- Email: [email protected]
Associations of body weight and weight change with cardiovascular events and mortality in patients with coronary heart disease
Atherosclerosis 274 (2018) 104e111
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Associations of body weight and weight change with cardiovascular events and mortality in patients with coronary heart disease Sheng-Yong Dong a, *, 1, Shuang-Tong Yan b, 1, Man-Liu Wang c, Zhi-Bing Li b, Lian-Qing Fang a, Qiang Zeng d, ** a
Healthcare Department, Agency for Offices Administration of PLA, Beijing, 100034, China Department of Geriatric Endocrinology, Chinese PLA General Hospital, Beijing, 100853, China Peking University-Tsinghua University Joint Center for Life Sciences, Beijing, China, Advanced Academy of Interdisciplinary Sciences, Peking University, Beijing, 100871, China d Health Management Institute, Chinese PLA General Hospital, Beijing, 100853, China b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 November 2017 Received in revised form 31 March 2018 Accepted 2 May 2018 Available online 5 May 2018
Background and aims: It is recommended that patients with coronary heart disease (CHD) pursue a normal body weight, while the effects of body weight and weight change on prognosis are still controversial. The present study was to assess these effects using a large-scale population with CHD in China. Methods: A total of 5276 patients with CHD were included from Jan 2000 to Dec 2014. Baseline and endpoint weights were measured. Outcomes including mortality and cardiovascular events were obtained. Results: Relative to patients with normal weight, risks for adverse outcomes were lowest in overweight patients and similar in obese patients. Hazard ratios (HRs) and 95% confidence interval (95% CI) for allcause death were 1.42 (1.06, 1.91) if overweight turned into normal weight and were 2.01 (1.28, 3.16) or 5.33 (2.81, 10.1) if obese turned into overweight or normal weight. Death risk increased with the extent of weight loss and moderate or large weight gain (p<0.05 for all). Similar results were found when risks for cardiovascular mortality and events were considered. Furthermore, these results remained significant when the patients were stratified by several covariates and even when several definitions of weight change were considered. Conclusions: Obesity did not increase adverse outcome risks in patients with CHD. Both weight loss and weight gain increased adverse outcome risks regardless of baseline body weight. The findings suggest that maintaining a stable weight may be a better strategy for the reduction of risks for cardiovascular outcomes and all-cause death in patients with CHD. © 2018 Elsevier B.V. All rights reserved.
Keywords: Coronary heart disease Body weight Obesity Weight change Mortality Cardiovascular event
1. Introduction Although cardiovascular disease (CVD) mortality declined dramatically over the past 2 decades in all high-income and some middle-income countries, CVDs remain the leading causes of premature death and chronic disability in the world [1,2]. The majority
* Corresponding author. Healthcare Department, Agency for Offices Administration of PLA, 2 Aimin Street, Xicheng District, Beijing, 100034, China. ** Corresponding author. Health Management Institute, Chinese PLA General Hospital, 28 Fuxing Road, Beijing, 100853, China. E-mail addresses: [email protected] (S.-Y. Dong), [email protected] (Q. Zeng). 1 These authors contributed equally to this work. https://doi.org/10.1016/j.atherosclerosis.2018.05.007 0021-9150/© 2018 Elsevier B.V. All rights reserved.
of health lost to CVDs are attributable to coronary heart disease (CHD) and stroke [1]. It is concerning that trends in CHD mortality are no longer declining for many world regions, which may be due to rising rates of cardiovascular risk factors particularly obesity [1,3e5]. Previous studies reported that obesity increased risks for CHD [6,7], which was thought to result in an increase of adverse outcomes [8]. Based on these findings, guidelines recommend and the literature encourage patients with CHD to pursue a normal body weight, especially in overweight or obese patients [9e12]. However, in recent years, increasing efforts have been made to assess the potential effects of obesity and weight change on a variety of health outcomes, while the findings are controversial [13e17].
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
Several studies have shown that overweight and obesity are related to better prognosis in patients with established CHD [18,19]. Other studies have shown that loss of weight in overweight or obese patients with CHD does not result in lower mortality and may even have worse prognosis [16,20]. These controversial findings raise questions regarding the effect of obesity on prognosis and the value of weight loss in patients with CHD. Furthermore, many of above studies had relatively small number of participants [16,17,19], and there are currently no large-scale studies involving the Asian population. Finally, data on analyses of associations between changes in body weight categories (i.e., body weight changes from normal weight to overweight) and prognosis in patients with CHD are limited. Therefore, we investigated associations of body weight and weight change with outcomes using a large-scale population with CHD in China. We specifically explored how risks for adverse outcomes change with changes in body weight categories, and also decomposed changes in risks of weight change for deaths and cardiovascular events into the contributions of sex, age, and all other factors. 2. Patients and methods 2.1. Patients The present study recruited the patients with established CHD who visited the Health Department of Agency for Offices Administration of PLA or the Chinese PLA General Hospital from Jan 2000 to Dec 2014. Eligible patients were men and women, 18e90 years of age, with clinically evident CHD, defined by one or more of the following: previous myocardial infarction (MI), the presence of angina pectoris with angiographic evidence of >50% stenosis in at least 1 major coronary artery, previous percutaneous coronary intervention (PCI), and a history of coronary artery bypass grafting (CABG). The exclusion criteria included a history of lung disease, kidney disease, liver disease, heart failure, mental illness, or cancer, or inability to participate in the study. After the exclusion of 104 patients who were lost to follow-up and 130 patients with missing data on covariates, complete data were obtained in 5276 patients. All known diseases were obtained from hospital medical records and confirmed by a medical record review. Written informed consent was obtained from all of the included patients. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Medical Ethics Committee of Chinese PLA General Hospital. The study population flow is explained in Supplemental Fig. 1. 2.2. Body weight and weight change Baseline and endpoint weights were obtained from hospital medical records. Self-reported endpoint weights through telephone visit were used in 185 participants because body weights at the endpoint were not able to be obtained from their hospital medical records. Baseline characteristics were comparable between these participants and the participants whose endpoint weights were obtained from hospital medical records. Body mass index (BMI) was calculated in kg/m2 and was categorized into: underweight, BMI <18.5 kg/m2; normal weight, 18.5e23.9 kg/m2; overweight, 24e27.9 kg/m2; and obese, 28 kg/m2, according to the recommendation by the Working Group on Obesity in China [21]. Weight change was defined by absolute weight change (kg). Absolute weight change was calculated by subtracting weight at baseline from weight at endpoint and were categorized into: weight stability (1 kg loss or gain), small weight gain (1e5 kg gain), moderate weight gain (5e10 kg gain), large weight gain
105
(>10 kg gain), small weight loss (1e5 kg loss), moderate weight loss (5e10 kg loss), and large weight loss (>10 kg loss). 2.3. Covariates Data on smoking status and family medical history were conducted via face-to-face interviews. Individual medical history and medication use were obtained from hospital medical records. Fasting glucose, serum lipids, serum creatinine, high-sensitivity Creactive protein (CRP), and N-terminal pro-B-type natriuretic peptide (NT-proBNP) were measured by routine laboratory methods. 2.4. Follow-up and adverse outcomes Participants were followed until Dec 31, 2016. The follow-up protocol included a combination of hospital medical records, telephone contacts with patients or family members, and death certificates. The primary outcomes were all-cause mortality and cardiovascular mortality. Cardiovascular mortality was defined as death attributable to fatal myocardial infarction, sudden cardiac death, or stroke. The second outcomes were cardiovascular events, including myocardial infarction, stroke, and coronary revascularization (i.e., PCI and/or CABG). Patients with more than one event were assigned the highest-ranked event based on the previous list of cardiovascular events. The diagnosis of myocardial infarction was based on symptoms, elevated cardiac marker of myocardial necrosis, and the presence or absence of diagnostic electrocardiogram changes [22]. Sudden cardiac death was diagnosed as unexpected natural death due to cardiac causes within 1 h of onset of acute symptoms. Stroke included ischemic stroke and hemorrhagic stroke, which were diagnosed according to World Health Organization criteria [23]. All deaths were classified using the tenth revision of the International Classification of Disease and confirmed through death certificates using personal identity card number. All events were confirmed by hospital medical records and adjudicated by an event adjudication committee who were blinded to the measurements of body weight and weight change. 2.5. Statistics Continuous variables are expressed as the mean ± standard deviation (SD) or median (interquartile range) and categorical variables as percentages. Descriptive statistics were performed using one-way analysis of variance or chi-square tests. Nonnormally distributed variables were log transformed before analyses. Age- and sex-standardized mortality and cardiovascular events per 1000 person years across BMI levels and weight change levels were calculated, respectively. Multivariate Cox proportional hazards models were used to assess associations of BMI levels and weight change with outcomes. Associations between changes in BMI levels and outcomes were also assessed. Hazard ratio (HR) and 95% confidence interval (CI) were shown. In sensitivity analyses, associations of weight change with allcause mortality were assessed in the patients stratified by baseline covariates. Furthermore, associations of the percentage of change in weight and average weight change per year with outcomes were assessed, respectively. Calculations and categories of the percentage of changing in weight and average weight change per year are described in the Supplemental material. Finally, the levels of baseline blood pressure, serum lipids, and fasting glucose were further adjusted in Cox models. All analyses were conducted using SPSS 18.0 (SPSS Inc., Chicago, Illinois), and a two-sided value of p < 0.05 was considered statistically significant.
