.99 kg/m2) Obese III (BMI ≥40 kg/m2) Categorical variables, counts (%)  Number of patientsa 1312 (18.5) 2786 (39.3) 1866 (26.3) 729 (10.3) 398 (5.6)  Female 462 (35.2) 672 (24.1)* 531 (28.5)* 280 (38.4) 188 (47.2)*  Smoking 836 (63.7) 1722 (61.8) 1133 (60.7)*** 445 (61) 237 (59.5)  Diabetes 315 (24) 839 (30.1)* 785 (42.1)* 359 (49.2)* 249 (62.6)*   Insulin dependence 99 (7.5) 260 (9.3)*** 247 (13.2)* 132 (18.1)* 97 (24.4)*  Hypercholesterolaemia 816 (62.2) 1994 (71.6)* 1365 (73.2)* 565 (77.5)* 294 (73.9)*  Hypertension 977 (74.5) 2196 (78.8)* 1572 (84.2)* 615 (84.4)* 360 (90.5)*  Peripheral vascular disease 261 (19.9) 442 (15.9)* 280 (15)* 91 (12.5)* 65 (16.3)  Cerebrovascular disease 357 (27.2) 641 (23)* 412 (22.1)* 129 (17.7)* 74 (18.6)*  Chronic lung disease 339 (25.8) 491 (17.6)* 325 (17.4)* 153 (21)** 92 (23.1)  Myocardial infarction 703 (53.6) 1544 (55.4) 1018 (54.6) 367 (50.3) 218 (54.8)  Congestive heart failure 143 (10.9) 265 (9.5) 203 (10.9) 97 (13.3) 70 (17.6)*  2-Vessel disease 290 (22.1) 580 (20.8) 415 (22.2) 169 (23.2) 86 (21.6)  3-Vessel disease 1022 (77.9) 2206 (79.2) 1451 (77.8) 560 (76.8) 312 (78.4)  Left main disease 336 (25.6) 658 (23.6) 403 (21.6)* 140 (19.2)* 79 (19.8)**  Percutaneous coronary intervention 203 (15.5) 517 (18.6)** 365 (19.6)* 157 (21.5)* 66 (16.6)  Emergency 76 (5.8) 155 (5.6) 100 (5.4) 35 (4.8) 22 (5.5)  CABG + other surgeries 132 (10.1) 256 (9.2) 176 (9.4) 61 (8.4) 34 (8.5)   Other non-cardiac surgeries 101 (7.7) 165 (5.9)** 109 (5.8)** 158 (21.5)* 21 (5.3)   Other cardiac surgeries 38 (2.9) 100 (3.6) 77 (4.1)*** 40 (5.5)* 14 (3.5)  Off-pump 85 (6.5) 158 (5.7) 91 (4.9)*** 41 (5.6) 28 (7)  Surgery erab ** * *   1995–2000 562 (42.8) 1114 (40) 717 (38.4) 224 (30.7) 104 (26.1)   2000–2005 619 (47.2) 1345 (48.3) 918 (49.2) 397 (54.5) 224 (56.3)   2005–2012 131 (10) 327 (11.7) 231 (12.4) 108 (14.8) 70 (17.6)  Stroke 104 (7.9) 226 (8.1) 148 (7.9) 48 (6.6) 22 (5.5)  All arterial grafts 77 (5.9) 210 (7.5)** 172 (9.2)* 65 (8.9)** 48 (12.1)*  Bilateral internal mammary graft 31 (2.4) 87 (3.1) 55 (2.9) 12 (1.6) 2 (0.5)**  Radial artery graft 387 (29.5) 1110 (39.8)* 823 (44.1)* 369 (50.6)* 228 (57.3)*  Ejection fraction <35%b 119 (9.1) 247 (8.9) 164 (8.8) 65 (8.9) 49 (12.3)*** Continuous variables, mean ± SD  Age (years) 67.1 ± 10.8 65.2 ± 10.6* 62.9 ± 10.0* 60.4 ± 9.9* 58.9 ± 9.3*  BSA (m2)b 1.78 ± 0.17 1.98 ± 0.17* 2.13 ± 0.19* 2.25 ± 0.21* 2.41 ± 0.23*  Ejection fraction (%) 49.2 ± 11.2 49.1 ± 11.2 49.2 ± 11.1 49.2 ± 10.9 47.8 ± 11.8**  Perfusion time (min) 79 ± 34 82 ± 35* 83 ± 34* 85 ± 35* 84 ± 43**  Cross-clamp time (min)b 49 ± 23 52 ± 25* 53 ± 25* 54 ±  25* 54 ± 29*  Revascularization index 0.59 ± 0.77 0.67 ± 0.79* 0.67 ± 0.81** 0.74 ± 0.84* 0.63 ± 0.78  Number of grafts 3.37 ± 0.84 3.46 ± 0.86* 3.44 ± 0.90** 3.50 ± 0.90* 3.42 ± 0.88   Arterialb 1.39 ± 0.64 1.55 ± 0.71* 1.60 ± 0.72* 1.69 ± 0.78* 1.72 ± 0.72*   Veinb 1.98 ± 0.94 1.92 ± 0.97 ** 1.85 ± 0.99* 1.82 ± 0.95* 1.70 ± 0.99*  Year of surgeryc 7.31 ± 4.07 7.7 ± 4.13* 7.81 ± 4.10* 8.58 ± 4.10* 9.04 ± 3.99* Normal (BMI 18.5– 24.99 kg/m2) Overweight (BMI 25– 29.99 kg/m2) Obese I (BMI 30– 34.99 kg/m2) Obese II (BMI 35– 39.99 kg/m2) Obese III (BMI ≥40 kg/m2) Categorical variables, counts (%)  Number of patientsa 1312 (18.5) 2786 (39.3) 1866 (26.3) 729 (10.3) 398 (5.6)  Female 462 (35.2) 672 (24.1)* 531 (28.5)* 280 (38.4) 188 (47.2)*  Smoking 836 (63.7) 1722 (61.8) 1133 (60.7)*** 445 (61) 237 (59.5)  Diabetes 315 (24) 839 (30.1)* 785 (42.1)* 359 (49.2)* 249 (62.6)*   Insulin dependence 99 (7.5) 260 (9.3)*** 247 (13.2)* 132 (18.1)* 97 (24.4)*  Hypercholesterolaemia 816 (62.2) 1994 (71.6)* 1365 (73.2)* 565 (77.5)* 294 (73.9)*  Hypertension 977 (74.5) 2196 (78.8)* 1572 (84.2)* 615 (84.4)* 360 (90.5)*  Peripheral vascular disease 261 (19.9) 442 (15.9)* 280 (15)* 91 (12.5)* 65 (16.3)  Cerebrovascular disease 357 (27.2) 641 (23)* 412 (22.1)* 129 (17.7)* 74 (18.6)*  Chronic lung disease 339 (25.8) 491 (17.6)* 325 (17.4)* 153 (21)** 92 (23.1)  Myocardial infarction 703 (53.6) 1544 (55.4) 1018 (54.6) 367 (50.3) 218 (54.8)  Congestive heart failure 143 (10.9) 265 (9.5) 203 (10.9) 97 (13.3) 70 (17.6)*  2-Vessel disease 290 (22.1) 580 (20.8) 415 (22.2) 169 (23.2) 86 (21.6)  3-Vessel disease 1022 (77.9) 2206 (79.2) 1451 (77.8) 560 (76.8) 312 (78.4)  Left main disease 336 (25.6) 658 (23.6) 403 (21.6)* 140 (19.2)* 79 (19.8)**  Percutaneous coronary intervention 203 (15.5) 517 (18.6)** 365 (19.6)* 157 (21.5)* 66 (16.6)  Emergency 76 (5.8) 155 (5.6) 100 (5.4) 35 (4.8) 22 (5.5)  CABG + other surgeries 132 (10.1) 256 (9.2) 176 (9.4) 61 (8.4) 34 (8.5)   Other non-cardiac surgeries 101 (7.7) 165 (5.9)** 109 (5.8)** 158 (21.5)* 21 (5.3)   Other cardiac surgeries 38 (2.9) 100 (3.6) 77 (4.1)*** 40 (5.5)* 14 (3.5)  Off-pump 85 (6.5) 158 (5.7) 91 (4.9)*** 41 (5.6) 28 (7)  Surgery erab ** * *   1995–2000 562 (42.8) 1114 (40) 717 (38.4) 224 (30.7) 104 (26.1)   2000–2005 619 (47.2) 1345 (48.3) 918 (49.2) 397 (54.5) 224 (56.3)   2005–2012 131 (10) 327 (11.7) 231 (12.4) 108 (14.8) 70 (17.6)  Stroke 104 (7.9) 226 (8.1) 148 (7.9) 48 (6.6) 22 (5.5)  All arterial grafts 77 (5.9) 210 (7.5)** 172 (9.2)* 65 (8.9)** 48 (12.1)*  Bilateral internal mammary graft 31 (2.4) 87 (3.1) 55 (2.9) 12 (1.6) 2 (0.5)**  Radial artery graft 387 (29.5) 1110 (39.8)* 823 (44.1)* 369 (50.6)* 228 (57.3)*  Ejection fraction <35%b 119 (9.1) 247 (8.9) 164 (8.8) 65 (8.9) 49 (12.3)*** Continuous variables, mean ± SD  Age (years) 67.1 ± 10.8 65.2 ± 10.6* 62.9 ± 10.0* 60.4 ± 9.9* 58.9 ± 9.3*  BSA (m2)b 1.78 ± 0.17 1.98 ± 0.17* 2.13 ± 0.19* 2.25 ± 0.21* 2.41 ± 0.23*  Ejection fraction (%) 49.2 ± 11.2 49.1 ± 11.2 49.2 ± 11.1 49.2 ± 10.9 47.8 ± 11.8**  Perfusion time (min) 79 ± 34 82 ± 35* 83 ± 34* 85 ± 35* 84 ± 43**  Cross-clamp time (min)b 49 ± 23 52 ± 25* 53 ± 25* 54 ±  25* 54 ± 29*  Revascularization index 0.59 ± 0.77 0.67 ± 0.79* 0.67 ± 0.81** 0.74 ± 0.84* 0.63 ± 0.78  Number of grafts 3.37 ± 0.84 3.46 ± 0.86* 3.44 ± 0.90** 3.50 ± 0.90* 3.42 ± 0.88   Arterialb 1.39 ± 0.64 1.55 ± 0.71* 1.60 ± 0.72* 1.69 ± 0.78* 1.72 ± 0.72*   Veinb 1.98 ± 0.94 1.92 ± 0.97 ** 1.85 ± 0.99* 1.82 ± 0.95* 1.70 ± 0.99*  Year of surgeryc 7.31 ± 4.07 7.7 ± 4.13* 7.81 ± 4.10* 8.58 ± 4.10* 9.04 ± 3.99* * P-value <0.01. ** P-value = 0.01–0.05. *** P-value = 0.05–0.10. a Reflect percentage of the overall study population (%). b Variables were not used in multivariable risk adjustment as they are highly correlated with other model variables. c Year of surgery: 1 = 1994, 2 = 1996, … , 19 = 2012. BMI: body mass index; BSA: body surface area; CABG: coronary artery bypass grafting; SD: standard deviation. Table 1: Patient characteristics and operative data for the 5 body habitus (BMI category) groups Normal (BMI 18.5– 24.99 kg/m2) Overweight (BMI 25– 29.99 kg/m2) Obese I (BMI 30– 34.99 kg/m2) Obese II (BMI 35– 39.99 kg/m2) Obese III (BMI ≥40 kg/m2) Categorical variables, counts (%)  Number of patientsa 1312 (18.5) 2786 (39.3) 1866 (26.3) 729 (10.3) 398 (5.6)  Female 462 (35.2) 672 (24.1)* 531 (28.5)* 280 (38.4) 188 (47.2)*  Smoking 836 (63.7) 1722 (61.8) 1133 (60.7)*** 445 (61) 237 (59.5)  Diabetes 315 (24) 839 (30.1)* 785 (42.1)* 359 (49.2)* 249 (62.6)*   Insulin dependence 99 (7.5) 260 (9.3)*** 247 (13.2)* 132 (18.1)* 97 (24.4)*  Hypercholesterolaemia 816 (62.2) 1994 (71.6)* 1365 (73.2)* 565 (77.5)* 294 (73.9)*  Hypertension 977 (74.5) 2196 (78.8)* 1572 (84.2)* 615 (84.4)* 360 (90.5)*  Peripheral vascular disease 261 (19.9) 442 (15.9)* 280 (15)* 91 (12.5)* 65 (16.3)  Cerebrovascular disease 357 (27.2) 641 (23)* 412 (22.1)* 129 (17.7)* 74 (18.6)*  Chronic lung disease 339 (25.8) 491 (17.6)* 325 (17.4)* 153 (21)** 92 (23.1)  Myocardial infarction 703 (53.6) 1544 (55.4) 1018 (54.6) 367 (50.3) 218 (54.8)  Congestive heart failure 143 (10.9) 265 (9.5) 203 (10.9) 97 (13.3) 70 (17.6)*  2-Vessel disease 290 (22.1) 580 (20.8) 415 (22.2) 169 (23.2) 86 (21.6)  3-Vessel disease 1022 (77.9) 2206 (79.2) 1451 (77.8) 560 (76.8) 312 (78.4)  Left main disease 336 (25.6) 658 (23.6) 403 (21.6)* 140 (19.2)* 79 (19.8)**  Percutaneous coronary intervention 203 (15.5) 517 (18.6)** 365 (19.6)* 157 (21.5)* 66 (16.6)  Emergency 76 (5.8) 155 (5.6) 100 (5.4) 35 (4.8) 22 (5.5)  CABG + other surgeries 132 (10.1) 256 (9.2) 176 (9.4) 61 (8.4) 34 (8.5)   Other non-cardiac surgeries 101 (7.7) 165 (5.9)** 109 (5.8)** 158 (21.5)* 21 (5.3)   Other cardiac surgeries 38 (2.9) 100 (3.6) 77 (4.1)*** 40 (5.5)* 14 (3.5)  Off-pump 85 (6.5) 158 (5.7) 91 (4.9)*** 41 (5.6) 28 (7)  Surgery erab ** * *   1995–2000 562 (42.8) 1114 (40) 717 (38.4) 224 (30.7) 104 (26.1)   2000–2005 619 (47.2) 1345 (48.3) 918 (49.2) 397 (54.5) 224 (56.3)   2005–2012 131 (10) 327 (11.7) 231 (12.4) 108 (14.8) 70 (17.6)  Stroke 104 (7.9) 226 (8.1) 148 (7.9) 48 (6.6) 22 (5.5)  All arterial grafts 77 (5.9) 210 (7.5)** 172 (9.2)* 65 (8.9)** 48 (12.1)*  Bilateral internal mammary graft 31 (2.4) 87 (3.1) 55 (2.9) 12 (1.6) 2 (0.5)**  Radial artery graft 387 (29.5) 1110 (39.8)* 823 (44.1)* 369 (50.6)* 228 (57.3)*  Ejection fraction <35%b 119 (9.1) 247 (8.9) 164 (8.8) 65 (8.9) 49 (12.3)*** Continuous variables, mean ± SD  Age (years) 67.1 ± 10.8 65.2 ± 10.6* 62.9 ± 10.0* 60.4 ± 9.9* 58.9 ± 9.3*  BSA (m2)b 1.78 ± 0.17 1.98 ± 0.17* 2.13 ± 0.19* 2.25 ± 0.21* 2.41 ± 0.23*  Ejection fraction (%) 49.2 ± 11.2 49.1 ± 11.2 49.2 ± 11.1 49.2 ± 10.9 47.8 ± 11.8**  Perfusion time (min) 79 ± 34 82 ± 35* 83 ± 34* 85 ± 35* 84 ± 43**  Cross-clamp time (min)b 49 ± 23 52 ± 25* 53 ± 25* 54 ±  25* 54 ± 29*  Revascularization index 0.59 ± 0.77 0.67 ± 0.79* 0.67 ± 0.81** 0.74 ± 0.84* 0.63 ± 0.78  Number of grafts 3.37 ± 0.84 3.46 ± 0.86* 3.44 ± 0.90** 3.50 ± 0.90* 3.42 ± 0.88   Arterialb 1.39 ± 0.64 1.55 ± 0.71* 1.60 ± 0.72* 1.69 ± 0.78* 1.72 ± 0.72*   Veinb 1.98 ± 0.94 1.92 ± 0.97 ** 1.85 ± 0.99* 1.82 ± 0.95* 1.70 ± 0.99*  Year of surgeryc 7.31 ± 4.07 7.7 ± 4.13* 7.81 ± 4.10* 8.58 ± 4.10* 9.04 ± 3.99* Normal (BMI 18.5– 24.99 kg/m2) Overweight (BMI 25– 29.99 kg/m2) Obese I (BMI 30– 34.99 kg/m2) Obese II (BMI 35– 39.99 kg/m2) Obese III (BMI ≥40 kg/m2) Categorical variables, counts (%)  Number of patientsa 1312 (18.5) 2786 (39.3) 1866 (26.3) 729 (10.3) 398 (5.6)  Female 462 (35.2) 672 (24.1)* 531 (28.5)* 280 (38.4) 188 (47.2)*  Smoking 836 (63.7) 1722 (61.8) 1133 (60.7)*** 445 (61) 237 (59.5)  Diabetes 315 (24) 839 (30.1)* 785 (42.1)* 359 (49.2)* 249 (62.6)*   Insulin dependence 99 (7.5) 260 (9.3)*** 247 (13.2)* 132 (18.1)* 97 (24.4)*  Hypercholesterolaemia 816 (62.2) 1994 (71.6)* 1365 (73.2)* 565 (77.5)* 294 (73.9)*  Hypertension 977 (74.5) 2196 (78.8)* 1572 (84.2)* 615 (84.4)* 360 (90.5)*  Peripheral vascular disease 261 (19.9) 442 (15.9)* 280 (15)* 91 (12.5)* 65 (16.3)  Cerebrovascular disease 357 (27.2) 641 (23)* 412 (22.1)* 129 (17.7)* 74 (18.6)*  Chronic lung disease 339 (25.8) 491 (17.6)* 325 (17.4)* 153 (21)** 92 (23.1)  Myocardial infarction 703 (53.6) 1544 (55.4) 1018 (54.6) 367 (50.3) 218 (54.8)  Congestive heart failure 143 (10.9) 265 (9.5) 203 (10.9) 97 (13.3) 70 (17.6)*  2-Vessel disease 290 (22.1) 580 (20.8) 415 (22.2) 169 (23.2) 86 (21.6)  3-Vessel disease 1022 (77.9) 2206 (79.2) 1451 (77.8) 560 (76.8) 312 (78.4)  Left main disease 336 (25.6) 658 (23.6) 403 (21.6)* 140 (19.2)* 79 (19.8)**  Percutaneous coronary intervention 203 (15.5) 517 (18.6)** 365 (19.6)* 157 (21.5)* 66 (16.6)  Emergency 76 (5.8) 155 (5.6) 100 (5.4) 35 (4.8) 22 (5.5)  CABG + other surgeries 132 (10.1) 256 (9.2) 176 (9.4) 61 (8.4) 34 (8.5)   Other non-cardiac surgeries 101 (7.7) 165 (5.9)** 109 (5.8)** 158 (21.5)* 21 (5.3)   Other cardiac surgeries 38 (2.9) 100 (3.6) 77 (4.1)*** 40 (5.5)* 14 (3.5)  Off-pump 85 (6.5) 158 (5.7) 91 (4.9)*** 41 (5.6) 28 (7)  Surgery erab ** * *   1995–2000 562 (42.8) 1114 (40) 717 (38.4) 224 (30.7) 104 (26.1)   2000–2005 619 (47.2) 1345 (48.3) 918 (49.2) 397 (54.5) 224 (56.3)   2005–2012 131 (10) 327 (11.7) 231 (12.4) 108 (14.8) 70 (17.6)  Stroke 104 (7.9) 226 (8.1) 148 (7.9) 48 (6.6) 22 (5.5)  All arterial grafts 77 (5.9) 210 (7.5)** 172 (9.2)* 65 (8.9)** 48 (12.1)*  Bilateral internal mammary graft 31 (2.4) 87 (3.1) 55 (2.9) 12 (1.6) 2 (0.5)**  Radial artery graft 387 (29.5) 1110 (39.8)* 823 (44.1)* 369 (50.6)* 228 (57.3)*  Ejection fraction <35%b 119 (9.1) 247 (8.9) 164 (8.8) 65 (8.9) 49 (12.3)*** Continuous variables, mean ± SD  Age (years) 67.1 ± 10.8 65.2 ± 10.6* 62.9 ± 10.0* 60.4 ± 9.9* 58.9 ± 9.3*  BSA (m2)b 1.78 ± 0.17 1.98 ± 0.17* 2.13 ± 0.19* 2.25 ± 0.21* 2.41 ± 0.23*  Ejection fraction (%) 49.2 ± 11.2 49.1 ± 11.2 49.2 ± 11.1 49.2 ± 10.9 47.8 ± 11.8**  Perfusion time (min) 79 ± 34 82 ± 35* 83 ± 34* 85 ± 35* 84 ± 43**  Cross-clamp time (min)b 49 ± 23 52 ± 25* 53 ± 25* 54 ±  25* 54 ± 29*  Revascularization index 0.59 ± 0.77 0.67 ± 0.79* 0.67 ± 0.81** 0.74 ± 0.84* 0.63 ± 0.78  Number of grafts 3.37 ± 0.84 3.46 ± 0.86* 3.44 ± 0.90** 3.50 ± 0.90* 3.42 ± 0.88   Arterialb 1.39 ± 0.64 1.55 ± 0.71* 1.60 ± 0.72* 1.69 ± 0.78* 1.72 ± 0.72*   Veinb 1.98 ± 0.94 1.92 ± 0.97 ** 1.85 ± 0.99* 1.82 ± 0.95* 1.70 ± 0.99*  Year of surgeryc 7.31 ± 4.07 7.7 ± 4.13* 7.81 ± 4.10* 8.58 ± 4.10* 9.04 ± 3.99* * P-value <0.01. ** P-value = 0.01–0.05. *** P-value = 0.05–0.10. a Reflect percentage of the overall study population (%). b Variables were not used in multivariable risk adjustment as they are highly correlated with other model variables. c Year of surgery: 1 = 1994, 2 = 1996, … , 19 = 2012. BMI: body mass index; BSA: body surface area; CABG: coronary artery bypass grafting; SD: standard deviation. A 2-sided P-value <0.05 was used uniformly to indicate significance. Statistical analysis was conducted using the IBM SPSS Statistics for Windows, Version 23.0 (Armonk, NY: IBM Corp. and StataCorp. 2017) and the Stata Statistical Software: Release 15 (College Station, TX: StataCorp LLC) for competing risk analysis. RESULTS The study population consisted of 7091 patients including 1312 patients in the Normal group (18.6%), 2786 patients in the Overweight group (39.3%), 1866 patients in the Obese I group (26.3%), 729 patients in the Obese II group (10.3%) and 398 patients in the Obese III group (5.6%). Prevalence of diabetes, hypertension and heart failure [Normal (10.9%) vs Obese III (17.6%)] was progressively higher with increasing BMI despite age being systematically lower [Normal (67.1 ± 10.8) vs Obese III (58.9 ± 9.3); years] (Table 1). Multiarterial grafts were frequently used in this series, and the rate increased with BMI given their younger age [Normal (29.5%) vs Obese III (57.3%)]. Mean follow-up was 9.5 ± 4.6 years. A total of 4063 (57.3%) deaths were documented overall, of which CV deaths accounted for 20.3%, non-CV deaths for 28.4% and 8.6% of deaths were of unknown cause. Unadjusted 30-day mortality was 1.38% and was similar across BMI categories [Normal (1.68%), Overweight (1.29%), Obese I (1.23%), Obese II (1.23%) and Obese III (1.37%)]. Overall and cause-specific mortality cumulative incidence functions Figure 1 shows the cumulative incidence functions for all-cause mortality partitioned to its CV, non-CV and unknown cause components for the entire population and then for the Normal, Overweight, Obese I, Obese II and Obese III subgroups. Compared to Normal (61.0%), the estimated 15-year unadjusted all-cause mortality was significantly lower for the Overweight [(56.6%), P < 0.001], Obese I [(54.7%), P < 0.001] and Obese II [(57.2%), P = 0.007] groups, whereas it was similar for the Obese III [(62.8%), P = 0.91] group. Figure 1: View largeDownload slide Fifteen-year cumulative incidence functions showing all-cause mortality and cause-specific mortality for all patients (n = 7091) and for all the 5 body habitus sub-cohorts. CABG: coronary artery bypass grafting; CV: cardiovascular; non-CV: non-cardiovascular; Unk: unknown cause of death. Figure 1: View largeDownload slide Fifteen-year cumulative incidence functions showing all-cause mortality and cause-specific mortality for all patients (n = 7091) and for all the 5 body habitus sub-cohorts. CABG: coronary artery bypass grafting; CV: cardiovascular; non-CV: non-cardiovascular; Unk: unknown cause of death. The unadjusted and comprehensively covariate-adjusted hazard ratios across BMI categories are summarized in Fig. 2—with ‘Normal’ designated as the reference group and accounting for competing risks in case of cause-specific (CV and non-CV) mortality calculations. Figure 2 also shows results assuming proportional hazard over the entire 15-year follow-up (overall) and after relaxing