- Email: [email protected]
Impact of Gait Speed on the Obesity Paradox in Older Patients With Cardiovascular Disease
ARTICLE IN PRESS CLINICAL RESEARCH STUDY
Impact of Gait Speed on the Obesity Paradox in Older Patients With Cardiovascular Disease Takeshi Nakamura, PT, MSca, Kentaro Kamiya, PT, PhDa,b, Atsuhiko Matsunaga, PT, PhDa,b, Nobuaki Hamazaki, PT, PhDc, Ryota Matsuzawa, PT, PhDd, Kohei Nozaki, PT, MScc, Masashi Yamashita, PT, MSca, Emi Maekawa, MD, PhDe, Chiharu Noda, MD, PhDe, Minako Yamaoka-Tojo, MD, PhDa,b, Junya Ako, MD, PhDe a
Department of Rehabilitation Sciences, Kitasato University Graduate School of Medical Sciences, Sagamihara, Japan; bDepartment of Rehabilitation, School of Allied Health Sciences, Kitasato University Sagamihara, Japan; cDepartment of Rehabilitation, Kitasato University Hospital, Sagamihara, Japan; dDepartment of Physical Therapy, School of Rehabilitation, Hyogo University of Health Sciences, Sagamihara, Japan; eDepartment of Cardiovascular Medicine, Kitasato University School of Medicine, Sagamihara, Japan.
ABSTRACT PURPOSE: The purpose of this study was to determine whether gait speed affects the obesity paradox in older patients with cardiovascular disease. METHODS: The study population consisted of 2224 patients ≥ 60 years old with cardiovascular disease admitted to hospital between May 1, 2006, and January 31, 2018. Body mass index (BMI) and gait speed before hospital discharge were determined, and patients were divided into two groups: slow and preserved gait speed (≤0.8 and > 0.8 m/s, respectively), according to the algorithm for sarcopenia diagnosis. The slow and preserved gait speed groups were also further subdivided according to BMI: < 18.5 kg/m2, 18.5−24.9 kg/m2, and BMI ≥ 25.0 kg/m2. The study endpoint was all-cause mortality. RESULTS: The study population (male: 66.7%) had a mean age of 73.1 § 7.6 years. Over a median followup period of 1.69 years (interquartile range 0.67−3.67 years), 283 patients died. Higher BMI was associated with favorable prognosis in the group with preserved gait speed but not in the group with slow gait speed after adjusting for other prognostic factors. Adding BMI to the clinical model significantly increased the area under the receiver operating characteristic curve in the group with preserved gait speed (0.744 vs 0.726, P = 0.028) but not in the group with slow gait speed (0.716 vs 0.716, P = 0.789). CONCLUSIONS: Higher BMI was consistently associated with favorable prognosis in patients with cardiovascular disease and preserved gait speed but not in those with slow gait speed. These findings indicated that physical frailty influences the obesity paradox in older patients with cardiovascular disease. Ó 2019 Elsevier Inc. All rights reserved. The American Journal of Medicine (2019) 000:1-8 KEYWORDS: Cardiovascular disease; Gait speed; Obesity paradox; Older; Prognosis
INTRODUCTION
Funding: This study was supported by a Grant for Clinical and Epidemiologic Research of the Joint Project of Japan Heart Foundation and the Japanese Society of Cardiovascular Disease Prevention Sponsored by AstraZeneca. Conflicts of Interest: None. Author contributions: All authors had access to the data and a role in writing this manuscript. Requests for reprints should be addressed to, Kentaro Kamiya, PT, PhD, Department of Rehabilitation, Kitasato University School of Allied Health Sciences, 1-15-1 Kitasato, Minami-ku, Sagamihara, Kanagawa, 252-0375, Japan E-mail address: [email protected]
0002-9343/© 2019 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.amjmed.2019.06.047
Although high body mass index (BMI) is known to be associated with elevated risk of cardiovascular disease,1,2 patients who are overweight and obese show better survival prognosis following the development of cardiovascular disease, including heart failure and coronary artery disease, in a phenomenon known as the “obesity paradox.”3,4 In contrast, high BMI is also a risk factor for both disability and falls5,6; therefore, high BMI does not necessarily have a protective effect against health disorders in older people with reduced physical function.7 The obesity paradox was reported in middle-aged and older patients with cardiovascular disease with preserved physical function.1,8−10 However, it is unclear whether the
2
ARTICLE IN PRESS
The American Journal of Medicine, Vol 000, No 000, && 2019
obesity paradox is present in older patients with cardiovascular disease with reduced physical function, such as those with slow gait speed. The present study was performed to determine whether the obesity paradox is also present in older patients with cardiovascular disease with slow gait speed.
METHODS Study Population This study involved a retrospective review of 2224 consecutive patients ≥60 years of age who were admitted for cardiovascular disease to the Cardiovascular Center of Kitasato University Hospital between May 1, 2006, and January 31, 2018, in all of whom usual gait speed had been evaluated during the hospitalization. The study was performed in accordance with the tenets of the Declaration of Helsinki and the protocol received approval from the Ethics Committee of Kitasato University Hospital.
Usual Gait Speed Measurement
Usual gait speed was determined by timing the patients walking at normal speed over the middle 10 m of a 16-m walkway. The use of walking aids, such as a cane, was permitted during the test. Patients were divided into slow gait speed (≤0.8 m/s) and preserved gait speed (>0.8 m/s) groups according to the European Working Group on Sarcopenia in Older People and Asian Working Group for Sarcopenia algorithm for diagnosis of CLINICAL SIGNIFICANCE sarcopenia.13,14
Higher body mass index (BMI) was consistently associated with favorable prognosis in patients with cardiovascular disease and preserved gait speed. In contrast, BMI showed no association with prognosis in the group with slow gait speed. Physical frailty, such as slow gait speed, may influence the obesity paradox in older patients with cardiovascular disease.
Data Collection Data on all variables were obtained from the patients’ electronic medical records. We recorded the clinical details for all patients, including medication use, comorbidities, demographic characteristics, and both echocardiographic and biochemical data immediately before hospital discharge. We defined diagnosis of heart failure before admission to hospital as prior heart failure. A stadiometer was used to measure height to the nearest 0.1 cm, and a calibrated weighing scale was used to measure weight to the nearest 0.1 kg at hospital discharge. BMI was calculated according to the formula: BMI = body weight (kg)/height (m2). Patients were divided according to the BMI categories recommended by the Japan Society for the Study of Obesity as follows: BMI < 18.5 kg/m2 (underweight), BMI 18.5−24.9 kg/m2 (normal), BMI ≥ 25.0 kg/m2 (obesity).11 A commercially available immunoradiometric assay was used to measure B-type natriuretic peptide (BNP) concentration (Shionogi, Osaka, Japan).The estimated glomerular filtration rate (eGFR) was determined using the Japanese Society of Nephrology formula: men = 194 £ (serum creatinine)1.094 £ (age)0.287; women = eGFR (male) £ 0.739.12 Left ventricular ejection fraction was estimated by Simpson’s method on two-dimensional echocardiographs. All-cause mortality was used as the study end point, with the time to the end point calculated as the period (in days) between the date of usual gait speed determination and the event. Follow-up was carried out according to hospital discharge and ended with the last clinical evaluation or with the patients’ death. If a patient died outside the hospital where they were being followed up, medical records of the event and a report of death were considered.