106
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
A total of 5276 patients with complete data met the study criteria and were analysed in the present study. The mean baseline weight of the patients was 71.0 ± 11.5 kg. Among all of the patients, 20.4% were obese, 45.5% were overweight, and 1.8% were underweight by the BMI criteria (Table 1). The percentage of patients with hypertension or diabetes increased with BMI levels being increased (p < 0.01 for all). The percentage of patients who received PCI, CABG, or medication use also increased with the increase of BMI levels (p < 0.05 for all).
increased if BMI levels at the endpoint decreased one level or more levels, regardless of BMI levels at baseline (Supplemental Table 2). After adjustment for covariates, risk for all-cause death increased 57% if body weight changed from normal weight to underweight (P ¼ 0.019) (Fig. 1). The HRs (95% CI) for all-cause death were 1.42 (1.06, 1.91) or 3.16 (1.52, 6.55) if body weight changed from overweight to normal weight or underweight, respectively. If body weight changed from obese to overweight, normal weight, or underweight, mortality risk increased to 2.01 (1.28, 3.16), 5.33 (2.81, 10.1), or 7.94 (1.14, 60.8), respectively. Similarly, multivariate analyses shown that risks for cardiovascular mortality and cardiovascular events increased if BMI levels decreased from the baseline to the endpoint.
3.2. BMI levels, changes in BMI levels, and outcomes
3.3. Weight change and outcomes
During 21,251 person-years of follow-up, we identified 540 deaths of all-cause, 216 cardiovascular deaths, and 1116 cardiovascular events (Supplemental Table 1). The age- and sexstandardized rates of all-cause death, cardiovascular death, and cardiovascular events were lowest in overweight patients (23, 9, and 60 per 1000 person years, respectively) and highest in underweight patients (44, 22, and 131 per 1000 person years, respectively). As shown in Table 2, risks for all-cause mortality, cardiovascular mortality, and cardiovascular events reduced 24%, 27%, and 19%, respectively, in overweight patients, while these risks increased 78%, 130%, and 88%, respectively, in underweight patients, relative to patients with normal weight after adjustment for covariates (p < 0.05 for all). Furthermore, obese patients had similar risks for adverse outcomes compared to normal weight patients in multivariate analyses. Death rate and the prevalence of cardiovascular events decreased if BMI levels at the endpoint increased one level but
The mean (SD) of weight change over a mean follow-up of 4 years was a loss of 1.9 kg (6.8 kg). The age- and sex-standardized all-cause death, cardiovascular death, and cardiovascular events per 1000 person years were lowest in participants with weight stability (1 kg loss or gain) and increased with the extent of weight loss or weight gain (Supplemental Fig. 2). Relative to weight stability, the HRs (95% CI) of small, moderate, and larger weight loss for all-cause mortality were 1.66 (1.26, 2.20), 1.85 (1.40, 2.44), and 2.91 (2.15, 3.93), respectively (Table 3). Patients with moderate or large weight gain also had a 48% or 95% increase in risk for all-cause death (p < 0.05). Similar results were found when associations of cardiovascular mortality and cardiovascular events with weight change were considered.
3. Results 3.1. Baseline characteristics
3.4. Sensitivity analyses When participants were stratified by sex, age groups (<60 years,
Table 1 Baseline characteristics.
Age, year Female Current smoking Hypertension Diabetes Stroke Family history of coronary disease Body weight, kg BMI, kg/m2 Systolic BP, mmHg Diastolic BP, mmHg Fasting glucose, mmol/l Total cholesterol, mmol/l LDL cholesterol, mmol/l HDL cholesterol, mmol/l Triglyceride, mmol/la Serum creatinine, mmol/la High-sensitivity CRP, mg/dla NT-proBNP, pg/mla Previous MI Previous PCI Previous CABG Medication use Statin Aspirin ACE inhibitor/ARB b-Blocker
Total (n ¼ 5276)
Underweight (n ¼ 93)
Normal weight (n ¼ 1708)
Overweight (n ¼ 2398)
Obese (n ¼ 1076)
p value
67.6 ± 11.8 1211 (23.0) 1023 (19.4) 3630 (68.8) 1355 (25.7) 868 (16.5) 292 (5.5) 71.0 ± 11.5 25.4 ± 3.4 134 ± 18 77 ± 11 6.21 ± 2.37 4.56 ± 1.09 2.52 ± 0.87 1.18 ± 0.35 1.36 (0.99, 1.85) 82 (69, 96) 0.20 (0.10, 0.83) 234 (80.884) 1374 (26.0) 1876 (35.6) 430 (8.2)
73.7 ± 9.2 26 (28.0) 12 (12.9) 51 (54.8) 17 (18.3) 19 (20.4) 1 (1.1) 48.1 ± 5.6 17.3 ± 1.1 128 ± 23 71 ± 13 5.78 ± 1.80 4.50 ± 1.05 2.42 ± 0.78 1.45 ± 0.49 0.85 (0.66, 0.96) 81 (65, 91) 0.30 (0.10, 1.86) 1293 (297, 2832) 26 (28.0) 21 (22.6) 3 (3.2)
69.5 ± 11.3 447 (26.2) 299 (17.5) 1070 (62.6) 398 (23.3) 285 (16.7) 77 (4.5) 61.4 ± 6.8 22.1 ± 1.4 132 ± 19 74 ± 11 6.11 ± 2.30 4.62 ± 1.16 2.55 ± 0.90 1.27 ± 0.39 1.17 (0.87, 1.59) 80 (67, 95) 0.20 (0.10, 1.00) 293 (101, 1287) 467 (27.3) 562 (32.9) 130 (7.6)
67.4 ± 11.5 486 (20.3) 474 (19.8) 1687 (70.4) 616 (25.7) 394 (16.4) 118 (4.9) 72.7 ± 6.9 25.9 ± 1.1 134 ± 18 76 ± 11 6.22 ± 2.39 4.55 ± 1.10 2.50 ± 0.86 1.15 ± 0.32 1.42 (1.06, 1.91) 82 (70, 96) 0.20 (0.10, 0.70) 213 (74, 836) 621 (25.9) 870 (36.3) 192 (8.0)
64.4 ± 12.7 252 (23.4) 238 (22.1) 822 (76.4) 324 (30.1) 170 (15.8) 96 (8.9) 84.3 ± 9.8 30.2 ± 2.1 135 ± 17 78 ± 10 6.40 ± 2.45 4.50 ± 1.09 2.50 ± 0.84 1.08 ± 0.30 1.63 (1.17, 2.06) 82 (71, 96) 0.21 (0.10, 0.78) 148 (65, 562) 260 (24.2) 423 (39.3) 105 (9.8)
<0.001 0.011 0.001 <0.001 <0.001 0.371 <0.001 <0.001 <0.001 <0.001 <0.001 0.015 0.845 0.494 <0.001 <0.001 0.186 0.396 <0.001 0.058 <0.001 0.018
3189 3944 2767 1816
40 56 37 25
957 (56.0) 1233 (72.1) 829 (48.5) 529 (31.0)
1504 (62.7) 1812 (75.6) 1303 (54.3) 884 (36.9)
688 843 598 378
<0.001 <0.001 <0.001 0.002
(60.4) (74.8) (52.4) (34.4)
(43.0) (60.2) (39.8) (26.9)
(63.9) (78.3) (55.6) (35.1)
Variables are shown as n (%), mean ± standard deviation, or median (quartiles). BMI, body mass index; BP, blood pressure; LDL, low-density lipoprotein; HDL, high-density lipoprotein; CRP, C-reactive protein; NT-proBNP, N-terminal pro-B-type natriuretic peptide; MI, myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass grafting; ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker. a These variables were log transformed before being tested.
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
107
Table 2 Associations between BMI levels and outcomes. Underweight HR (95% CI) All-cause mortality Model 1 1.86 (1.21, Model 2 1.78 (1.15, Cardiovascular mortality Model 1 2.50 (1.32, Model 2 2.30 (1.22, Cardiovascular events Model 1 1.76 (1.23, Model 2 1.88 (1.31,
Normal weight p
Overweight
Obese
HR (95% CI)
p
HR (95% CI)
p
2.86) 2.76)
0.005 0.010
Ref Ref
0.75 (0.62, 0.91) 0.76 (0.62, 0.92)
0.003 0.005
0.91 (0.71, 1.16) 0.90 (0.70, 1.15)
0.455 0.383
4.76) 4.34)
0.005 0.010
Ref Ref
0.74 (0.54, 0.99) 0.73 (0.53, 0.99)
0.045 0.046
1.09 (0.75, 1.58) 1.01 (0.69, 1.47)
0.660 0.968
2.51) 2.70)
0.002 0.001
Ref Ref
0.82 (0.72, 0.94) 0.81 (0.71, 0.93)
0.005 0.003
0.99 (0.84, 1.16) 0.95 (0.81, 1.12)
0.852 0.552
Model 1 was adjusted for age, sex, current smoking, hypertension, diabetes, stroke, family history of coronary disease, previous PCI, previous CABG, and medication use. Model 2 was further adjusted for serum creatinine, high-sensitivity CRP, and NT-proBNP at baseline.
Fig. 1. Risks of changes in BMI levels. For all-cause mortality (A), cardiovascular mortality (B), and cardiovascular events (C). Scales on the x-axis represent hazard ratios and 95% confidence interval adjusted for age, sex, current smoking, hypertension, diabetes, stroke, family history of coronary disease, previous PCI, previous CABG, medication use, serum creatinine, high-sensitivity CRP, and NTproBNP at baseline.