this assumption by time segmenting the analysis to Early mortality, Intermediate mortality and Late mortality. Generally, the unadjusted all-cause, CV and non-CV death data showed a significant and appreciable partial obesity paradox [Fig. 2 (left panels)] whereby Overweight and Obese I and Obese II groups showed superior outcomes (mortality hazard ratios <1), whereas the Obese III group had unadjusted hazard ratios similar to the Normal group. This paradox was (i) the largest in case of early outcomes, (ii) present but relatively reduced in the intermediate term (1–8 years) and (iii) absent in the late term. Figure 2: View largeDownload slide Comparison of mortality risk stratified by the body habitus groups for all-cause, CV and non-CV deaths shown for entire follow-up [overall (0–15 years) and for time-segmented follow-up (Early: 0–1 year, Intermediate: 1–8 years and Late: 8–15 years)]. Open symbols indicate unadjusted hazard ratios with or without accounting for competing risk of death, closed symbols indicate comprehensive risk-adjusted hazard ratios with or without accounting for competing risk of death and error bars represent 95% CIs. CI: confidence interval; CV: cardiovascular; HR: hazard ratio; Ob: obese; OW: overweight. Figure 2: View largeDownload slide Comparison of mortality risk stratified by the body habitus groups for all-cause, CV and non-CV deaths shown for entire follow-up [overall (0–15 years) and for time-segmented follow-up (Early: 0–1 year, Intermediate: 1–8 years and Late: 8–15 years)]. Open symbols indicate unadjusted hazard ratios with or without accounting for competing risk of death, closed symbols indicate comprehensive risk-adjusted hazard ratios with or without accounting for competing risk of death and error bars represent 95% CIs. CI: confidence interval; CV: cardiovascular; HR: hazard ratio; Ob: obese; OW: overweight. Risk-adjustment modified the mortality outcome—BMI relationships such that the magnitude of the paradoxical (improved) outcomes at higher BMI were generally reduced in magnitude and limited to Overweight and Obese I patients, whereas severely obese patients (Obese III) showed significantly worse outcomes than Normal BMI. The complementary time-segmented and cause-specific analysis suggested that a higher BMI paradox was mostly due to superior intermediate survival. The overall 15-year mortality demonstrated the presence of the J-shaped protective partial ‘obesity paradox’ in the Overweight and Obese I showing lower mortality as evidenced by the following adjusted hazard ratios (AHR): [AHR (95% CI) = 0.88 (0.79–0.98), P = 0.017 and AHR = 0.88 (0.78–0.99), P = 0.032, respectively]. The morbidly obese patients (Obese III) had a higher mortality [AHR = 1.28 (1.06–1.55), P = 0.012]. The Obese II group showed an increased trend for overall mortality, but this did not reach statistical significance [AHR = 1.13 (0.97–1.33)]. The protective nature of the Overweight and Obese I category was driven by decreased CV mortality and non-CV mortality although this trend did not reach significance in the Obese I patients (Fig. 2). The detrimental impact on survival in the Obese III category was driven by increased CV mortality only [AHR = 1.47 (1.08–2.00), P = 0.015] (Table 2; Fig. 2). Table 2: Overall and time-segmented risk-adjusted hazard ratios for all-cause and cause-specific mortality across the BMI groups PH regression, HR (95% CI) Time-segmented analysis, HR (95% CI) Overall (0–15 years) Early (0–1 year) Intermediate (1–8 years) Late (8–15 years) All deaths  N (ref) 1 1 1 1  Overweight 0.88 (0.79–0.98) 0.75 (0.56–1.00) 0.82 (0.71–0.96) 0.99 (0.83–1.18)  Obese I 0.88 (0.78–0.99) 0.84 (0.6–1.19) 0.80 (0.67–0.95) 0.97 (0.80–1.18)  Obese II 1.13 (0.97–1.33) 0.74 (0.44–1.22) 1.16 (0.94–1.44) 1.21 (0.93–1.57)  Obese III 1.28 (1.06–1.55) 1.91 (1.19–3.06) 1.24 (0.95–1.62) 1.08 (0.77–1.52) CV deaths  N (ref) 1 1 1 1  Overweight 0.83 (0.70–1.00) 0.76 (0.53–1.09) 0.61 (0.46–0.81) 1.19 (0.87–1.62)  Obese I 0.90 (0.73–1.10) 0.87 (0.57–1.34) 0.73 (0.53–1.00) 1.09 (0.77–1.55)  Obese II 1.11 (0.85–1.45) 0.99 (0.56–1.76) 0.95 (0.64–1.42) 1.43 (0.90–2.27)  Obese III 1.47 (1.08–2.00) 2.31 (1.31–4.07) 1.20 (0.77–1.86) 1.20 (0.64–2.24) Non-CV deaths  N (ref) 1 1 1 1  Overweight 0.86 (0.74–1.00) 0.57 (0.31–1.07) 0.85 (0.69–1.04) 0.82 (0.64–1.04)  Obese I 0.85 (0.72–1.01) 0.60 (0.30–1.23) 0.77 (0.61–0.98) 0.91 (0.69–1.19)  Obese II 1.02 (0.81–1.28) 0.24 (0.05–1.08) 1.22 (0.91–1.64) 0.90 (0.62–1.30)  Obese III 0.88 (0.65–1.18) 1.01 (0.34–3.04) 1.07 (0.72–1.58) 0.74 (0.45–1.22) PH regression, HR (95% CI) Time-segmented analysis, HR (95% CI) Overall (0–15 years) Early (0–1 year) Intermediate (1–8 years) Late (8–15 years) All deaths  N (ref) 1 1 1 1  Overweight 0.88 (0.79–0.98) 0.75 (0.56–1.00) 0.82 (0.71–0.96) 0.99 (0.83–1.18)  Obese I 0.88 (0.78–0.99) 0.84 (0.6–1.19) 0.80 (0.67–0.95) 0.97 (0.80–1.18)  Obese II 1.13 (0.97–1.33) 0.74 (0.44–1.22) 1.16 (0.94–1.44) 1.21 (0.93–1.57)  Obese III 1.28 (1.06–1.55) 1.91 (1.19–3.06) 1.24 (0.95–1.62) 1.08 (0.77–1.52) CV deaths  N (ref) 1 1 1 1  Overweight 0.83 (0.70–1.00) 0.76 (0.53–1.09) 0.61 (0.46–0.81) 1.19 (0.87–1.62)  Obese I 0.90 (0.73–1.10) 0.87 (0.57–1.34) 0.73 (0.53–1.00) 1.09 (0.77–1.55)  Obese II 1.11 (0.85–1.45) 0.99 (0.56–1.76) 0.95 (0.64–1.42) 1.43 (0.90–2.27)  Obese III 1.47 (1.08–2.00) 2.31 (1.31–4.07) 1.20 (0.77–1.86) 1.20 (0.64–2.24) Non-CV deaths  N (ref) 1 1 1 1  Overweight 0.86 (0.74–1.00) 0.57 (0.31–1.07) 