Statistical Analysis
Continuous variables are expressed as the mean § standard deviation, and non-normally distributed variables are presented as the median (interquartile range). Categorical variables are expressed as frequencies and percentages. Multivariate normal imputation for missing values was performed using JMP 13.2 software (SAS Institute Inc., Cary, NC).15,16 The following variables were incorporated into the imputation model: age, sex, BMI, diagnosis, eGFR, log BNP, albumin, hemoglobin, left ventricular ejection fraction, smoking, hypertension, diabetes mellitus, dyslipidemia and usual gait speed, and prior heart failure. We compared the baseline characteristics by one-way analysis of variance, Kruskal−Wallis test, unpaired Student’s t-test, or Mann−Whitney U test for continuous variables and chi-squared test or Fisher exact test for dichotomous variables, as appropriate. The cumulative incidence of mortality during follow-up period was calculated according to BMI categories based on the Kaplan−Meier curves in the slow gait speed and preserved gait speed groups. Intergroup differences were examined for significance by the log-rank test. Univariate and multivariate Cox regression analyses were used to evaluate the prognostic capabilities of BMI categories. Age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure were included in multivariate analyses. The associations between BMI and mortality risk were examined using Cox regression model adjusted for age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure with spline functions. In addition, to examine the potential effect modification on the association of BMI categories with mortality, we performed subgroup analyses of BMI categories with adjustment for age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure in various subgroups, including history of heart failure, ischemic heart disease, and cardiac surgery, age (stratified at 75 years), sex, and left ventricular ejection fraction (stratified at 50%). To avoid overfitting, all potential confounding factors of BMI, which include age, sex,
Nakamura et al
ARTICLE IN PRESS
Gait Speed and Obesity Paradox in Older Cardiac Patients
eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure, were reduced to one composite characteristic by applying a propensity score in subgroup analyses.17 We examined whether BMI had complementary predictive capability to age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure (clinical model) by constructing receiver operating characteristic (ROC) curves for all-cause mortality using the clinical model only and clinical model + BMI. The areas under the curve (AUC) were compared according to the method of DeLong et al.18 An AUC between 0.7 and 0.8 is considered acceptable discrimination, between 0.8 and 0.9 is considered excellent discrimination, and greater than 0.9 is considered outstanding discrimination.19 The continuous net reclassification improvement (cNRI) and integrated discrimination improvement (IDI)—sensitive statistical methods to quantify improvement of a model when a new variable is added—were used to calculate the increase in information by each BMI component compared to the prognostic factors.20 Two-tailed P <0.05 was taken to indicate statistical significance. Analyses were performed using R version 3.1.2 (R Foundation for Statistical Computing, Vienna, Austria) and JMP Pro 13 (SAS Institute Inc., Cary, NC).
RESULTS Study Population Baseline characteristics for all participants and for groups stratified according to gait speed and BMI are shown in Table 1 and Supplemental Table S1 (available online). The study population had a mean age of 73 § 8 years, and the gender distribution was 67% male and 33% female. Patients were hospitalized as a result of heart failure (35.1%), coronary artery disease (25.9%), cardiac surgery (21.7%), and other clinical entities (18.1%). Seventy-two percent of the patients were prescribed beta-blockers at discharge and 68% received angiotensin-converting enzyme inhibitors or angiotensin receptor blockers. Baseline characteristics for all participants and for groups stratified according to gait speed and BMI are shown in Table 1 and Supplemental Table S1. The study population had a mean age of 73 § 8 years, and the gender distribution was 67% male and 33% female. Patients were hospitalized as a result of heart failure (35.1%), coronary artery disease (25.9%), cardiac surgery (21.7%), and other clinical entities (18.1%). Seventy-two percent of the patients were prescribed beta-blockers at discharge and 68% received angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs).
Associations of BMI With All-Cause Mortality in Slow and Preserved Gait Speed Groups A total of 283 deaths occurred during the follow-up period (median 1.69 years, interquartile range 0.67-3.67 years, 49.5 deaths per 1000 person-years). The Kaplan−Meier
3
survival curves showed that higher BMI was associated with favorable prognosis in the preserved gait speed group (Figure 1A; log-rank P <0.001). In contrast, no such associations were observed between BMI categories and prognosis in the slow gait speed group (Figure 1B; log-rank P = 0.126). Table 2 shows the results of univariate and multivariate Cox regression analyses for all-cause mortality. On Cox regression analyses, the hazard ratio in BMI < 18.5 and BMI ≥25.0 in the preserved gait speed group were 2.06 (95% confidence interval [CI]: 1.40−3.03, P < 0.001) and 0.55 (95% CI: 0.35−0.88, P = 0.013), respectively, compared to BMI 18.5−24.9 after adjustment for the effects of age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure. On Cox regression analyses, the hazard ratio in BMI < 18.5 and BMI ≥ 25.0 in the preserved gait speed group were 2.06 (95% CI: 1.40−3.03, P <0.001) and 0.55 (95% CI: 0.35−0.88, P = 0.013), respectively, compared to BMI 18.5-24.9 after adjusting for the effects of age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure. On the other hand, in the fully adjusted model, the hazard ratio in BMI <18.5 and BMI ≥25.0 in the slow gait speed group were 1.04 (95% CI: 0.62−1.75, P = 0.875) and 0.68 (95% CI: 0.34−1.35, P = 0.271), respectively, compared to BMI 18.5−24.9 (Table 2). Figure 2 shows the association of BMI with mortality in slow and preserved gait speed groups. The CI of BMI after adjustment for age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure was wide at all BMI levels in the group with slow gait speed. In contrast, mortality risk decreased with increasing BMI in fully adjusted models in the group with preserved gait speed. Figure 3 shows that BMI < 18.5 was significantly associated with mortality compared to BMI 18.5-24.9 across various subgroups in the group with preserved gait speed after adjusting for age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure (Figure 3), except in patients who < 75 years old and with left ventricular ejection fraction ≥50. In contrast, BMI consistently showed no association with prognosis in the group with slow gait speed (Figure 3).