60e74 years, and 75 years), smoking status, hypertension, diabetes, and stroke, respectively, the positive associations of weight
loss or weight gain with risk for all-cause death were unchanged in all subgroups (Supplemental Table 3). These associations were also
Data are shown as hazard ratios and 95% confidence interval. Model 1 was adjusted for age, sex, height, current smoking, hypertension, diabetes, stroke, family history of coronary disease, previous PCI, previous CABG, and medication use. Model 2 was further adjusted for serum creatinine, high-sensitivity CRP, and NT-proBNP at baseline.
1.77 (1.27, 2.89) 1.70 (1.22, 2.77) 1.37 (1.02, 1.70) 1.32 (1.01, 1.64) 1.30 (1.03, 1.63) 1.28 (1.03, 1.62) 1.56 (1.21, 1.96) 1.55 (1.20, 1.94)
Ref Ref
1.10 (0.88, 1.38) 1.08 (0.86, 1.36)
3.64 (1.85, 7.15) 3.08 (1.56, 6.08) 1.90 (1.04, 3.55) 1.78 (0.98, 3.25) 1.88 (1.24, 2.85) 1.61 (1.06, 2.45) 1.95 (1.27, 2.99) 1.58 (1.03, 2.44)
Ref Ref
1.86 (1.02, 3.38) 1.71 (0.92, 3.20)
2.12 (1.31, 3.44) 1.95 (1.20, 3.17) 1.63 (1.11, 2.40) 1.48 (1.00, 2.18) 1.33 (0.90, 1.98) 1.24 (0.84, 1.84) Ref Ref 1.88 (1.42, 2.47) 1.66 (1.26, 2.20) 2.23 (1.69, 2.93) 1.85 (1.40, 2.44)
1 kg & 1 kg (n ¼ 1454) 5 kg & < 1 kg (n ¼ 1175) 10 kg & < 5 kg (n ¼ 925) < 10 kg (n ¼ 425)
Weight change
Table 3 Associations between weight change and outcomes.
All-cause mortality Model 1 3.54 (2.63, 4.77) Model 2 2.91 (2.15, 3.93) Cardiovascular mortality Model 1 2.71 (1.68, 4.38) Model 2 2.23 (1.37, 3.63) Cardiovascular events Model 1 2.17 (1.29, 3.29) Model 2 2.11 (1.22, 3.15)
>1 kg & < 5 kg (n ¼ 593)
10 kg (n ¼ 216)
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
5 kg & < 10 kg (n ¼ 488)
108
significant no matter whether body weight at baseline was underweight, normal weight, overweight, or obese (Fig. 2). When the percentage of change in weight was used to define weight change, risks for death and cardiovascular events were lowest in participants with weight stability (3% loss or gain) and increased with the extent of weight loss or weight gain (Supplemental Tables 4 and 5). Similar results were found when average weight change per year were considered (Supplemental Tables 6 and 7). Finally, when the levels of baseline blood pressure, serum lipids, and fasting glucose were further adjusted in Cox regression models, the associations of baseline BMI levels and weight change with adverse outcomes were unchanged (Supplemental Tables 8 and 9). 4. Discussion The present study indicated that being overweight at baseline was related to less risks for mortality and cardiovascular events. Risks for adverse outcomes increased with the decrease of BMI levels and with the increase of weight loss or weight gain during the follow-up. These associations were independent of baseline BMI levels. Obesity is an independent risk factor for many chronic disorders, such as diabetes, CHD, cancer, and neuromuscular and skeletal disorders [6,24,25]. Epidemiologic studies have identified associations of both overweight and obesity with higher all-cause mortality in the general population [3,26]. However, an obesity paradox has been reported in which overweight and obesity is associated with a decreased risk for morbidity and mortality among patients with established CHD [18,19]. Similar findings have been described in patients with heart failure, hypertension, or atrial fibrillation [27,28]. Reasons for these findings remain in debate. In our study, although overweight and obese patients had higher prevalence of diabetes and hypertension, they were younger and were more likely to receive target lesion revascularization and medication therapies, similar to the literature [29]. Several studies supposed that these characteristics might be an explanation for the “protective” effect of obesity [29,30]. Furthermore, previous studies reported that metabolic abnormality was closer to CVD morbidity and mortality compared to obesity [31,32]. We speculate that obesity increases risks for chronic diseases morbidity but its effect on adverse outcomes would be attenuated after chronic diseases have been developed. Subsequently, the protective effects of overweight, such as nutritional reserve and stores of lean mass, may prevail over its adverse effects on cardiovascular outcomes in patients with established CHD in our study. Although weight loss is recommended in obese patients [9], even for those with established CHD [10,12], the question of whether such change in body weight is beneficial to health is controversial [15,16]. Several studies have shown that weight loss increases in cardiorespiratory fitness and improves in insulin resistance and related cardiometabolic risk factors [14,28], which may suggest favorable prognostic effects in patients with established CHD [11]. However, observations from recent studies and the present study show that the long-term clinical outcomes are not as expected [16,20,33]. Compared with the previous studies, our study population is larger, which helped us firstly analysed of the effect of changes in BMI levels on prognosis in patients with CHD. Interestingly, decreases in BMI levels contributed to increases in adverse outcomes, regardless of baseline BMI levels. Previous study reported that weight loss interventions, such as exercise and caloric restriction, might lead to loss of muscle mass and result in impaired physical function [34,35]. We speculate that
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
109
Fig. 2. Stratified analyses of associations between weight change and all-cause mortality. Scales on the y-axis are log transformed and represent hazard ratios and 95% confidence interval. Age, sex, height, current smoking, hypertension, diabetes, stroke, family history of coronary disease, previous PCI, previous CABG, medication use, serum creatinine, high-sensitivity CRP, and NT-proBNP at baseline were adjusted. The covariate was not included in models when the participants were stratified by this covariate.
these adverse effects may cause the increases in risks for adverse outcomes in patients with CHD. However, our study was focused on cardiovascular outcomes and all-cause death. We cannot rule out the possibility of a different effect of weight loss on other health outcomes in patients with CHD. Finally, we also found that risks for cardiovascular events and mortality increased with the increase of weight gain, which was in line with the literature [36,37]. Although the association between small weight gain and adverse outcomes was not significant in our study, it is now widely accepted that moderate or severe weight gain results in adipocyte hypertrophy and an increased in obesityrelated disorders, which cause the increasing risks for cardiovascular outcomes [38]. Furthermore, smoking cessation is recommended after CHD. Smoking cessation, however, is known to cause
weight gain [39]. Although our stratified analyses showed that large weight gain increased mortality risk both in non- or exsmokers and in current smokers, further investigations are needed to elucidate the cardiovascular benefits achieved by smoking cessation during the follow-up and the untoward effects of weight gain in patients with CHD. The present study has several limitations. First, the duration of obesity and weight change over time before the observational period are important confounders for the assessment of obesityrelated adverse prognosis, but our study lacks these data, which may have caused to misestimate the impact of obesity or weight change on adverse outcome risks. Second, although the measurement of weight change in our study was also used in many previous studies [20,40,41], weight change may not well reflect weight
110
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
fluctuation during the follow-up. Weight fluctuation was reported to be a better marker in terms of assessment of adverse outcomes risks [15,36]. Third, the effects on outcomes between intentional and unintentional weight changes were reported to be different [16], but our study was not able to differentiate between intentional and unintentional weight changes. We speculate that a majority of patients with CAD have an intention for controlling their weight based on suggestions from their physicians. Fourth, metabolic dysfunction is known to be independently associated with CHD risk and may play an important role in heterogeneous effects of obesity on prognosis. Although metabolic abnormalities, such as hypertension and diabetes, were adjusted in our multivariate regression models, their roles in the effects of obesity on adverse outcomes need further investigation in patients with CHD. Finally, other measurements of adiposity, such as waist circumference and body fat percentage could not be obtained in our study. The literature reported that high body fat percentage was associated with increased mortality even in those with low BMI [42], suggesting that other measurements of adiposity may help explain the heterogeneous effects of obesity on adverse outcomes. In conclusion, the present study found that obesity did not increase risks for cardiovascular events and mortality in patients with CHD and that risks for adverse cardiovascular outcomes increased with the increase of weight loss or weight gain. Our results suggest that the recommendation of maintaining a normal body weight for patients with CHD should be reconsidered. Maintaining a stable weight may be a better strategy for the reduction of risks for cardiovascular outcomes and all-cause death in this population, regardless of baseline body weight. Further investigations on heterogeneous effects of obesity and weight change are needed.
[5]
[6]
[7]
[8]
[9] [10]
[11] [12]
[13]
[14] [15] [16]
[17]
Conflicts of interest [18]
The authors declared they do not have anything to disclose regarding conflict of interest with respect to this manuscript. Author contributions Sheng-Yong Dong and Qiang Zeng had full access to all the data in the study and designed the study. Shuang-Tong Yan drafted the manuscript. Man-Liu Wang, Zhi-Bing Li, and Lian-Qing Fang collected the data. Sheng-Yong Dong and Qiang Zeng analyzed data and revised the manuscript.
[19]
[20]
[21]
[22]
Acknowledgements The authors thank all participants for their important contributions.
[23]
Appendix A. Supplementary data
[24]
Supplementary data related to this article can be found at https://doi.org/10.1016/j.atherosclerosis.2018.05.007.