0.85 (0.69–1.04) 0.82 (0.64–1.04)  Obese I 0.85 (0.72–1.01) 0.60 (0.30–1.23) 0.77 (0.61–0.98) 0.91 (0.69–1.19)  Obese II 1.02 (0.81–1.28) 0.24 (0.05–1.08) 1.22 (0.91–1.64) 0.90 (0.62–1.30)  Obese III 0.88 (0.65–1.18) 1.01 (0.34–3.04) 1.07 (0.72–1.58) 0.74 (0.45–1.22) CI: confidence interval; CV: cardiovascular; HR: hazard ratio; PH: proportional hazard. Bold values indicate significance at P-value = 0.05. Table 2: Overall and time-segmented risk-adjusted hazard ratios for all-cause and cause-specific mortality across the BMI groups PH regression, HR (95% CI) Time-segmented analysis, HR (95% CI) Overall (0–15 years) Early (0–1 year) Intermediate (1–8 years) Late (8–15 years) All deaths  N (ref) 1 1 1 1  Overweight 0.88 (0.79–0.98) 0.75 (0.56–1.00) 0.82 (0.71–0.96) 0.99 (0.83–1.18)  Obese I 0.88 (0.78–0.99) 0.84 (0.6–1.19) 0.80 (0.67–0.95) 0.97 (0.80–1.18)  Obese II 1.13 (0.97–1.33) 0.74 (0.44–1.22) 1.16 (0.94–1.44) 1.21 (0.93–1.57)  Obese III 1.28 (1.06–1.55) 1.91 (1.19–3.06) 1.24 (0.95–1.62) 1.08 (0.77–1.52) CV deaths  N (ref) 1 1 1 1  Overweight 0.83 (0.70–1.00) 0.76 (0.53–1.09) 0.61 (0.46–0.81) 1.19 (0.87–1.62)  Obese I 0.90 (0.73–1.10) 0.87 (0.57–1.34) 0.73 (0.53–1.00) 1.09 (0.77–1.55)  Obese II 1.11 (0.85–1.45) 0.99 (0.56–1.76) 0.95 (0.64–1.42) 1.43 (0.90–2.27)  Obese III 1.47 (1.08–2.00) 2.31 (1.31–4.07) 1.20 (0.77–1.86) 1.20 (0.64–2.24) Non-CV deaths  N (ref) 1 1 1 1  Overweight 0.86 (0.74–1.00) 0.57 (0.31–1.07) 0.85 (0.69–1.04) 0.82 (0.64–1.04)  Obese I 0.85 (0.72–1.01) 0.60 (0.30–1.23) 0.77 (0.61–0.98) 0.91 (0.69–1.19)  Obese II 1.02 (0.81–1.28) 0.24 (0.05–1.08) 1.22 (0.91–1.64) 0.90 (0.62–1.30)  Obese III 0.88 (0.65–1.18) 1.01 (0.34–3.04) 1.07 (0.72–1.58) 0.74 (0.45–1.22) PH regression, HR (95% CI) Time-segmented analysis, HR (95% CI) Overall (0–15 years) Early (0–1 year) Intermediate (1–8 years) Late (8–15 years) All deaths  N (ref) 1 1 1 1  Overweight 0.88 (0.79–0.98) 0.75 (0.56–1.00) 0.82 (0.71–0.96) 0.99 (0.83–1.18)  Obese I 0.88 (0.78–0.99) 0.84 (0.6–1.19) 0.80 (0.67–0.95) 0.97 (0.80–1.18)  Obese II 1.13 (0.97–1.33) 0.74 (0.44–1.22) 1.16 (0.94–1.44) 1.21 (0.93–1.57)  Obese III 1.28 (1.06–1.55) 1.91 (1.19–3.06) 1.24 (0.95–1.62) 1.08 (0.77–1.52) CV deaths  N (ref) 1 1 1 1  Overweight 0.83 (0.70–1.00) 0.76 (0.53–1.09) 0.61 (0.46–0.81) 1.19 (0.87–1.62)  Obese I 0.90 (0.73–1.10) 0.87 (0.57–1.34) 0.73 (0.53–1.00) 1.09 (0.77–1.55)  Obese II 1.11 (0.85–1.45) 0.99 (0.56–1.76) 0.95 (0.64–1.42) 1.43 (0.90–2.27)  Obese III 1.47 (1.08–2.00) 2.31 (1.31–4.07) 1.20 (0.77–1.86) 1.20 (0.64–2.24) Non-CV deaths  N (ref) 1 1 1 1  Overweight 0.86 (0.74–1.00) 0.57 (0.31–1.07) 0.85 (0.69–1.04) 0.82 (0.64–1.04)  Obese I 0.85 (0.72–1.01) 0.60 (0.30–1.23) 0.77 (0.61–0.98) 0.91 (0.69–1.19)  Obese II 1.02 (0.81–1.28) 0.24 (0.05–1.08) 1.22 (0.91–1.64) 0.90 (0.62–1.30)  Obese III 0.88 (0.65–1.18) 1.01 (0.34–3.04) 1.07 (0.72–1.58) 0.74 (0.45–1.22) CI: confidence interval; CV: cardiovascular; HR: hazard ratio; PH: proportional hazard. Bold values indicate significance at P-value = 0.05. Comprehensive multivariable Cox regression model analyses using BMI as a continuous non-linear variable corroborated the 5 BMI category analyses (Fig. 3). A cubic function was found to be the best fit for the overall follow-up analysis; however, a quadratic function was the best fit for some of the time-segmented and cause-specific death analyses. The late follow-up time period showed no significant correlation with BMI as continuous function regardless of the function used (linear, quadratic and cubic). In the overall (0–15 years), early (0–1 year) and intermediate (1–8 years) analyses, there was evidence of the consistent obesity paradox with the adjusted hazard ratios at slightly above normal BMI [25–30], being less than 1. Also, with the exception of early non-CV mortality, it appears that a higher BMI (>37 kg/m2) is associated with worse outcome (Fig. 3). Based on the continuous model, a BMI between 28.5 and 29 kg/m2 is associated with the best outcomes, mostly due to less early mortality and intermediate-term mortality. Figure 3: View largeDownload slide Association between BMI and mortality (all-cause, CV and non-CV) across the BMI spectrum (continuous variable with cubic outcome association) for overall and time-segmented follow-up periods following multivariable competing risk regression. Shaded area represents 95% CI. Symbol with error bars represent HR with 95% CIs for BMI when categorized into the body habitus groups (same as risk-adjusted data in Fig. 2). Dagger represents results of model fit with a forced BMI cubic function (a1*BMI ± a2*BMI2 ± a3*BMI3), where the model coefficient a3 (for BMI3) was not significant, whereas a2 (for BMI2) was significant and double dagger indicates that the BMI outcome association is not significant for any order. BMI: body mass index; CI: confidence interval; CV: cardiovascular; HR: hazard ratio. Figure 3: View largeDownload slide Association between BMI and mortality (all-cause, CV and non-CV) across the BMI spectrum (continuous variable with cubic outcome association) for overall and time-segmented follow-up periods following multivariable competing risk regression. Shaded area represents 95% CI. Symbol with error bars represent HR with 95% CIs for BMI when categorized into the body habitus groups (same as risk-adjusted data in Fig. 2). Dagger represents results of model fit with a forced BMI cubic function (a1*BMI ± a2*BMI2 ± a3*BMI3), where the model coefficient