Additive Prognostic Predictive Capability of BMI to the Clinical Model As shown in Table 3, the logistic regression models of the clinical model only and clinical model + BMI were subjected to ROC curve analyses in the slow and preserved gait speed groups. The AUC on ROC curve analysis was greater with clinical model + BMI (0.744; 95% CI: 0.703−0.781) than the clinical model only (0.724; 95% CI: 0.684−0.765; P = 0.028) in the group with preserved gait speed. The addition of BMI to clinical model was also associated with significant increases in both cNRI (0.263; 95% CI: 0.114−0.413; P < 0.001) and IDI (0.012; 95% CI: 0.006−0.019; P <0.001) for all-cause mortality in the group with preserved gait speed. However, the addition of
ARTICLE IN PRESS
The American Journal of Medicine, Vol 000, No 000, && 2019
4 Table 1
Patient Characteristics
Variable
Missing Data
Overall
Preserved Gait Speed (>0.8 m/s)
P Value
P Value
Slow gait speed (≤ 0.8 m/s)
BMI < 18.5
BMI 18.5−24.9
BMI ≥ 25.0
BMI <18.5
BMI 18.5−24.9
BMI ≥ 25.0
(n = 112)
(n = 301)
(n = 71)
n (%)
(n = 2224)
(n = 209)
(n = 1156)
(n = 375)
Age (yrs) Male, n (%) Body weight (kg) Height (cm) BMI (kg/m2) SBP (mmHg) History of disease, n (%) Heart failure Ischemic heart disease Cardiac surgery Hypertension Diabetes mellitus Dyslipidemia Prior heart failure Smoking, n (%) LVEF (%) eGFR (mL/min/1.73 m2) B-type natriuretic peptide (pg/mL)
0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 57 (2.5)
73.1§7.6 1484 (66.7) 56.9 §11.5 159.8§8.9 22.2§3.7 121 §26
73.2§7.0 104 (49.8) 43.2§5.6 158.6§8.7 17.1 §1.2 116 §27
71.7 §6.8 874 (75.6) 57.2 §7.3 158.6 §8.7 21.8 §1.7 122 § 27
70.1§6.4 269 (71.7) 70.9§9.4 160.8§8.5 27.4§2.4 126§ 29
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001
79.8 §8.0 41 (36.6) 39.6 §5.6 154.3 §9.4 16.6 §1.3 112 §29
79.0§7.3 158 (52.5) 51.9§7.7 155.2§9.5 21.5§1.8 118§ 25
77.4§6.5 38 (53.5) 69.1§11.1 155.6§9.6 28.5§3.7 119 §20
0.108 0.011 <0.001 0.590 <0.001 0.074
0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 69 (3.1) 188 (8.5) 11 (0.1) 215 (9.7)
493 (42.6) 659 (57.0) 511 (44.2) 801 (69.3) 392 (33.9) 585 (50.6) 253 (21.9) 735 (63.6) 54.2 §14.2 52.7 §19.6 168.1 [71.6, 352.8] 3.6 §0.5 12.1 §2.0 96 § 30 1.14 §0.21 128 (11.1)
159 (42.4) 131 (34.9) 144 (38.4) 292 (77.9) 178 (47.5) 239 (63.7) 80 (21.3) 241 (64.3) 55.2 §13.4 53.7§18.2 140.2 [66.5, 277.8] 3.7§0.4 12.5§1.9 91 §30 1.13§0.20 21 ( 5.6)
74 (66.1) 41 (36.6) 32 (28.6) 66 (58.9) 23 (20.5) 32 (28.6) 43 (38.4) 49 (43.8) 52.0 §15.9 51.2 §25.4 367.9 [123.3, 831.9] 3.2 §0.5 11.1 §1.9 95 §34 0.60 §0.15 21 (18.8)
183 (60.8) 143 (47.5) 122 (40.5) 221 (73.4) 121 (40.2) 135 (44.9) 94 (31.2) 137 (45.5) 53.4§15.8 45.6§22.0 326.6 [156.0, 724.5] 3.3§0.5 11.1§1.8 89 §33 0.60§0.15 66 (21.9)
48 (67.6) 46 (64.8) 28 (39.4) 56 (78.9) 36 (50.7) 38 (53.5) 23 (32.4) 39 (54.9) 55.4§13.3 41.6§16.1 229.3 [122.1, 468.0] 3.3§0.5 11.5§1.8 89 §30 0.59§0.14 10 (14.1)
0.422 0.001 0.078 0.004 <0.001 0.001 0.385 0.289 0.322 0.011 0.034
15 (0.1) 12 (0.1) 248 (11.2) 0 (0.0) 0 (0.0)
124 (59.3) 71 (33.9) 92 (44.0) 106 (50.7) 42 (20.1) 58 (27.8) 58 (27.8) 102 (48.8) 52.4 §15.3 58.2§21.9 211.7 [107.3, 509.6] 3.4§0.5 11.5§1.7 98 §32 1.10§0.20 37 (17.7)
< 0.001 < 0.001 0.135 < 0.001 < 0.001 < 0.001 0.146 < 0.001 0.061 0.001 < 0.001
Albumin (g/dL) Hemoglobin (g/dL) LDL cholesterol (mg/dL) Gait speed (m/sec) All-cause mortality, n (%)
1081 (48.6) 1204 (54.1) 929 (41.8) 1542 (69.3) 792 (35.6) 1087 (48.9) 551 (24.8) 1303 (58.6) 53.8 §14.3 52.0§20.5 210.1 [92.9, 409.1] 3.5§0.5 11.9§2.0 94§31 1.02§0.29 283 (12.7)
< 0.001 < 0.001 0.024 0.043 < 0.001
0.110 0.148 0.342 0.947 0.308
Values are the means § SD, median [interquartile range], or number (%). BMI = body mass index; eGFR = estimated glomerular filtration rate; LDL = lowdensity lipoprotein; LVEF = left ventricular ejection fraction; SBP = systolic blood pressure.
BMI to the clinical model did not improve the AUC, cNRI, or IDI in the group with slow gait speed.
DISCUSSION The results of the present study indicated that higher BMI was consistently associated with favorable prognosis, and BMI showed complementary prognostic predictive capability to preexisting prognostic risk factors in patients with cardiovascular disease and preserved gait speed but not in those with slow gait speed. These results suggest that there is no obesity paradox in patients with cardiovascular disease and slow gait speed. Most previous studies suggested that there was obesity paradox in patients with cardiovascular disease, including heart failure and coronary artery disease.21−23 Although some reports suggested that the obesity paradox could be explained by several factors, such as body composition,24 exercise capacity,8,10 chronic inflammation,25 or nutritional status,26 it was unclear whether there was an obesity paradox in older patients with cardiovascular disease and lower levels of physical function. Gait speed is well established as a reliable, valid, sensitive, and specific measurement, and slow gait is a core component of frailty and sarcopenia.13,27 Therefore, measurement of gait speed has been widely used
to evaluate physical function in older adults. In this study, we measured gait speed as a means of determining physical function and investigated whether gait speed affected the obesity paradox in older patients with cardiovascular disease. To our knowledge, this is first report indicating differences in prognostic value of BMI between patients with cardiovascular disease and preserved gait speed compared to those with slow gait speed. Gait speed could be a useful indicator of mortality risk and is a potential therapeutic target because gait speed is an independent predictor of operative mortality,28,29 and improvement in gait speed is associated with reduced rates of rehospitalization and mortality in patients with cardiovascular disease.30 There are several possible underlying mechanisms for the association between BMI and prognosis in patients with cardiovascular disease and slow gait speed. Body composition may affect prognosis in patients with cardiovascular disease and slow gait speed. Some studies showed that gait speed was positively correlated to muscle mass and negatively correlated to fat mass in community-dwelling older adults31−34 and in patients with heart failure.35 Recently, the presence of both muscle wasting and high fat mass has been defined as sarcopenic obesity, which is related to lower limb muscle function, inflammation, or physical activity. Therefore, it is related to poorer outcome
Nakamura et al
ARTICLE IN PRESS
Gait Speed and Obesity Paradox in Older Cardiac Patients
5
Figure 1 Kaplan-Meier curves for all-cause mortality according to body mass index (BMI) categories in the preserved and slow gait speed groups.