[25] [26]
References [1] G.A. Roth, C. Johnson, A. Abajobir, et al., Global, regional, and national burden of cardiovascular diseases for 10 causes, 1990 to 2015, J. Am. Coll. Cardiol. 70 (1) (2017) 1e25. [2] G.A. Roth, M.D. Huffman, A.E. Moran, et al., Global and regional patterns in cardiovascular mortality from 1990 to 2013, Circulation 132 (17) (2015) 1667e1678. [3] GBD 2015 Obesity Collaborators, A. Afshin, M.H. Forouzanfar, et al., Health effects of overweight and obesity in 195 countries over 25 years, N. Engl. J. Med. 377 (1) (2017) 13e27. [4] NCD Risk Factor Collaboration (NCD-RisC), Trends in adult body-mass index in 200 countries from 1975 to 2014: a pooled analysis of 1698 population-based
[27]
[28]
[29]
[30]
measurement studies with 19.2 million participants, Lancet 387 (10026) (2016) 1377e1396. GBD 2016 Risk Factors Collaborators, Global, regional, and national comparative risk assessment of 84 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016, Lancet 390 (10100) (2017) 1345e1422. C. Lassale, I. Tzoulaki, K.G.M. Moons, et al., Separate and combined associations of obesity and metabolic health with coronary heart disease: a panEuropean case-cohort analysis, Eur. Heart J. 39 (5) (2018) 397e406. A.J. Flint, F.B. Hu, R.J. Glynn, et al., Excess weight and the risk of incident coronary heart disease among men and women, Obesity (Silver Spring) 18 (2) (2010) 377e383. K. Bibbins-Domingo, P. Coxson, M.J. Pletcher, et al., Adolescent overweight and future adult coronary heart disease, N. Engl. J. Med. 357 (23) (2007) 2371e2379. B.M. Wolfe, E. Kvach, R.H. Eckel, Treatment of obesity: weight loss and bariatric surgery, Circ. Res. 118 (11) (2016) 1844e1855. J. Perk, G. De Backer, H. Gohlke, et al., European guidelines on cardiovascular disease prevention in clinical practice (version 2012). The fifth joint task force of the European society of cardiology and other societies on cardiovascular disease prevention in clinical practice (constituted by representatives of nine societies and by invited experts), Eur. Heart J. 33 (13) (2012) 1635e1701. P.A. Ades, P.D. Savage, Potential benefits of weight loss in coronary heart disease, Prog. Cardiovasc. Dis. 56 (4) (2014) 448e456. S.C. Smith Jr., E.J. Benjamin, R.O. Bonow, et al., AHA/ACCF secondary prevention and risk reduction therapy for patients with coronary and other atherosclerotic vascular disease: 2011 update: a guideline from the american heart association and american college of cardiology foundation, Circulation 124 (22) (2011) 2458e2473. C.J. Lavie, R.V. Milani, H.O. Ventura, Obesity and cardiovascular disease: risk factor, paradox, and impact of weight loss, J. Am. Coll. Cardiol. 53 (21) (2009) 1925e1932. T.D. Adams, L.E. Davidson, S.E. Litwin, et al., Weight and metabolic outcomes 12 Years after gastric bypass, N. Engl. J. Med. 377 (12) (2017) 1143e1155. S. Bangalore, R. Fayyad, R. Laskey, et al., Body-weight fluctuations and outcomes in coronary disease, N. Engl. J. Med. 376 (14) (2017) 1332e1340. Q.R. Pack, J.P. Rodriguez-Escudero, R.J. Thomas, et al., The prognostic importance of weight loss in coronary artery disease: a systematic review and metaanalysis, Mayo Clin. Proc. 89 (10) (2014) 1368e1377. J. Sierra-Johnson, A. Romero-Corral, V.K. Somers, et al., Prognostic importance of weight loss in patients with coronary heart disease regardless of initial body mass index, Eur. J. Cardiovasc. Prev. Rehabil. 15 (3) (2008) 336e340. J. Niedziela, B. Hudzik, N. Niedziela, et al., The obesity paradox in acute coronary syndrome: a meta-analysis, Eur. J. Epidemiol. 29 (11) (2014) 801e812. C.J. Lavie, A. De Schutter, D. Patel, et al., Body composition and coronary heart disease mortalityean obesity or a lean paradox? Mayo Clin. Proc. 86 (9) (2011) 857e864. F. Lopez-Jimenez, C.O. Wu, X. Tian, et al., Weight change after myocardial infarctionethe enhancing recovery in coronary heart disease patients (ENRICHD) experience, Am. Heart J. 155 (3) (2008) 478e484. B.F. Zhou, Cooperative Meta-Analysis Group of the Working Group on Obesity in China, Predictive values of body mass index and waist circumference for risk factors of certain related diseases in Chinese adultsestudy on optimal cutoff points of body mass index and waist circumference in Chinese adults, Biomed. Environ. Sci. 15 (1) (2002) 83e96. C.P. Cannon, R.G. Brindis, B.R. Chaitman, et al., 2013 ACCF/AHA key data elements and definitions for measuring the clinical management and outcomes of patients with acute coronary syndromes and coronary artery disease: a report of the American College of Cardiology Foundation/American Heart Association Task Force on clinical data standards (writing committee to develop acute coronary syndromes and coronary artery disease clinical data standards), J. Am. Coll. Cardiol. 61 (9) (2013) 992e1025. K. Aho, P. Harmsen, S. Hatano, et al., Cerebrovascular disease in the community: results of a WHO collaborative study, Bull. World Health Organ. 58 (1) (1980) 113e130. R. Caleyachetty, G.N. Thomas, K.A. Toulis, et al., Metabolically healthy obese and incident cardiovascular disease events among 3.5 million men and women, J. Am. Coll. Cardiol. 70 (12) (2017) 1429e1437. C.E. Ndumele, L. Cobb, M. Lazo, et al., Weight history and subclinical myocardial damage, Clin. Chem. 64 (1) (2018) 201e209. Global BMI Mortality Collaboration, E. Di Angelantonio, N. Bhupathiraju Sh, et al., Body-mass index and all-cause mortality: individual-participant-data meta-analysis of 239 prospective studies in four continents, Lancet 388 (10046) (2016) 776e786. S.M. Artham, C.J. Lavie, R.V. Milani, et al., Obesity and hypertension, heart failure, and coronary heart disease-risk factor, paradox, and recommendations for weight loss, Ochsner J. 9 (3) (2009) 124e132. C.J. Lavie, A. Pandey, D.H. Lau, et al., Obesity and atrial fibrillation prevalence, pathogenesis, and prognosis: effects of weight loss and exercise, J. Am. Coll. Cardiol. 70 (16) (2017) 2022e2035. A. Oreopoulos, F.A. McAlister, K. Kalantar-Zadeh, et al., The relationship between body mass index, treatment, and mortality in patients with established coronary artery disease: a report from APPROACH, Eur. Heart J. 30 (21) (2009) 2584e2592. B.A. Steinberg, C.P. Cannon, A.F. Hernandez, et al., Medical therapies and
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
[31]
[32]
[33]
[34] [35]
invasive treatments for coronary artery disease by body mass: the "obesity paradox" in the Get with the Guidelines database, Am. J. Cardiol. 100 (9) (2007) 1331e1335. Global Burden of Metabolic Risk Factors for Chronic Diseases Collaboration (BMI Mediated Effects), Y. Lu, K. Hajifathalian, et al., Metabolic mediators of the effects of body-mass index, overweight, and obesity on coronary heart disease and stroke: a pooled analysis of 97 prospective cohorts with 1.8 million participants, Lancet 383 (9921) (2014) 970e983. Q. Zeng, S.Y. Dong, M.L. Wang, et al., Obesity and novel cardiovascular markers in a population without diabetes and cardiovascular disease in China, Prev. Med. 91 (2016) 62e69. T. Moholdt, C.J. Lavie, J. Nauman, Sustained physical activity, not weight loss, associated with improved survival in coronary heart disease, J. Am. Coll. Cardiol. 71 (10) (2018) 1094e1101. B. Cao, Estimating the effects of obesity and weight change on mortality using a dynamic causal model, PLoS One 10 (6) (2015), e0129946. E.M. Weinheimer, L.P. Sands, W.W. Campbell, A systematic review of the separate and combined effects of energy restriction and exercise on fat-free mass in middle-aged and older adults: implications for sarcopenic obesity,
111
Nutr. Rev. 68 (7) (2010) 375e388. [36] E. Yu, S.H. Ley, J.E. Manson, et al., Weight history and all-cause and causespecific mortality in three prospective cohort studies, Ann. Intern. Med. 166 (9) (2017) 613e620. [37] K.F. Adams, M.F. Leitzmann, R. Ballard-Barbash, et al., Body mass and weight change in adults in relation to mortality risk, Am. J. Epidemiol. 179 (2) (2014) 135e144. [38] L.K. Forsythe, J.M. Wallace, M.B. Livingstone, Obesity and inflammation: the effects of weight loss, Nutr. Res. Rev. 21 (2) (2008) 117e133. [39] P. Dehghani, B. Habib, S.B. Windle, et al., Smokers and postcessation weight gain after acute coronary syndrome, J Am Heart Assoc 6 (4) (2017). [40] Z.B. Li, M.L. Wang, S.Y. Dong, et al., Effects of body mass index and weight change on mortality in older men with impaired glucose regulation, Exp. Gerontol. 89 (2017) 87e92. [41] S.Y. Park, L.R. Wilkens, G. Maskarinec, et al., Weight change in older adults and mortality: the Multiethnic Cohort Study, Int. J. Obes. 42 (2) (2018) 205e212. [42] R. Padwal, W.D. Leslie, L.M. Lix, et al., Relationship among body fat percentage, body mass index, and all-cause mortality: a cohort study, Ann. Intern. Med. 164 (8) (2016) 532e541.