a3 (for BMI3) was not significant, whereas a2 (for BMI2) was significant and double dagger indicates that the BMI outcome association is not significant for any order. BMI: body mass index; CI: confidence interval; CV: cardiovascular; HR: hazard ratio. DISCUSSION Our study supports the existence of a long-term (15 years) protective partial obesity paradox in CABG patients as evidenced by improved survival in the Overweight and Obese I cohorts, which appeared in the Early postoperative period in Overweight patients and in the Intermediate postoperative period in both the Overweight and Obese I patients. Using the competing risk analysis, the obesity paradox was associated with both CV and non-CV mortality risk reduction during the entire 0–15-year follow-up period in Overweight patients and during the Intermediate postoperative time period in Obese I patients. The morbidly obese patients (Obese III) were found to have increased mortality risk primarily due to worse CV survival in the early postoperative time interval. These results agree with previous reports documenting an obesity paradox after CABG in higher BMI patients [15, 16, 18, 19, 28] and worse survival in the morbidly obese patients [13, 18]. Stamou et al. [15] identified an obesity paradox among cardiac surgery patients noting a decreased short-term mortality and intermediate-term mortality among overweight patients. van Straten et al. [18] noted a long-term survival advantage in overweight CABG patients. In a Veterans Administration CABG population, Wagner et al. [28] noted optimal acute perioperative survival in patients with a BMI of 30 kg/m2. The identified partial obesity paradox characterized by a ‘J’-shaped relationship between our 5 BMI groups and long-term survival conflicts with other reports [12, 14, 17, 21] that found no such protective effect of increased BMI on survival. The discrepant findings between those studies and ours could be due to multiple factors including small sample sizes and short follow-up periods in the other studies [17], as well as using different BMI categorization [14] and not taking into account the continuous but non-linear relationship between outcome and BMI. Gurm et al. [12] analysed the impact of obesity on acute mortality and 5-year mortality in patients undergoing CABG and percutaneous intervention in the Bypass Angioplasty Revascularization Investigation (BARI) trial. In contrast to our observation, they identified that among overweight and mildly obese patients, CV mortality risk was increased, whereas all-cause mortality was unaffected by increasing BMI, suggesting, in fact, that non-CV mortality was reduced. Importantly, the BARI cohort did not reflect contemporary treatment modalities (low LITA use rate of 75–86%) as it included patients from the 1990s, and operative deaths were not included in the analysis. Le-Bert et al. [17], despite claims of identifying an obesity paradox in a small cohort of elderly CABG patients, did not find an association between BMI and acute in-hospital mortality. Prabhakar et al. [21], using the Society of Thoracic Surgeons Database-and relying only on a small number of covariates for a logistic regression analysis, found an increased perioperative mortality risk among mildly obese patients [HR = 1.21 (1.45–1.73)]. Long-term mortality was not analysed, and their patient population was characterized by low LITA utilization. As our study population included only LITA-based CABG with an appreciable rate of multiarterial grafting (41.6%), the 2 study populations are not comparable. Importantly, our outcomes were adjusted for the number of arterial grafts and the revascularization index, both of which were generally higher with rising BMI. Not doing so would have further exaggerated the protective effects of the obesity paradox. Benedetto et al. [14] found no effect of obesity on acute perioperative mortality, an observation that agrees with our findings. In propensity-matched groups, an overweight BMI was not protective against long-term mortality [HR = 1.05 (0.90–1.08)], and in contrast to our study, mildly obese patients (our Obese I cohort) had an increased risk of mortality [HR = 1.22 (1.07–2.66)] during a mean follow-up period of 7.6 years. The covariates in the propensity matching algorithm used by Benedetto et al. were limited to those included in the EuroSCORE and did not include specific grafting strategies (single arterial or multiarterial) or the revascularization index. They were unable to corroborate their outcomes in a subsequent Cox regression analysis, noting that overweight patients had decreased risk of long-term mortality [HR = 0.88 (0.80–0.97)], whereas mild obesity and morbid obesity did not influence long-term survival. They only all-cause peri-operative and long-term mortality without details on specific causes of death. The physiological basis of the obesity paradox remains undefined and is beyond the scope of this study, but a number of possible hypotheses have been proposed. Among these are the development of earlier but less complex coronary artery disease [24] (less left main coronary disease with increasing BMI in our series), altered cytokine and neuroendocrine profiles among obese patients with a favourable CV profile, increased adipose-tissue-derived soluble tumour necrosis factor (TNF) receptor production with a resultant decrease in the pathogenic effects of TNF, lower levels of atrial naturetic peptides, attenuated