compared to muscle wasting alone. Moreover, several studies suggested that mortality risk in subjects with higher BMI was likely the result of the high risk associated with fat mass, and increased fat mass did not have a protective effect against mortality risk.24,36 These studies suggested that patients with cardiovascular disease and a high BMI, especially those with slow gait speed, may have both low muscle mass and high fat mass, which would adversely affect their prognosis. Physical inactivity represents another possible mechanism underlying the relation between BMI and prognosis in patients with slow gait speed. The act of walking uses energy and places demands on multiple organ systems, including the heart, lungs, circulatory system, nervous system, and musculoskeletal system. A slow gait may reflect the high energy costs of walking and indicate damage to these systems, which may be responsible for reduced levels of physical inactivity in the older patient. Reduced physical Table 2
activity itself is known to be a strong prognostic factor,37,38 and therefore, slow gait speed may be associated with poor prognosis despite BMI in patients with cardiovascular disease. On the other hand, the results of the present study showed that higher BMI was associated better prognosis in patients with cardiovascular disease and preserved gait speed. Muscle mass showed a consistent protective effect in community-dwelling subjects and patients with cardiovascular disease. Previous studies suggested that muscle mass had a protective effect against mortality risk during prolonged illness by providing a reliable protein reserve39 or the metabolic effect of changes in skeletal muscle.40 These studies suggested that patients with cardiovascular disease, high BMI, and preserved gait speed may have high muscle mass and may, therefore, have good prognosis. Although several recent studies regarding the obesity paradox suggested that increasing or preserving body
Univariariate and Multivariate Cox Regression Analysis for All-Cause Mortality
Variable
Preserved gait speed BMI < 18.5 BMI 18.5−24.9 BMI ≥ 25.0 Slow gait speed BMI < 18.5 BMI 18.5−24.9 BMI ≥ 25.0
Univariate Cox Analysis
Multivariate Cox Analysis*
HR
95%CI
P Value
HR
95%CI
P Value
2.01 1.00 0.47
1.39−2.90 (Reference) 0.29−0.74
< 0.001 (Reference) 0.001
2.06 1.00 0.55
1.40−3.03 (Reference) 0.35−0.88
< 0.001 (Reference) 0.013
1.12 1.00 0.54
0.69−1.84 (Reference) 0.28−1.04
0.648 (Reference) 0.067
1.04 1.00 0.68
0.62−1.75 (Reference) 0.34−1.35
0.875 (Reference) 0.271
P values represent pairwise relationships relative to reference group. BMI = body mass index; CI = confidence interval; HR = hazard ratio. *Adjusted HR after adjustment for age, sex, estimated glomerular filtration rate, log B-type natirutic peptide, albumin, hemoglobin, left ventricular ejection fraction, smoking, and prior heart failure
6
ARTICLE IN PRESS
The American Journal of Medicine, Vol 000, No 000, && 2019
Figure 2 Associations of body mass index (BMI) with all-cause mortality in the preserved and slow gait speed groups. Notes: Dotted lines represent the 95% confidence intervals. Rug plots are shown along the x-axes of the graphs to depict the distributions of BMI. The adjusted models were adjusted for age, sex, estimated glomerular filtration rate, log B-type natriuretic peptide, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure.
weight should be recommended for older patients with cardiovascular disease,41,42 frailty status was not evaluated in these studies. Therefore, it was unknown whether frailty status influenced the obesity paradox in patients with
cardiovascular disease. On the other hand, previous studies have indicated that slow gait speed is more closely associated with increased mortality than decreased body weight in patients with cardiovascular disease.43,44 Moreover,
Figure 3 Forest plot of hazard ratio for association of body mass index (BMI) with all-cause mortality in the preserved and slowgait-speed groups. *Adjusted for applying a propensity score calculated by age, sex, estimated glomerular filtration rate, log B-type natriuretic peptide, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure. CI = confidence interval; HR = hazard ratio; IHD = ischemic heart disease; LVEF = left ventricular ejection fraction.
Nakamura et al Table 3
ARTICLE IN PRESS
Gait Speed and Obesity Paradox in Older Cardiac Patients
7
Predictive Value Analyses for All-Cause Mortality AUC
Preserved gait speed Clinical model Clinical model + BMI Slow gait speed Clinical model Clinical model + BMI
95%CI
P Value
cNRI
0.726 0.684−0.765 (Reference) 0.744 0.703−0.781 0.028 0.263
95%CI
P Value
0.114−0.413
(Reference) < 0.001 0.012
IDI
95%CI
P Value
0.006−0.019
(Reference) < 0.001
0.716 0.656−0.769 (Reference) (Reference) (Reference) 0.716 0.656−0.770 0.789 -0.013 -0.235−0.208 0.907 -0.0001 -0.0006−0.0004 0.758
P values represent pairwise relationships relative to reference group. BMI was added to Clinical model containing age, sex, estimated glomerular filtration rate, log B-type natriuretic peptide, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure. AUC = areas under the curve; BMI = body mass index; CI = confidence interval; cNRI = continuous net reclassification improvement; IDI = integrated discrimination improvement.
another study indicated that slow gait speed outperformed other multicomponent frailty scales, including body weight assessment, for predicting mortality in patients with cardiovascular disease, even if it is a single measurement.45 These studies and our results indicated that BMI does not necessarily provide increased value in patients with cardiovascular disease in whom gait speed is already compromised.
Study Limitations This study had several limitations, the most important of which are its retrospective design, limited follow-up period, and small sample size in the group with slow gait speed. In addition, only patients in whom gait speed had been measured were included in the study. However, patients with missing data may have more severe illness than those with complete data, and missing data can cause biased estimates. Body composition was not measured directly by magnetic resonance imaging or dual-energy X-ray absorptiometry. Finally, all subjects included in this study were Asian, so further studies are required to determine the generalizability of the results to other populations.
CONCLUSIONS Higher BMI was consistently associated with favorable prognosis in patients with cardiovascular disease and preserved gait speed but not in those with slow gait speed. In addition, BMI showed complementary prognostic predictive capability to preexisting prognostic risk factors in patients with preserved gait speed but not in those with slow gait speed. These findings indicate that physical frailty may influence the obesity paradox in older patients with cardiovascular disease.