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Associations of body weight and weight change with cardiovascular events and mortality in patients with coronary heart disease Sheng-Yong Dong a, *, 1, Shuang-Tong Yan b, 1, Man-Liu Wang c, Zhi-Bing Li b, Lian-Qing Fang a, Qiang Zeng d, ** a
Healthcare Department, Agency for Offices Administration of PLA, Beijing, 100034, China Department of Geriatric Endocrinology, Chinese PLA General Hospital, Beijing, 100853, China Peking University-Tsinghua University Joint Center for Life Sciences, Beijing, China, Advanced Academy of Interdisciplinary Sciences, Peking University, Beijing, 100871, China d Health Management Institute, Chinese PLA General Hospital, Beijing, 100853, China b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 November 2017 Received in revised form 31 March 2018 Accepted 2 May 2018 Available online 5 May 2018
Background and aims: It is recommended that patients with coronary heart disease (CHD) pursue a normal body weight, while the effects of body weight and weight change on prognosis are still controversial. The present study was to assess these effects using a large-scale population with CHD in China. Methods: A total of 5276 patients with CHD were included from Jan 2000 to Dec 2014. Baseline and endpoint weights were measured. Outcomes including mortality and cardiovascular events were obtained. Results: Relative to patients with normal weight, risks for adverse outcomes were lowest in overweight patients and similar in obese patients. Hazard ratios (HRs) and 95% confidence interval (95% CI) for allcause death were 1.42 (1.06, 1.91) if overweight turned into normal weight and were 2.01 (1.28, 3.16) or 5.33 (2.81, 10.1) if obese turned into overweight or normal weight. Death risk increased with the extent of weight loss and moderate or large weight gain (p<0.05 for all). Similar results were found when risks for cardiovascular mortality and events were considered. Furthermore, these results remained significant when the patients were stratified by several covariates and even when several definitions of weight change were considered. Conclusions: Obesity did not increase adverse outcome risks in patients with CHD. Both weight loss and weight gain increased adverse outcome risks regardless of baseline body weight. The findings suggest that maintaining a stable weight may be a better strategy for the reduction of risks for cardiovascular outcomes and all-cause death in patients with CHD. © 2018 Elsevier B.V. All rights reserved.
Keywords: Coronary heart disease Body weight Obesity Weight change Mortality Cardiovascular event
1. Introduction Although cardiovascular disease (CVD) mortality declined dramatically over the past 2 decades in all high-income and some middle-income countries, CVDs remain the leading causes of premature death and chronic disability in the world [1,2]. The majority
* Corresponding author. Healthcare Department, Agency for Offices Administration of PLA, 2 Aimin Street, Xicheng District, Beijing, 100034, China. ** Corresponding author. Health Management Institute, Chinese PLA General Hospital, 28 Fuxing Road, Beijing, 100853, China. E-mail addresses: [email protected] (S.-Y. Dong), [email protected] (Q. Zeng). 1 These authors contributed equally to this work. https://doi.org/10.1016/j.atherosclerosis.2018.05.007 0021-9150/© 2018 Elsevier B.V. All rights reserved.
of health lost to CVDs are attributable to coronary heart disease (CHD) and stroke [1]. It is concerning that trends in CHD mortality are no longer declining for many world regions, which may be due to rising rates of cardiovascular risk factors particularly obesity [1,3e5]. Previous studies reported that obesity increased risks for CHD [6,7], which was thought to result in an increase of adverse outcomes [8]. Based on these findings, guidelines recommend and the literature encourage patients with CHD to pursue a normal body weight, especially in overweight or obese patients [9e12]. However, in recent years, increasing efforts have been made to assess the potential effects of obesity and weight change on a variety of health outcomes, while the findings are controversial [13e17].
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
Several studies have shown that overweight and obesity are related to better prognosis in patients with established CHD [18,19]. Other studies have shown that loss of weight in overweight or obese patients with CHD does not result in lower mortality and may even have worse prognosis [16,20]. These controversial findings raise questions regarding the effect of obesity on prognosis and the value of weight loss in patients with CHD. Furthermore, many of above studies had relatively small number of participants [16,17,19], and there are currently no large-scale studies involving the Asian population. Finally, data on analyses of associations between changes in body weight categories (i.e., body weight changes from normal weight to overweight) and prognosis in patients with CHD are limited. Therefore, we investigated associations of body weight and weight change with outcomes using a large-scale population with CHD in China. We specifically explored how risks for adverse outcomes change with changes in body weight categories, and also decomposed changes in risks of weight change for deaths and cardiovascular events into the contributions of sex, age, and all other factors. 2. Patients and methods 2.1. Patients The present study recruited the patients with established CHD who visited the Health Department of Agency for Offices Administration of PLA or the Chinese PLA General Hospital from Jan 2000 to Dec 2014. Eligible patients were men and women, 18e90 years of age, with clinically evident CHD, defined by one or more of the following: previous myocardial infarction (MI), the presence of angina pectoris with angiographic evidence of >50% stenosis in at least 1 major coronary artery, previous percutaneous coronary intervention (PCI), and a history of coronary artery bypass grafting (CABG). The exclusion criteria included a history of lung disease, kidney disease, liver disease, heart failure, mental illness, or cancer, or inability to participate in the study. After the exclusion of 104 patients who were lost to follow-up and 130 patients with missing data on covariates, complete data were obtained in 5276 patients. All known diseases were obtained from hospital medical records and confirmed by a medical record review. Written informed consent was obtained from all of the included patients. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Medical Ethics Committee of Chinese PLA General Hospital. The study population flow is explained in Supplemental Fig. 1. 2.2. Body weight and weight change Baseline and endpoint weights were obtained from hospital medical records. Self-reported endpoint weights through telephone visit were used in 185 participants because body weights at the endpoint were not able to be obtained from their hospital medical records. Baseline characteristics were comparable between these participants and the participants whose endpoint weights were obtained from hospital medical records. Body mass index (BMI) was calculated in kg/m2 and was categorized into: underweight, BMI <18.5 kg/m2; normal weight, 18.5e23.9 kg/m2; overweight, 24e27.9 kg/m2; and obese, 28 kg/m2, according to the recommendation by the Working Group on Obesity in China [21]. Weight change was defined by absolute weight change (kg). Absolute weight change was calculated by subtracting weight at baseline from weight at endpoint and were categorized into: weight stability (1 kg loss or gain), small weight gain (1e5 kg gain), moderate weight gain (5e10 kg gain), large weight gain
105
(>10 kg gain), small weight loss (1e5 kg loss), moderate weight loss (5e10 kg loss), and large weight loss (>10 kg loss). 2.3. Covariates Data on smoking status and family medical history were conducted via face-to-face interviews. Individual medical history and medication use were obtained from hospital medical records. Fasting glucose, serum lipids, serum creatinine, high-sensitivity Creactive protein (CRP), and N-terminal pro-B-type natriuretic peptide (NT-proBNP) were measured by routine laboratory methods. 2.4. Follow-up and adverse outcomes Participants were followed until Dec 31, 2016. The follow-up protocol included a combination of hospital medical records, telephone contacts with patients or family members, and death certificates. The primary outcomes were all-cause mortality and cardiovascular mortality. Cardiovascular mortality was defined as death attributable to fatal myocardial infarction, sudden cardiac death, or stroke. The second outcomes were cardiovascular events, including myocardial infarction, stroke, and coronary revascularization (i.e., PCI and/or CABG). Patients with more than one event were assigned the highest-ranked event based on the previous list of cardiovascular events. The diagnosis of myocardial infarction was based on symptoms, elevated cardiac marker of myocardial necrosis, and the presence or absence of diagnostic electrocardiogram changes [22]. Sudden cardiac death was diagnosed as unexpected natural death due to cardiac causes within 1 h of onset of acute symptoms. Stroke included ischemic stroke and hemorrhagic stroke, which were diagnosed according to World Health Organization criteria [23]. All deaths were classified using the tenth revision of the International Classification of Disease and confirmed through death certificates using personal identity card number. All events were confirmed by hospital medical records and adjudicated by an event adjudication committee who were blinded to the measurements of body weight and weight change. 2.5. Statistics Continuous variables are expressed as the mean ± standard deviation (SD) or median (interquartile range) and categorical variables as percentages. Descriptive statistics were performed using one-way analysis of variance or chi-square tests. Nonnormally distributed variables were log transformed before analyses. Age- and sex-standardized mortality and cardiovascular events per 1000 person years across BMI levels and weight change levels were calculated, respectively. Multivariate Cox proportional hazards models were used to assess associations of BMI levels and weight change with outcomes. Associations between changes in BMI levels and outcomes were also assessed. Hazard ratio (HR) and 95% confidence interval (CI) were shown. In sensitivity analyses, associations of weight change with allcause mortality were assessed in the patients stratified by baseline covariates. Furthermore, associations of the percentage of change in weight and average weight change per year with outcomes were assessed, respectively. Calculations and categories of the percentage of changing in weight and average weight change per year are described in the Supplemental material. Finally, the levels of baseline blood pressure, serum lipids, and fasting glucose were further adjusted in Cox models. All analyses were conducted using SPSS 18.0 (SPSS Inc., Chicago, Illinois), and a two-sided value of p < 0.05 was considered statistically significant.