sympathetic and renin–angiotensin activity, decreased systemic vascular resistance and enhanced metabolic reserves that maybe beneficial during stress-induced catabolic states such as the postoperative period [23]. The fact that weight-loss programmes have been associated with increased adverse outcomes maybe considered as indirect evidence for the existence of an obesity paradox [29]. Our findings of increased mortality in the morbidly obese (Obese III) CABG patients confirms those of multiple other reports [12–14, 18, 21, 30], and this was principally driven by increased CV mortality within the first postoperative year. The pathophysiology behind this observation is also undefined and beyond the scope of this study but is unlikely related to postoperative complications, as the acute perioperative outcomes were equivalent and uniformly excellent across all BMI categories. When survival was analysed as a function of BMI as a continuous variable, we identified a BMI of 29 kg/m2 corresponding to the lowest overall mortality rate (Fig. 3). In such an analysis, rather than a ‘J’-shaped relationship between survival and the 5 distinct BMI groups, the survival curves showed a U-shaped relationship between BMI and mortality with increased risk at both extremes, a finding that we have reported on earlier [13]. Benedetto et al. [14] similarly identified a BMI range of 27–29 kg/m2 as being maximally protective for late mortality. Wagner et al. [28] noted a similar U-shaped relationship between acute perioperative mortality and BMI with a nadir mortality point at a BMI of 30 kg/m2. A question that remains unanswered is whether delaying surgery to affect weight reduction may ameliorate the increased mortality risk among the morbidly obese patients. Limitations Our study has a number of limitations. As this is a retrospective analysis, only associations can be established, and we are unable to assess the specific physiological mechanisms that may account for our findings. All observational studies are subject to well-recognized biases including patient selection bias, and conclusions might be altered by exclusion of covariates that may be pertinent to the analysis but, nevertheless, are not included in the study data set. In addition, although BMI is a convenient and widely used parameter to assess obesity, other parameters of obesity such as adiposity, body fatness, waist circumference and weight-to-height ratio, all of which have been shown to impact mortality [23], were not used in our analysis, and it is not known how, or if at all, the inclusion of these measures of obesity may have impacted our conclusions. Furthermore, it is important to note that body mass index is not a static patient characteristic and is subject to change over time. Any changes in BMI over the study period were unavailable to us, and only the values at the index operation were utilized in the analysis. Also, we are unable to assess the patients’ long-term medication use and risk reduction strategies on long-term survival. The impact of any repeat revascularization strategies on long-term survival was not included in our study. Finally, BMI does not provide a distinction between all obese patients and metabolically healthy obese patients who may have a distinctly different prognosis for any given BMI [23]. CONCLUSION In conclusion, our study demonstrates the presence of a partial obesity paradox up to 15 years postoperatively in CABG patients. Patients who are overweight and mildly obese show improved survival, with a decreased risk of CV mortality and non-CV mortality in the entire 0–15-year postoperative period and in the 1–8-year postoperative period, respectively. Conversely, morbidly obese patients had significantly worse survival driven by increased CV mortality within a year of the index operation. Conflict of interest: none declared. REFERENCES 1 Flegal KM , Carroll MD , Ogden CL , Johnson CL. Prevalence and trends in obesity among US adults, 1999-2000 . JAMA 2002 ; 288 : 1723 – 7 . Google Scholar CrossRef Search ADS PubMed 2 Ogden CL , Carroll MD , Kit BK , Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011-2012 . JAMA 2014 ; 311 : 806 – 14 . Google Scholar CrossRef Search ADS PubMed 3 Manson JE , Bassuk SS. Obesity in the United States: a fresh look at its high toll . JAMA 2003 ; 289 : 229 – 30 . Google Scholar CrossRef Search ADS PubMed 4 Lavie CJ , Milani RV , Ventura HO. Obesity and cardiovascular disease: risk factor, paradox, and impact of weight loss . J Am Coll Cardiol 2009 ; 53 : 1925 – 32 . Google Scholar CrossRef Search ADS PubMed 5 5, Zacharias A , Schwann TA , Riordan CJ , Durham SJ , Shah AS , Habib RH. Obesity and risk of new-onset atrial fibrillation after cardiac surgery . Circulation 2005 ; 112 : 3247 – 55 . Google Scholar CrossRef Search ADS PubMed 6 National Institutes of Diabetes and Digestive and Kidney Diseases. Statistics Related to Overweight and Obesity. NIH Publ No.96-4158. Update 2000 . 