References 1. Lavie CJ, Alpert MA, Arena R, Mehra MR, Milani RV, Ventura HO. Impact of obesity and the obesity paradox on prevalence and prognosis in heart failure. JACC Heart Fail 2013;1:93–102. 2. Krauss RM, Winston M, Fletcher BJ, Grundy SM. Obesity: impact on cardiovascular disease. Circulation 1998;98:1472–6. 3. Gruberg L, Weissman NJ, Waksman R, et al. The impact of obesity on the short-term and long-term outcomes after percutaneous coronary intervention: the obesity paradox? J Am Coll Cardiol 2002;39: 578–84.
4. Clark AL, Fonarow GC, Horwich TB. Obesity and the obesity paradox in heart failure. Prog Cardiovasc Dis 2014;56:409–14. 5. Mitchell RJ, Lord SR, Harvey LA, Close JC. Obesity and falls in older people: mediating effects of disease, sedentary behavior, mood, pain and medication use. Arch Gerontol Geriatr 2015;60:52–8. 6. Kosar CM, Thomas KS, Gozalo PL, Mor V. Higher level of obesity Is associated with intensive personal care assistance in the nursing home. J Am Med Dir Assoc 2018;19:1015–9. 7. Yamashita M, Kamiya K, Matsunaga A, et al. Prognostic value of sarcopenic obesity estimated by computed tomography in patients with cardiovascular disease and undergoing surgery. J Cardiol 2019. https://doi.org/10.1016/j.jjcc.2019.02.010. 8. Lavie CJ, Cahalin LP, Chase P, et al. Impact of cardiorespiratory fitness on the obesity paradox in patients with heart failure. Mayo Clin Proc 2013;88:251–8. 9. Piepoli MF, Corra U, Veglia F, et al. Exercise tolerance can explain the obesity paradox in patients with systolic heart failure: data from the MECKI Score Research Group. Eur J Heart Fail 2016;18:545–53. 10. McAuley PA, Keteyian SJ, Brawner CA, et al. Exercise capacity and the obesity paradox in heart failure: The FIT (Henry Ford Exercise Testing) Project. Mayo Clin Proc 2018;93:701–8. 11. Examination Committee of Criteria for ’Obesity Disease’ in Japan; Japan Society for the Study of Obesity. New criteria for ’obesity disease’ in Japan. Circ J 2002;66:987–92. 12. Ando Y, Ito S, Uemura O, et al. CKD Clinical Practice Guidebook. The essence of treatment for CKD patients. Clin Exp Nephrol 2009;13: 191–248. 13. Chen LK, Liu LK, Woo J, et al. Sarcopenia in Asia: consensus report of the Asian Working Group for Sarcopenia. J Am Med Dir Assoc 2014;15:95–101. 14. Cruz-Jentoft AJ, Baeyens JP, Bauer JM, et al. Sarcopenia: European consensus on definition and diagnosis: Report of the European Working Group on Sarcopenia in Older People. Age Ageing 2010;39: 412–23. 15. Schafer J, Strimmer K. A shrinkage approach to large-scale covariance matrix estimation and implications for functional genomics. Stat Appl Genet Mol Biol 2005;4:Article32. 16. Koch E, Finne K, Eikrem O, et al. Transcriptome-proteome integration of archival human renal cell carcinoma biopsies enables identification of molecular mechanisms. Am J Physiol Renal Physiol 2019;316:F1053–67. 17. Rubin DB. Estimating causal effects from large data sets using propensity scores. Ann Intern Med 1997;127:757–63. 18. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837–45. 19. Hosmer DW, Lemeshow S. Applied Logistic Regression. 2nd ed. New York, NY: Wiley; 2000. 20. Pencina MJ, D’Agostino R.B. Sr., D’Agostino R.B. Jr., Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008;27:157–72 [discussion 207-212].
8
ARTICLE IN PRESS
21. Mariscalco G, Wozniak MJ, Dawson AG, et al. Body mass index and mortality among adults undergoing cardiac surgery: a nationwide study with a systematic review and meta-analysis. Circulation 2017;135:850–63. 22. Moholdt T, Lavie CJ, Nauman J. Interaction of physical activity and body mass index on mortality in coronary heart disease: data from the Nord-Trondelag Health Study. Am J Med 2017;130:949–57. 23. Kamiya K, Masuda T, Matsue Y, et al. Complementary role of arm circumference to body mass index in risk stratification in heart failure. JACC Heart Fail 2016;4:265–73. 24. Lee DH, Keum N, Hu FB, et al. Predicted lean body mass, fat mass, and all cause and cause specific mortality in men: prospective US cohort study. BMJ 2018;362:k2575. 25. Antonopoulos AS, Tousoulis D. The molecular mechanisms of obesity paradox. Cardiovasc Res 2017;113:1074–86. 26. Casas-Vara A, Santolaria F, Fernandez-Bereciartua A, GonzalezReimers E, Garcia-Ochoa A, Martinez-Riera A. The obesity paradox in elderly patients with heart failure: analysis of nutritional status. Nutrition 2012;28:616–22. 27. Huded CP, Huded JM, Friedman JL, et al. Frailty status and outcomes after transcatheter aortic valve implantation. Am J Cardiol 2016;117: 1966–71. 28. Afilalo J, Kim S, O’Brien S, et al. Gait speed and operative mortality in older adults following cardiac surgery. JAMA Cardiol 2016;1:314–21. 29. Pulignano G, Del Sindaco D, Di Lenarda A, et al. Incremental value of gait speed in predicting prognosis of older adults with heart failure. insights from the IMAGE-HF Study JACC Heart Fail 2016;4:289–98. 30. Grazzi G, Mazzoni G, Myers J, et al. Improved walking speed is associated with lower hospitalisation rates in patients in an exercise-based secondary prevention programme. Heart 2016;102:1902–8. 31. Fiser WM, Hays NP, Rogers SC, et al. Energetics of walking in elderly people: factors related to gait speed. J Gerontol A Biol Sci Med Sci 2010;65:1332–7. 32. McLean RR, Shardell MD, Alley DE, et al. Criteria for clinically relevant weakness and low lean mass and their longitudinal association with incident mobility impairment and mortality: the foundation for the National Institutes of Health (FNIH) sarcopenia project. J Gerontol A Biol Sci Med Sci 2014;69:576–83. 33. Meng P, Hu YX, Fan L, et al. Sarcopenia and sarcopenic obesity among men aged 80 years and older in Beijing: prevalence and its association with functional performance. Geriatr Gerontol Int 2014;14:29–35 [Suppl 1]. 34. Stoever K, Heber A, Eichberg S, Brixius K. Sarcopenia and predictors of skeletal muscle mass in elderly men with and without obesity. Gerontol Geriatr Med 2017;3:2333721417713637.