106
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
A total of 5276 patients with complete data met the study criteria and were analysed in the present study. The mean baseline weight of the patients was 71.0 ± 11.5 kg. Among all of the patients, 20.4% were obese, 45.5% were overweight, and 1.8% were underweight by the BMI criteria (Table 1). The percentage of patients with hypertension or diabetes increased with BMI levels being increased (p < 0.01 for all). The percentage of patients who received PCI, CABG, or medication use also increased with the increase of BMI levels (p < 0.05 for all).
increased if BMI levels at the endpoint decreased one level or more levels, regardless of BMI levels at baseline (Supplemental Table 2). After adjustment for covariates, risk for all-cause death increased 57% if body weight changed from normal weight to underweight (P ¼ 0.019) (Fig. 1). The HRs (95% CI) for all-cause death were 1.42 (1.06, 1.91) or 3.16 (1.52, 6.55) if body weight changed from overweight to normal weight or underweight, respectively. If body weight changed from obese to overweight, normal weight, or underweight, mortality risk increased to 2.01 (1.28, 3.16), 5.33 (2.81, 10.1), or 7.94 (1.14, 60.8), respectively. Similarly, multivariate analyses shown that risks for cardiovascular mortality and cardiovascular events increased if BMI levels decreased from the baseline to the endpoint.
3.2. BMI levels, changes in BMI levels, and outcomes
3.3. Weight change and outcomes
During 21,251 person-years of follow-up, we identified 540 deaths of all-cause, 216 cardiovascular deaths, and 1116 cardiovascular events (Supplemental Table 1). The age- and sexstandardized rates of all-cause death, cardiovascular death, and cardiovascular events were lowest in overweight patients (23, 9, and 60 per 1000 person years, respectively) and highest in underweight patients (44, 22, and 131 per 1000 person years, respectively). As shown in Table 2, risks for all-cause mortality, cardiovascular mortality, and cardiovascular events reduced 24%, 27%, and 19%, respectively, in overweight patients, while these risks increased 78%, 130%, and 88%, respectively, in underweight patients, relative to patients with normal weight after adjustment for covariates (p < 0.05 for all). Furthermore, obese patients had similar risks for adverse outcomes compared to normal weight patients in multivariate analyses. Death rate and the prevalence of cardiovascular events decreased if BMI levels at the endpoint increased one level but
The mean (SD) of weight change over a mean follow-up of 4 years was a loss of 1.9 kg (6.8 kg). The age- and sex-standardized all-cause death, cardiovascular death, and cardiovascular events per 1000 person years were lowest in participants with weight stability (1 kg loss or gain) and increased with the extent of weight loss or weight gain (Supplemental Fig. 2). Relative to weight stability, the HRs (95% CI) of small, moderate, and larger weight loss for all-cause mortality were 1.66 (1.26, 2.20), 1.85 (1.40, 2.44), and 2.91 (2.15, 3.93), respectively (Table 3). Patients with moderate or large weight gain also had a 48% or 95% increase in risk for all-cause death (p < 0.05). Similar results were found when associations of cardiovascular mortality and cardiovascular events with weight change were considered.
3. Results 3.1. Baseline characteristics
3.4. Sensitivity analyses When participants were stratified by sex, age groups (<60 years,
Table 1 Baseline characteristics.
Age, year Female Current smoking Hypertension Diabetes Stroke Family history of coronary disease Body weight, kg BMI, kg/m2 Systolic BP, mmHg Diastolic BP, mmHg Fasting glucose, mmol/l Total cholesterol, mmol/l LDL cholesterol, mmol/l HDL cholesterol, mmol/l Triglyceride, mmol/la Serum creatinine, mmol/la High-sensitivity CRP, mg/dla NT-proBNP, pg/mla Previous MI Previous PCI Previous CABG Medication use Statin Aspirin ACE inhibitor/ARB b-Blocker
Total (n ¼ 5276)
Underweight (n ¼ 93)
Normal weight (n ¼ 1708)
Overweight (n ¼ 2398)
Obese (n ¼ 1076)
p value
67.6 ± 11.8 1211 (23.0) 1023 (19.4) 3630 (68.8) 1355 (25.7) 868 (16.5) 292 (5.5) 71.0 ± 11.5 25.4 ± 3.4 134 ± 18 77 ± 11 6.21 ± 2.37 4.56 ± 1.09 2.52 ± 0.87 1.18 ± 0.35 1.36 (0.99, 1.85) 82 (69, 96) 0.20 (0.10, 0.83) 234 (80.884) 1374 (26.0) 1876 (35.6) 430 (8.2)
73.7 ± 9.2 26 (28.0) 12 (12.9) 51 (54.8) 17 (18.3) 19 (20.4) 1 (1.1) 48.1 ± 5.6 17.3 ± 1.1 128 ± 23 71 ± 13 5.78 ± 1.80 4.50 ± 1.05 2.42 ± 0.78 1.45 ± 0.49 0.85 (0.66, 0.96) 81 (65, 91) 0.30 (0.10, 1.86) 1293 (297, 2832) 26 (28.0) 21 (22.6) 3 (3.2)
69.5 ± 11.3 447 (26.2) 299 (17.5) 1070 (62.6) 398 (23.3) 285 (16.7) 77 (4.5) 61.4 ± 6.8 22.1 ± 1.4 132 ± 19 74 ± 11 6.11 ± 2.30 4.62 ± 1.16 2.55 ± 0.90 1.27 ± 0.39 1.17 (0.87, 1.59) 80 (67, 95) 0.20 (0.10, 1.00) 293 (101, 1287) 467 (27.3) 562 (32.9) 130 (7.6)
67.4 ± 11.5 486 (20.3) 474 (19.8) 1687 (70.4) 616 (25.7) 394 (16.4) 118 (4.9) 72.7 ± 6.9 25.9 ± 1.1 134 ± 18 76 ± 11 6.22 ± 2.39 4.55 ± 1.10 2.50 ± 0.86 1.15 ± 0.32 1.42 (1.06, 1.91) 82 (70, 96) 0.20 (0.10, 0.70) 213 (74, 836) 621 (25.9) 870 (36.3) 192 (8.0)
64.4 ± 12.7 252 (23.4) 238 (22.1) 822 (76.4) 324 (30.1) 170 (15.8) 96 (8.9) 84.3 ± 9.8 30.2 ± 2.1 135 ± 17 78 ± 10 6.40 ± 2.45 4.50 ± 1.09 2.50 ± 0.84 1.08 ± 0.30 1.63 (1.17, 2.06) 82 (71, 96) 0.21 (0.10, 0.78) 148 (65, 562) 260 (24.2) 423 (39.3) 105 (9.8)
<0.001 0.011 0.001 <0.001 <0.001 0.371 <0.001 <0.001 <0.001 <0.001 <0.001 0.015 0.845 0.494 <0.001 <0.001 0.186 0.396 <0.001 0.058 <0.001 0.018
3189 3944 2767 1816
40 56 37 25
957 (56.0) 1233 (72.1) 829 (48.5) 529 (31.0)
1504 (62.7) 1812 (75.6) 1303 (54.3) 884 (36.9)
688 843 598 378
<0.001 <0.001 <0.001 0.002
(60.4) (74.8) (52.4) (34.4)
(43.0) (60.2) (39.8) (26.9)
(63.9) (78.3) (55.6) (35.1)
Variables are shown as n (%), mean ± standard deviation, or median (quartiles). BMI, body mass index; BP, blood pressure; LDL, low-density lipoprotein; HDL, high-density lipoprotein; CRP, C-reactive protein; NT-proBNP, N-terminal pro-B-type natriuretic peptide; MI, myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass grafting; ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker. a These variables were log transformed before being tested.
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
107
Table 2 Associations between BMI levels and outcomes. Underweight HR (95% CI) All-cause mortality Model 1 1.86 (1.21, Model 2 1.78 (1.15, Cardiovascular mortality Model 1 2.50 (1.32, Model 2 2.30 (1.22, Cardiovascular events Model 1 1.76 (1.23, Model 2 1.88 (1.31,
Normal weight p
Overweight
Obese
HR (95% CI)
p
HR (95% CI)
p
2.86) 2.76)
0.005 0.010
Ref Ref
0.75 (0.62, 0.91) 0.76 (0.62, 0.92)
0.003 0.005
0.91 (0.71, 1.16) 0.90 (0.70, 1.15)
0.455 0.383
4.76) 4.34)
0.005 0.010
Ref Ref
0.74 (0.54, 0.99) 0.73 (0.53, 0.99)
0.045 0.046
1.09 (0.75, 1.58) 1.01 (0.69, 1.47)
0.660 0.968
2.51) 2.70)
0.002 0.001
Ref Ref
0.82 (0.72, 0.94) 0.81 (0.71, 0.93)
0.005 0.003
0.99 (0.84, 1.16) 0.95 (0.81, 1.12)
0.852 0.552
Model 1 was adjusted for age, sex, current smoking, hypertension, diabetes, stroke, family history of coronary disease, previous PCI, previous CABG, and medication use. Model 2 was further adjusted for serum creatinine, high-sensitivity CRP, and NT-proBNP at baseline.
Fig. 1. Risks of changes in BMI levels. For all-cause mortality (A), cardiovascular mortality (B), and cardiovascular events (C). Scales on the x-axis represent hazard ratios and 95% confidence interval adjusted for age, sex, current smoking, hypertension, diabetes, stroke, family history of coronary disease, previous PCI, previous CABG, medication use, serum creatinine, high-sensitivity CRP, and NTproBNP at baseline.