7 Rosenbaum M , Leibel RL , Hirsch J. Obesity . N Engl J Med 1997 ; 337 : 396 – 407 . Google Scholar CrossRef Search ADS PubMed 8 Poirier P , Giles TD , Bray GA , Hong Y , Stern JS , Pi SFX. Obesity and cardiovascular disease: pathophysiology, evaluation, and effect of weight loss: an update of the 1997 American Heart Association Scientific Statement on Obesity and Heart Disease from the Obesity Committee of the Council on Nutrition, Physical Activity, and Metabolism . Circulation 2006 ; 113 : 898 – 918 . Google Scholar CrossRef Search ADS PubMed 9 Park J , Ahmadi S-F , Streja E , Molnar MZ , Flegal KM , Gillen D et al. . Obesity paradox in end-stage kidney disease patients . Prog Cardiovasc Dis 2014 ; 56 : 415 – 25 . Google Scholar CrossRef Search ADS PubMed 10 Chittal P , Babu AS , Lavie CJ. Obesity paradox: does fat alter outcomes in chronic obstructive pulmonary disease? COPD 2015 ; 12 : 14 – 18 . Google Scholar CrossRef Search ADS PubMed 11 Galal W , van Gestel YRBM , Hoeks SE , Sin DD , Winkel TA , Bax JJ et al. . The obesity paradox in patients with peripheral arterial disease: the influence of chronic obstructive pulmonary disease . Chest 2008 ; 134 : 925 – 30 . Google Scholar CrossRef Search ADS PubMed 12 Gurm HS , Whitlow PL , Kip KE ; BARI Investigators . The impact of body mass index on short and long-term outcomes in patients undergoing coronary revascularization . J Am Coll Cardiol 2002 ; 39 : 834 – 40 . Google Scholar CrossRef Search ADS PubMed 13 Habib RH , Zacharias A , Schwann TA , Riordan CJ , Durham SJ , Shah A. Effects of obesity and small body size on operative and long-term outcomes of coronary artery bypass surgery: a propensity-matched analysis . Ann Thorac Surg 2005 ; 79 : 1976 – 86 . Google Scholar CrossRef Search ADS PubMed 14 Benedetto U , Danese C , Codispoti M. Obesity paradox in coronary artery bypass grafting: myth or reality? J Thorac Cardiovasc Surg 2014 ; 147 : 1517 – 23 . Google Scholar CrossRef Search ADS PubMed 15 Stamou SC , Nussbaum M , Stiegel RM , Reames MK , Skipper ER , Robicsek F et al. . Effect of body mass index on outcomes after cardiac surgery: is there an obesity paradox? Ann Thorac Surg 2011 ; 91 : 42 – 7 . Google Scholar CrossRef Search ADS PubMed 16 Sung S-H , Wu T-C , Huang C-H , Lin S-J , Chen J-W. Prognostic impact of body mass index in patients undergoing coronary artery bypass surgery . Heart 2011 ; 97 : 648 – 54 . Google Scholar CrossRef Search ADS PubMed 17 Le-Bert G , Santana O , Pineda AM , Zamora C , Lamas GA , Lamelas J. The obesity paradox in elderly obese patients undergoing coronary artery bypass surgery . Interact CardioVasc Thorac Surg 2011 ; 13 : 124 – 7 . Google Scholar CrossRef Search ADS PubMed 18 van Straten AHM , Bramer S , Soliman Hamad MA , van Zundert AAJ , Martens EJ , Schönberger JPAM et al. . Effect of body mass index on early and late mortality after coronary artery bypass grafting . Ann Thorac Surg 2010 ; 89 : 30 – 7 . Google Scholar CrossRef Search ADS PubMed 19 Gruberg L , Mercado N , Milo S , Boersma E , Disco C , van Es G-A et al. . Impact of body mass index on the outcome of patients with multivessel disease randomized to either coronary artery bypass grafting or stenting in the ARTS trial: the obesity paradox II? Am J Cardiol 2005 ; 95 : 439 – 44 . Google Scholar CrossRef Search ADS PubMed 20 Alam M , Siddiqui S , Lee V-V , Elayda MA , Nambi V , Yang EY et al. . Isolated coronary artery bypass grafting in obese individuals: a propensity matched analysis of outcomes . Circ J 2011 ; 75 : 1378 – 85 . Google Scholar CrossRef Search ADS PubMed 21 Prabhakar G , Haan CK , Peterson ED , Coombs LP , Cruzzavala JL , Murray GF. The risks of moderate and extreme obesity for coronary artery bypass grafting outcomes: a study from the Society of Thoracic Surgeons Database . Ann Thorac Surg 2002 ; 74 : 1125 – 31 . Google Scholar CrossRef Search ADS PubMed 22 Hartrumpf M , Kuehnel RU , Albes JM. The obesity paradox is still there: a risk analysis of over 15 000 cardiosurgical patients based on body mass index . Interact CardioVasc Thorac Surg 2017 ; 25 : 18 – 24 . Google Scholar CrossRef Search ADS PubMed 23 Lavie CJ , McAuley PA , Church TS , Milani RV , Blair SN. Obesity and cardiovascular diseases: implications regarding fitness, fatness, and severity in the obesity paradox . J Am Coll Cardiol 2014 ; 63 : 1345 – 54 . Google Scholar CrossRef Search ADS PubMed 24 Sturm R , Wells KB. Does obesity contribute as much to morbidity as poverty or smoking? Public Health 2001 ; 115 : 229 – 35 . Google Scholar CrossRef Search ADS PubMed 25 Lavie CJ , Messerli FH. Cardiovascular adaption to obesity and hypertension . Chest 1986 ; 90 : 275 – 9 . Google Scholar CrossRef Search ADS PubMed 26 Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser., 2000 ; 894: i–xii, 1–253. 27 Schwann TA , Habib RH , Zacharias A , Parenteau GL , Riordan CJ , Durham SJ et al. . Effects of body size on operative, intermediate, and long-term outcomes after coronary artery bypass operation . Ann Thorac Surg 2001 ; 71 : 521 – 30 . Google Scholar CrossRef Search ADS PubMed 28 Wagner BD , Grunwald GK , Rumsfeld JS , Hill JO , Ho PM , Wyatt HR et al. . Relationship of body mass index with outcomes after coronary artery bypass graft surgery . Ann Thorac Surg 2007 ; 84 : 10 – 16 . Google Scholar CrossRef Search ADS PubMed 29 Fonarow GC , Horwich TB , Hamilton MA , MacLellan WR , Tillisch JH. Obesity, weight reduction, and survival in heart failure: reply . J Am Coll Cardiol 2002 ; 39 : 1563 – 4 . Google Scholar CrossRef Search ADS PubMed 30 Sun X , Hill PC , Bafi AS , Garcia JM , Haile E , Corso PJ et al. . Is cardiac surgery safe in extremely obese patients (body mass index 50 or greater)? Ann Thorac Surg 2009 ; 87 : 540 – 7 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

European Journal of Cardio-Thoracic SurgeryOxford University Press

Published: Jun 2, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off