The American Journal of Medicine, Vol 000, No 000, && 2019 35. Fulster S, Tacke M, Sandek A, et al. Muscle wasting in patients with chronic heart failure: results from the studies investigating co-morbidities aggravating heart failure (SICA-HF). Eur Heart J 2013;34:512–9. 36. Iliodromiti S, Celis-Morales CA, Lyall DM, et al. The impact of confounding on the associations of different adiposity measures with the incidence of cardiovascular disease: a cohort study of 296 535 adults of white European descent. Eur Heart J 2018;39:1514–20. 37. Hegde SM, Claggett B, Shah AM, et al. Physical activity and prognosis in the TOPCAT Trial (Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist). Circulation 2017;136:982–92. 38. Stewart RAH, Held C, Hadziosmanovic N, et al. Physical activity and mortality in patients with stable coronary heart disease. J Am Coll Cardiol 2017;70:1689–700. 39. Stini WA. Body composition and longevity: is there a longevous morphotype? Med Anthropol 1991;13:215–29. 40. Briasoulis A, Bakris G. The future of interventional management of hypertension: threats and opportunities. Curr Vasc Pharmacol 2014;12:69–76. 41. Takiguchi M, Yoshihisa A, Miura S, et al. Impact of body mass index on mortality in heart failure patients. Eur J Clin Investig 2014;44: 1197–205. 42. Hamaguchi S, Tsuchihashi-Makaya M, Kinugawa S, et al. Body mass index is an independent predictor of long-term outcomes in patients hospitalized with heart failure in Japan. Circ J 2010;74:2605–11. 43. Rodriguez-Pascual C, Paredes-Galan E, Ferrero-Martinez AI, et al. The frailty syndrome is associated with adverse health outcomes in very old patients with stable heart failure: a prospective study in six Spanish hospitals. Int J Cardiol 2017;236:296–303. 44. Matsuzawa Y, Konishi M, Akiyama E, et al. Association between gait speed as a measure of frailty and risk of cardiovascular events after myocardial infarction. J Am Coll Cardiol 2013;61:1964–72. 45. Afilalo J, Mottillo S, Eisenberg MJ, et al. Addition of frailty and disability to cardiac surgery risk scores identifies elderly patients at high risk of mortality or major morbidity. Circ Cardiovasc Qual Outcomes 2012;5:222–8.
SUPPLEMENTARY DATA Supplementary data to this article can be found online at https://doi.org/10.1016/j.amjmed.2019.06.047.
Nakamura et al
ARTICLE IN PRESS
Gait Speed and Obesity Paradox in Older Cardiac Patients
8.e1
Supplementary Table S1 Patient characteristics Variable
Missing data
Preserved gait speed (> 0.8 m/s)
Slow gait speed ( 0.8 m/s)
Age (yrs) Male, n (%) Body weight (kg) Height (cm) BMI (kg/m2) Systolic blood pressure (mmHg) History of disease, n (%) Hheart failure Ischemic heart disease Cardiac surgery Hypertension Diabetes mellitus Dyslipidemia Prior heart failure Smoking, n (%) Left ventricular ejection fraction (%) eGFR (mL/min/1.73 m2) B-type natriuretic peptide (pg/mL) Albumin (g/dL) Hemoglobin (g/dL) LDL cholesterol (mg/dL) Gait speed (m/sec) All-cause mortality, n (%)
n (%)
(n=1740)
(n=484)
P Value
0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 57 (2.5)
71.5 ±6.8 1247 (71.7) 58.4 ± 11.0 161.1 ± 8.3 22.5 ± 3.5 122 ± 27
78.9 ± 8.0 237 (49.0) 51.6 ± 11.9 155.1 ± 9.5 21.4 ± 4.1 117 ± 26
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001
0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 69 (3.1) 188 (8.5) 11 (0.1) 215 (9.7) 15 (0.1) 12 (0.1) 248 (11.2) 0 (0.0) 0 (0.0)
776 (44.6) 974 (56.0) 747 (42.9) 1199 (68.9) 612 (35.2) 882 (50.7) 391 (22.5) 1078 (62.0) 53.9 ± 14.0 53.6 ± 19.7 168.55 [72.60, 349.03] 3.6 ± 0.5 12.1 ± 2.0 95 ± 31 1.13 ± 0.21 186 (10.7)
305 (63.0) 230 (47.5) 182 (37.6) 343 (70.9) 180 (37.2) 205 (42.4) 160 (33.1) 225 (46.5) 53.2 ± 15.4 46.3 ± 22.3 310.80 [145.05, 668.05] 3.3 ± 0.5 11.2 ± 1.8 90 ± 33 0.60 ± 0.15 97 (20.0)
0.001 0.001 0.037 0.435 0.421 0.001 <0.001 <0.001 0.300 <0.001 <0.001 <0.001 <0.001 0.006 <0.001 <0.001
Values are the means ± SD, median [interquartile range], or number (%). BMI: Body mass index; eGFR: estimated glomerular filtration rate
Impact of Gait Speed on the Obesity Paradox in Older Patients With Cardiovascular Disease Takeshi Nakamura, PT, MSca, Kentaro Kamiya, PT, PhDa,b, Atsuhiko Matsunaga, PT, PhDa,b, Nobuaki Hamazaki, PT, PhDc, Ryota Matsuzawa, PT, PhDd, Kohei Nozaki, PT, MScc, Masashi Yamashita, PT, MSca, Emi Maekawa, MD, PhDe, Chiharu Noda, MD, PhDe, Minako Yamaoka-Tojo, MD, PhDa,b, Junya Ako, MD, PhDe a
Department of Rehabilitation Sciences, Kitasato University Graduate School of Medical Sciences, Sagamihara, Japan; bDepartment of Rehabilitation, School of Allied Health Sciences, Kitasato University Sagamihara, Japan; cDepartment of Rehabilitation, Kitasato University Hospital, Sagamihara, Japan; dDepartment of Physical Therapy, School of Rehabilitation, Hyogo University of Health Sciences, Sagamihara, Japan; eDepartment of Cardiovascular Medicine, Kitasato University School of Medicine, Sagamihara, Japan.