60e74 years, and 75 years), smoking status, hypertension, diabetes, and stroke, respectively, the positive associations of weight
loss or weight gain with risk for all-cause death were unchanged in all subgroups (Supplemental Table 3). These associations were also
Data are shown as hazard ratios and 95% confidence interval. Model 1 was adjusted for age, sex, height, current smoking, hypertension, diabetes, stroke, family history of coronary disease, previous PCI, previous CABG, and medication use. Model 2 was further adjusted for serum creatinine, high-sensitivity CRP, and NT-proBNP at baseline.
1.77 (1.27, 2.89) 1.70 (1.22, 2.77) 1.37 (1.02, 1.70) 1.32 (1.01, 1.64) 1.30 (1.03, 1.63) 1.28 (1.03, 1.62) 1.56 (1.21, 1.96) 1.55 (1.20, 1.94)
Ref Ref
1.10 (0.88, 1.38) 1.08 (0.86, 1.36)
3.64 (1.85, 7.15) 3.08 (1.56, 6.08) 1.90 (1.04, 3.55) 1.78 (0.98, 3.25) 1.88 (1.24, 2.85) 1.61 (1.06, 2.45) 1.95 (1.27, 2.99) 1.58 (1.03, 2.44)
Ref Ref
1.86 (1.02, 3.38) 1.71 (0.92, 3.20)
2.12 (1.31, 3.44) 1.95 (1.20, 3.17) 1.63 (1.11, 2.40) 1.48 (1.00, 2.18) 1.33 (0.90, 1.98) 1.24 (0.84, 1.84) Ref Ref 1.88 (1.42, 2.47) 1.66 (1.26, 2.20) 2.23 (1.69, 2.93) 1.85 (1.40, 2.44)
1 kg & 1 kg (n ¼ 1454) 5 kg & < 1 kg (n ¼ 1175) 10 kg & < 5 kg (n ¼ 925) < 10 kg (n ¼ 425)
Weight change
Table 3 Associations between weight change and outcomes.
All-cause mortality Model 1 3.54 (2.63, 4.77) Model 2 2.91 (2.15, 3.93) Cardiovascular mortality Model 1 2.71 (1.68, 4.38) Model 2 2.23 (1.37, 3.63) Cardiovascular events Model 1 2.17 (1.29, 3.29) Model 2 2.11 (1.22, 3.15)
>1 kg & < 5 kg (n ¼ 593)
10 kg (n ¼ 216)
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
5 kg & < 10 kg (n ¼ 488)
108
significant no matter whether body weight at baseline was underweight, normal weight, overweight, or obese (Fig. 2). When the percentage of change in weight was used to define weight change, risks for death and cardiovascular events were lowest in participants with weight stability (3% loss or gain) and increased with the extent of weight loss or weight gain (Supplemental Tables 4 and 5). Similar results were found when average weight change per year were considered (Supplemental Tables 6 and 7). Finally, when the levels of baseline blood pressure, serum lipids, and fasting glucose were further adjusted in Cox regression models, the associations of baseline BMI levels and weight change with adverse outcomes were unchanged (Supplemental Tables 8 and 9). 4. Discussion The present study indicated that being overweight at baseline was related to less risks for mortality and cardiovascular events. Risks for adverse outcomes increased with the decrease of BMI levels and with the increase of weight loss or weight gain during the follow-up. These associations were independent of baseline BMI levels. Obesity is an independent risk factor for many chronic disorders, such as diabetes, CHD, cancer, and neuromuscular and skeletal disorders [6,24,25]. Epidemiologic studies have identified associations of both overweight and obesity with higher all-cause mortality in the general population [3,26]. However, an obesity paradox has been reported in which overweight and obesity is associated with a decreased risk for morbidity and mortality among patients with established CHD [18,19]. Similar findings have been described in patients with heart failure, hypertension, or atrial fibrillation [27,28]. Reasons for these findings remain in debate. In our study, although overweight and obese patients had higher prevalence of diabetes and hypertension, they were younger and were more likely to receive target lesion revascularization and medication therapies, similar to the literature [29]. Several studies supposed that these characteristics might be an explanation for the “protective” effect of obesity [29,30]. Furthermore, previous studies reported that metabolic abnormality was closer to CVD morbidity and mortality compared to obesity [31,32]. We speculate that obesity increases risks for chronic diseases morbidity but its effect on adverse outcomes would be attenuated after chronic diseases have been developed. Subsequently, the protective effects of overweight, such as nutritional reserve and stores of lean mass, may prevail over its adverse effects on cardiovascular outcomes in patients with established CHD in our study. Although weight loss is recommended in obese patients [9], even for those with established CHD [10,12], the question of whether such change in body weight is beneficial to health is controversial [15,16]. Several studies have shown that weight loss increases in cardiorespiratory fitness and improves in insulin resistance and related cardiometabolic risk factors [14,28], which may suggest favorable prognostic effects in patients with established CHD [11]. However, observations from recent studies and the present study show that the long-term clinical outcomes are not as expected [16,20,33]. Compared with the previous studies, our study population is larger, which helped us firstly analysed of the effect of changes in BMI levels on prognosis in patients with CHD. Interestingly, decreases in BMI levels contributed to increases in adverse outcomes, regardless of baseline BMI levels. Previous study reported that weight loss interventions, such as exercise and caloric restriction, might lead to loss of muscle mass and result in impaired physical function [34,35]. We speculate that
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
109
Fig. 2. Stratified analyses of associations between weight change and all-cause mortality. Scales on the y-axis are log transformed and represent hazard ratios and 95% confidence interval. Age, sex, height, current smoking, hypertension, diabetes, stroke, family history of coronary disease, previous PCI, previous CABG, medication use, serum creatinine, high-sensitivity CRP, and NT-proBNP at baseline were adjusted. The covariate was not included in models when the participants were stratified by this covariate.
these adverse effects may cause the increases in risks for adverse outcomes in patients with CHD. However, our study was focused on cardiovascular outcomes and all-cause death. We cannot rule out the possibility of a different effect of weight loss on other health outcomes in patients with CHD. Finally, we also found that risks for cardiovascular events and mortality increased with the increase of weight gain, which was in line with the literature [36,37]. Although the association between small weight gain and adverse outcomes was not significant in our study, it is now widely accepted that moderate or severe weight gain results in adipocyte hypertrophy and an increased in obesityrelated disorders, which cause the increasing risks for cardiovascular outcomes [38]. Furthermore, smoking cessation is recommended after CHD. Smoking cessation, however, is known to cause
weight gain [39]. Although our stratified analyses showed that large weight gain increased mortality risk both in non- or exsmokers and in current smokers, further investigations are needed to elucidate the cardiovascular benefits achieved by smoking cessation during the follow-up and the untoward effects of weight gain in patients with CHD. The present study has several limitations. First, the duration of obesity and weight change over time before the observational period are important confounders for the assessment of obesityrelated adverse prognosis, but our study lacks these data, which may have caused to misestimate the impact of obesity or weight change on adverse outcome risks. Second, although the measurement of weight change in our study was also used in many previous studies [20,40,41], weight change may not well reflect weight
110
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
fluctuation during the follow-up. Weight fluctuation was reported to be a better marker in terms of assessment of adverse outcomes risks [15,36]. Third, the effects on outcomes between intentional and unintentional weight changes were reported to be different [16], but our study was not able to differentiate between intentional and unintentional weight changes. We speculate that a majority of patients with CAD have an intention for controlling their weight based on suggestions from their physicians. Fourth, metabolic dysfunction is known to be independently associated with CHD risk and may play an important role in heterogeneous effects of obesity on prognosis. Although metabolic abnormalities, such as hypertension and diabetes, were adjusted in our multivariate regression models, their roles in the effects of obesity on adverse outcomes need further investigation in patients with CHD. Finally, other measurements of adiposity, such as waist circumference and body fat percentage could not be obtained in our study. The literature reported that high body fat percentage was associated with increased mortality even in those with low BMI [42], suggesting that other measurements of adiposity may help explain the heterogeneous effects of obesity on adverse outcomes. In conclusion, the present study found that obesity did not increase risks for cardiovascular events and mortality in patients with CHD and that risks for adverse cardiovascular outcomes increased with the increase of weight loss or weight gain. Our results suggest that the recommendation of maintaining a normal body weight for patients with CHD should be reconsidered. Maintaining a stable weight may be a better strategy for the reduction of risks for cardiovascular outcomes and all-cause death in this population, regardless of baseline body weight. Further investigations on heterogeneous effects of obesity and weight change are needed.
[5]
[6]
[7]
[8]
[9] [10]
[11] [12]
[13]
[14] [15] [16]
[17]
Conflicts of interest [18]
The authors declared they do not have anything to disclose regarding conflict of interest with respect to this manuscript. Author contributions Sheng-Yong Dong and Qiang Zeng had full access to all the data in the study and designed the study. Shuang-Tong Yan drafted the manuscript. Man-Liu Wang, Zhi-Bing Li, and Lian-Qing Fang collected the data. Sheng-Yong Dong and Qiang Zeng analyzed data and revised the manuscript.
[19]
[20]
[21]
[22]
Acknowledgements The authors thank all participants for their important contributions.
[23]
Appendix A. Supplementary data
[24]
Supplementary data related to this article can be found at https://doi.org/10.1016/j.atherosclerosis.2018.05.007.