ABSTRACT PURPOSE: The purpose of this study was to determine whether gait speed affects the obesity paradox in older patients with cardiovascular disease. METHODS: The study population consisted of 2224 patients ≥ 60 years old with cardiovascular disease admitted to hospital between May 1, 2006, and January 31, 2018. Body mass index (BMI) and gait speed before hospital discharge were determined, and patients were divided into two groups: slow and preserved gait speed (≤0.8 and > 0.8 m/s, respectively), according to the algorithm for sarcopenia diagnosis. The slow and preserved gait speed groups were also further subdivided according to BMI: < 18.5 kg/m2, 18.5−24.9 kg/m2, and BMI ≥ 25.0 kg/m2. The study endpoint was all-cause mortality. RESULTS: The study population (male: 66.7%) had a mean age of 73.1 § 7.6 years. Over a median followup period of 1.69 years (interquartile range 0.67−3.67 years), 283 patients died. Higher BMI was associated with favorable prognosis in the group with preserved gait speed but not in the group with slow gait speed after adjusting for other prognostic factors. Adding BMI to the clinical model significantly increased the area under the receiver operating characteristic curve in the group with preserved gait speed (0.744 vs 0.726, P = 0.028) but not in the group with slow gait speed (0.716 vs 0.716, P = 0.789). CONCLUSIONS: Higher BMI was consistently associated with favorable prognosis in patients with cardiovascular disease and preserved gait speed but not in those with slow gait speed. These findings indicated that physical frailty influences the obesity paradox in older patients with cardiovascular disease. Ó 2019 Elsevier Inc. All rights reserved. The American Journal of Medicine (2019) 000:1-8 KEYWORDS: Cardiovascular disease; Gait speed; Obesity paradox; Older; Prognosis
INTRODUCTION
Funding: This study was supported by a Grant for Clinical and Epidemiologic Research of the Joint Project of Japan Heart Foundation and the Japanese Society of Cardiovascular Disease Prevention Sponsored by AstraZeneca. Conflicts of Interest: None. Author contributions: All authors had access to the data and a role in writing this manuscript. Requests for reprints should be addressed to, Kentaro Kamiya, PT, PhD, Department of Rehabilitation, Kitasato University School of Allied Health Sciences, 1-15-1 Kitasato, Minami-ku, Sagamihara, Kanagawa, 252-0375, Japan E-mail address: [email protected]
0002-9343/© 2019 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.amjmed.2019.06.047
Although high body mass index (BMI) is known to be associated with elevated risk of cardiovascular disease,1,2 patients who are overweight and obese show better survival prognosis following the development of cardiovascular disease, including heart failure and coronary artery disease, in a phenomenon known as the “obesity paradox.”3,4 In contrast, high BMI is also a risk factor for both disability and falls5,6; therefore, high BMI does not necessarily have a protective effect against health disorders in older people with reduced physical function.7 The obesity paradox was reported in middle-aged and older patients with cardiovascular disease with preserved physical function.1,8−10 However, it is unclear whether the
2
ARTICLE IN PRESS
The American Journal of Medicine, Vol 000, No 000, && 2019
obesity paradox is present in older patients with cardiovascular disease with reduced physical function, such as those with slow gait speed. The present study was performed to determine whether the obesity paradox is also present in older patients with cardiovascular disease with slow gait speed.
METHODS Study Population This study involved a retrospective review of 2224 consecutive patients ≥60 years of age who were admitted for cardiovascular disease to the Cardiovascular Center of Kitasato University Hospital between May 1, 2006, and January 31, 2018, in all of whom usual gait speed had been evaluated during the hospitalization. The study was performed in accordance with the tenets of the Declaration of Helsinki and the protocol received approval from the Ethics Committee of Kitasato University Hospital.
Usual Gait Speed Measurement
Usual gait speed was determined by timing the patients walking at normal speed over the middle 10 m of a 16-m walkway. The use of walking aids, such as a cane, was permitted during the test. Patients were divided into slow gait speed (≤0.8 m/s) and preserved gait speed (>0.8 m/s) groups according to the European Working Group on Sarcopenia in Older People and Asian Working Group for Sarcopenia algorithm for diagnosis of CLINICAL SIGNIFICANCE sarcopenia.13,14
Higher body mass index (BMI) was consistently associated with favorable prognosis in patients with cardiovascular disease and preserved gait speed. In contrast, BMI showed no association with prognosis in the group with slow gait speed. Physical frailty, such as slow gait speed, may influence the obesity paradox in older patients with cardiovascular disease.
Data Collection Data on all variables were obtained from the patients’ electronic medical records. We recorded the clinical details for all patients, including medication use, comorbidities, demographic characteristics, and both echocardiographic and biochemical data immediately before hospital discharge. We defined diagnosis of heart failure before admission to hospital as prior heart failure. A stadiometer was used to measure height to the nearest 0.1 cm, and a calibrated weighing scale was used to measure weight to the nearest 0.1 kg at hospital discharge. BMI was calculated according to the formula: BMI = body weight (kg)/height (m2). Patients were divided according to the BMI categories recommended by the Japan Society for the Study of Obesity as follows: BMI < 18.5 kg/m2 (underweight), BMI 18.5−24.9 kg/m2 (normal), BMI ≥ 25.0 kg/m2 (obesity).11 A commercially available immunoradiometric assay was used to measure B-type natriuretic peptide (BNP) concentration (Shionogi, Osaka, Japan).The estimated glomerular filtration rate (eGFR) was determined using the Japanese Society of Nephrology formula: men = 194 £ (serum creatinine)1.094 £ (age)0.287; women = eGFR (male) £ 0.739.12 Left ventricular ejection fraction was estimated by Simpson’s method on two-dimensional echocardiographs. All-cause mortality was used as the study end point, with the time to the end point calculated as the period (in days) between the date of usual gait speed determination and the event. Follow-up was carried out according to hospital discharge and ended with the last clinical evaluation or with the patients’ death. If a patient died outside the hospital where they were being followed up, medical records of the event and a report of death were considered.
Statistical Analysis
Continuous variables are expressed as the mean § standard deviation, and non-normally distributed variables are presented as the median (interquartile range). Categorical variables are expressed as frequencies and percentages. Multivariate normal imputation for missing values was performed using JMP 13.2 software (SAS Institute Inc., Cary, NC).15,16 The following variables were incorporated into the imputation model: age, sex, BMI, diagnosis, eGFR, log BNP, albumin, hemoglobin, left ventricular ejection fraction, smoking, hypertension, diabetes mellitus, dyslipidemia and usual gait speed, and prior heart failure. We compared the baseline characteristics by one-way analysis of variance, Kruskal−Wallis test, unpaired Student’s t-test, or Mann−Whitney U test for continuous variables and chi-squared test or Fisher exact test for dichotomous variables, as appropriate. The cumulative incidence of mortality during follow-up period was calculated according to BMI categories based on the Kaplan−Meier curves in the slow gait speed and preserved gait speed groups. Intergroup differences were examined for significance by the log-rank test. Univariate and multivariate Cox regression analyses were used to evaluate the prognostic capabilities of BMI categories. Age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure were included in multivariate analyses. The associations between BMI and mortality risk were examined using Cox regression model adjusted for age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure with spline functions. In addition, to examine the potential effect modification on the association of BMI categories with mortality, we performed subgroup analyses of BMI categories with adjustment for age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure in various subgroups, including history of heart failure, ischemic heart disease, and cardiac surgery, age (stratified at 75 years), sex, and left ventricular ejection fraction (stratified at 50%). To avoid overfitting, all potential confounding factors of BMI, which include age, sex,
Nakamura et al
ARTICLE IN PRESS
Gait Speed and Obesity Paradox in Older Cardiac Patients
eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure, were reduced to one composite characteristic by applying a propensity score in subgroup analyses.17 We examined whether BMI had complementary predictive capability to age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure (clinical model) by constructing receiver operating characteristic (ROC) curves for all-cause mortality using the clinical model only and clinical model + BMI. The areas under the curve (AUC) were compared according to the method of DeLong et al.18 An AUC between 0.7 and 0.8 is considered acceptable discrimination, between 0.8 and 0.9 is considered excellent discrimination, and greater than 0.9 is considered outstanding discrimination.19 The continuous net reclassification improvement (cNRI) and integrated discrimination improvement (IDI)—sensitive statistical methods to quantify improvement of a model when a new variable is added—were used to calculate the increase in information by each BMI component compared to the prognostic factors.20 Two-tailed P <0.05 was taken to indicate statistical significance. Analyses were performed using R version 3.1.2 (R Foundation for Statistical Computing, Vienna, Austria) and JMP Pro 13 (SAS Institute Inc., Cary, NC).