[25] [26]
References [1] G.A. Roth, C. Johnson, A. Abajobir, et al., Global, regional, and national burden of cardiovascular diseases for 10 causes, 1990 to 2015, J. Am. Coll. Cardiol. 70 (1) (2017) 1e25. [2] G.A. Roth, M.D. Huffman, A.E. Moran, et al., Global and regional patterns in cardiovascular mortality from 1990 to 2013, Circulation 132 (17) (2015) 1667e1678. [3] GBD 2015 Obesity Collaborators, A. Afshin, M.H. Forouzanfar, et al., Health effects of overweight and obesity in 195 countries over 25 years, N. Engl. J. Med. 377 (1) (2017) 13e27. [4] NCD Risk Factor Collaboration (NCD-RisC), Trends in adult body-mass index in 200 countries from 1975 to 2014: a pooled analysis of 1698 population-based
[27]
[28]
[29]
[30]
measurement studies with 19.2 million participants, Lancet 387 (10026) (2016) 1377e1396. GBD 2016 Risk Factors Collaborators, Global, regional, and national comparative risk assessment of 84 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016, Lancet 390 (10100) (2017) 1345e1422. C. Lassale, I. Tzoulaki, K.G.M. Moons, et al., Separate and combined associations of obesity and metabolic health with coronary heart disease: a panEuropean case-cohort analysis, Eur. Heart J. 39 (5) (2018) 397e406. A.J. Flint, F.B. Hu, R.J. Glynn, et al., Excess weight and the risk of incident coronary heart disease among men and women, Obesity (Silver Spring) 18 (2) (2010) 377e383. K. Bibbins-Domingo, P. Coxson, M.J. Pletcher, et al., Adolescent overweight and future adult coronary heart disease, N. Engl. J. Med. 357 (23) (2007) 2371e2379. B.M. Wolfe, E. Kvach, R.H. Eckel, Treatment of obesity: weight loss and bariatric surgery, Circ. Res. 118 (11) (2016) 1844e1855. J. Perk, G. De Backer, H. Gohlke, et al., European guidelines on cardiovascular disease prevention in clinical practice (version 2012). The fifth joint task force of the European society of cardiology and other societies on cardiovascular disease prevention in clinical practice (constituted by representatives of nine societies and by invited experts), Eur. Heart J. 33 (13) (2012) 1635e1701. P.A. Ades, P.D. Savage, Potential benefits of weight loss in coronary heart disease, Prog. Cardiovasc. Dis. 56 (4) (2014) 448e456. S.C. Smith Jr., E.J. Benjamin, R.O. Bonow, et al., AHA/ACCF secondary prevention and risk reduction therapy for patients with coronary and other atherosclerotic vascular disease: 2011 update: a guideline from the american heart association and american college of cardiology foundation, Circulation 124 (22) (2011) 2458e2473. C.J. Lavie, R.V. Milani, H.O. Ventura, Obesity and cardiovascular disease: risk factor, paradox, and impact of weight loss, J. Am. Coll. Cardiol. 53 (21) (2009) 1925e1932. T.D. Adams, L.E. Davidson, S.E. Litwin, et al., Weight and metabolic outcomes 12 Years after gastric bypass, N. Engl. J. Med. 377 (12) (2017) 1143e1155. S. Bangalore, R. Fayyad, R. Laskey, et al., Body-weight fluctuations and outcomes in coronary disease, N. Engl. J. Med. 376 (14) (2017) 1332e1340. Q.R. Pack, J.P. Rodriguez-Escudero, R.J. Thomas, et al., The prognostic importance of weight loss in coronary artery disease: a systematic review and metaanalysis, Mayo Clin. Proc. 89 (10) (2014) 1368e1377. J. Sierra-Johnson, A. Romero-Corral, V.K. Somers, et al., Prognostic importance of weight loss in patients with coronary heart disease regardless of initial body mass index, Eur. J. Cardiovasc. Prev. Rehabil. 15 (3) (2008) 336e340. J. Niedziela, B. Hudzik, N. Niedziela, et al., The obesity paradox in acute coronary syndrome: a meta-analysis, Eur. J. Epidemiol. 29 (11) (2014) 801e812. C.J. Lavie, A. De Schutter, D. Patel, et al., Body composition and coronary heart disease mortalityean obesity or a lean paradox? Mayo Clin. Proc. 86 (9) (2011) 857e864. F. Lopez-Jimenez, C.O. Wu, X. Tian, et al., Weight change after myocardial infarctionethe enhancing recovery in coronary heart disease patients (ENRICHD) experience, Am. Heart J. 155 (3) (2008) 478e484. B.F. Zhou, Cooperative Meta-Analysis Group of the Working Group on Obesity in China, Predictive values of body mass index and waist circumference for risk factors of certain related diseases in Chinese adultsestudy on optimal cutoff points of body mass index and waist circumference in Chinese adults, Biomed. Environ. Sci. 15 (1) (2002) 83e96. C.P. Cannon, R.G. Brindis, B.R. Chaitman, et al., 2013 ACCF/AHA key data elements and definitions for measuring the clinical management and outcomes of patients with acute coronary syndromes and coronary artery disease: a report of the American College of Cardiology Foundation/American Heart Association Task Force on clinical data standards (writing committee to develop acute coronary syndromes and coronary artery disease clinical data standards), J. Am. Coll. Cardiol. 61 (9) (2013) 992e1025. K. Aho, P. Harmsen, S. Hatano, et al., Cerebrovascular disease in the community: results of a WHO collaborative study, Bull. World Health Organ. 58 (1) (1980) 113e130. R. Caleyachetty, G.N. Thomas, K.A. Toulis, et al., Metabolically healthy obese and incident cardiovascular disease events among 3.5 million men and women, J. Am. Coll. Cardiol. 70 (12) (2017) 1429e1437. C.E. Ndumele, L. Cobb, M. Lazo, et al., Weight history and subclinical myocardial damage, Clin. Chem. 64 (1) (2018) 201e209. Global BMI Mortality Collaboration, E. Di Angelantonio, N. Bhupathiraju Sh, et al., Body-mass index and all-cause mortality: individual-participant-data meta-analysis of 239 prospective studies in four continents, Lancet 388 (10046) (2016) 776e786. S.M. Artham, C.J. Lavie, R.V. Milani, et al., Obesity and hypertension, heart failure, and coronary heart disease-risk factor, paradox, and recommendations for weight loss, Ochsner J. 9 (3) (2009) 124e132. C.J. Lavie, A. Pandey, D.H. Lau, et al., Obesity and atrial fibrillation prevalence, pathogenesis, and prognosis: effects of weight loss and exercise, J. Am. Coll. Cardiol. 70 (16) (2017) 2022e2035. A. Oreopoulos, F.A. McAlister, K. Kalantar-Zadeh, et al., The relationship between body mass index, treatment, and mortality in patients with established coronary artery disease: a report from APPROACH, Eur. Heart J. 30 (21) (2009) 2584e2592. B.A. Steinberg, C.P. Cannon, A.F. Hernandez, et al., Medical therapies and
S.-Y. Dong et al. / Atherosclerosis 274 (2018) 104e111
[31]
[32]
[33]
[34] [35]
invasive treatments for coronary artery disease by body mass: the "obesity paradox" in the Get with the Guidelines database, Am. J. Cardiol. 100 (9) (2007) 1331e1335. Global Burden of Metabolic Risk Factors for Chronic Diseases Collaboration (BMI Mediated Effects), Y. Lu, K. Hajifathalian, et al., Metabolic mediators of the effects of body-mass index, overweight, and obesity on coronary heart disease and stroke: a pooled analysis of 97 prospective cohorts with 1.8 million participants, Lancet 383 (9921) (2014) 970e983. Q. Zeng, S.Y. Dong, M.L. Wang, et al., Obesity and novel cardiovascular markers in a population without diabetes and cardiovascular disease in China, Prev. Med. 91 (2016) 62e69. T. Moholdt, C.J. Lavie, J. Nauman, Sustained physical activity, not weight loss, associated with improved survival in coronary heart disease, J. Am. Coll. Cardiol. 71 (10) (2018) 1094e1101. B. Cao, Estimating the effects of obesity and weight change on mortality using a dynamic causal model, PLoS One 10 (6) (2015), e0129946. E.M. Weinheimer, L.P. Sands, W.W. Campbell, A systematic review of the separate and combined effects of energy restriction and exercise on fat-free mass in middle-aged and older adults: implications for sarcopenic obesity,
111
Nutr. Rev. 68 (7) (2010) 375e388. [36] E. Yu, S.H. Ley, J.E. Manson, et al., Weight history and all-cause and causespecific mortality in three prospective cohort studies, Ann. Intern. Med. 166 (9) (2017) 613e620. [37] K.F. Adams, M.F. Leitzmann, R. Ballard-Barbash, et al., Body mass and weight change in adults in relation to mortality risk, Am. J. Epidemiol. 179 (2) (2014) 135e144. [38] L.K. Forsythe, J.M. Wallace, M.B. Livingstone, Obesity and inflammation: the effects of weight loss, Nutr. Res. Rev. 21 (2) (2008) 117e133. [39] P. Dehghani, B. Habib, S.B. Windle, et al., Smokers and postcessation weight gain after acute coronary syndrome, J Am Heart Assoc 6 (4) (2017). [40] Z.B. Li, M.L. Wang, S.Y. Dong, et al., Effects of body mass index and weight change on mortality in older men with impaired glucose regulation, Exp. Gerontol. 89 (2017) 87e92. [41] S.Y. Park, L.R. Wilkens, G. Maskarinec, et al., Weight change in older adults and mortality: the Multiethnic Cohort Study, Int. J. Obes. 42 (2) (2018) 205e212. [42] R. Padwal, W.D. Leslie, L.M. Lix, et al., Relationship among body fat percentage, body mass index, and all-cause mortality: a cohort study, Ann. Intern. Med. 164 (8) (2016) 532e541.