RESULTS Study Population Baseline characteristics for all participants and for groups stratified according to gait speed and BMI are shown in Table 1 and Supplemental Table S1 (available online). The study population had a mean age of 73 § 8 years, and the gender distribution was 67% male and 33% female. Patients were hospitalized as a result of heart failure (35.1%), coronary artery disease (25.9%), cardiac surgery (21.7%), and other clinical entities (18.1%). Seventy-two percent of the patients were prescribed beta-blockers at discharge and 68% received angiotensin-converting enzyme inhibitors or angiotensin receptor blockers. Baseline characteristics for all participants and for groups stratified according to gait speed and BMI are shown in Table 1 and Supplemental Table S1. The study population had a mean age of 73 § 8 years, and the gender distribution was 67% male and 33% female. Patients were hospitalized as a result of heart failure (35.1%), coronary artery disease (25.9%), cardiac surgery (21.7%), and other clinical entities (18.1%). Seventy-two percent of the patients were prescribed beta-blockers at discharge and 68% received angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs).
Associations of BMI With All-Cause Mortality in Slow and Preserved Gait Speed Groups A total of 283 deaths occurred during the follow-up period (median 1.69 years, interquartile range 0.67-3.67 years, 49.5 deaths per 1000 person-years). The Kaplan−Meier
3
survival curves showed that higher BMI was associated with favorable prognosis in the preserved gait speed group (Figure 1A; log-rank P <0.001). In contrast, no such associations were observed between BMI categories and prognosis in the slow gait speed group (Figure 1B; log-rank P = 0.126). Table 2 shows the results of univariate and multivariate Cox regression analyses for all-cause mortality. On Cox regression analyses, the hazard ratio in BMI < 18.5 and BMI ≥25.0 in the preserved gait speed group were 2.06 (95% confidence interval [CI]: 1.40−3.03, P < 0.001) and 0.55 (95% CI: 0.35−0.88, P = 0.013), respectively, compared to BMI 18.5−24.9 after adjustment for the effects of age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure. On Cox regression analyses, the hazard ratio in BMI < 18.5 and BMI ≥ 25.0 in the preserved gait speed group were 2.06 (95% CI: 1.40−3.03, P <0.001) and 0.55 (95% CI: 0.35−0.88, P = 0.013), respectively, compared to BMI 18.5-24.9 after adjusting for the effects of age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure. On the other hand, in the fully adjusted model, the hazard ratio in BMI <18.5 and BMI ≥25.0 in the slow gait speed group were 1.04 (95% CI: 0.62−1.75, P = 0.875) and 0.68 (95% CI: 0.34−1.35, P = 0.271), respectively, compared to BMI 18.5−24.9 (Table 2). Figure 2 shows the association of BMI with mortality in slow and preserved gait speed groups. The CI of BMI after adjustment for age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure was wide at all BMI levels in the group with slow gait speed. In contrast, mortality risk decreased with increasing BMI in fully adjusted models in the group with preserved gait speed. Figure 3 shows that BMI < 18.5 was significantly associated with mortality compared to BMI 18.5-24.9 across various subgroups in the group with preserved gait speed after adjusting for age, sex, eGFR, log BNP, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure (Figure 3), except in patients who < 75 years old and with left ventricular ejection fraction ≥50. In contrast, BMI consistently showed no association with prognosis in the group with slow gait speed (Figure 3).
Additive Prognostic Predictive Capability of BMI to the Clinical Model As shown in Table 3, the logistic regression models of the clinical model only and clinical model + BMI were subjected to ROC curve analyses in the slow and preserved gait speed groups. The AUC on ROC curve analysis was greater with clinical model + BMI (0.744; 95% CI: 0.703−0.781) than the clinical model only (0.724; 95% CI: 0.684−0.765; P = 0.028) in the group with preserved gait speed. The addition of BMI to clinical model was also associated with significant increases in both cNRI (0.263; 95% CI: 0.114−0.413; P < 0.001) and IDI (0.012; 95% CI: 0.006−0.019; P <0.001) for all-cause mortality in the group with preserved gait speed. However, the addition of
ARTICLE IN PRESS
The American Journal of Medicine, Vol 000, No 000, && 2019
4 Table 1
Patient Characteristics
Variable
Missing Data
Overall
Preserved Gait Speed (>0.8 m/s)
P Value
P Value
Slow gait speed (≤ 0.8 m/s)
BMI < 18.5
BMI 18.5−24.9
BMI ≥ 25.0
BMI <18.5
BMI 18.5−24.9
BMI ≥ 25.0
(n = 112)
(n = 301)
(n = 71)
n (%)
(n = 2224)
(n = 209)
(n = 1156)
(n = 375)
Age (yrs) Male, n (%) Body weight (kg) Height (cm) BMI (kg/m2) SBP (mmHg) History of disease, n (%) Heart failure Ischemic heart disease Cardiac surgery Hypertension Diabetes mellitus Dyslipidemia Prior heart failure Smoking, n (%) LVEF (%) eGFR (mL/min/1.73 m2) B-type natriuretic peptide (pg/mL)
0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 57 (2.5)
73.1§7.6 1484 (66.7) 56.9 §11.5 159.8§8.9 22.2§3.7 121 §26
73.2§7.0 104 (49.8) 43.2§5.6 158.6§8.7 17.1 §1.2 116 §27
71.7 §6.8 874 (75.6) 57.2 §7.3 158.6 §8.7 21.8 §1.7 122 § 27
70.1§6.4 269 (71.7) 70.9§9.4 160.8§8.5 27.4§2.4 126§ 29
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001
79.8 §8.0 41 (36.6) 39.6 §5.6 154.3 §9.4 16.6 §1.3 112 §29
79.0§7.3 158 (52.5) 51.9§7.7 155.2§9.5 21.5§1.8 118§ 25
77.4§6.5 38 (53.5) 69.1§11.1 155.6§9.6 28.5§3.7 119 §20
0.108 0.011 <0.001 0.590 <0.001 0.074
0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 69 (3.1) 188 (8.5) 11 (0.1) 215 (9.7)
493 (42.6) 659 (57.0) 511 (44.2) 801 (69.3) 392 (33.9) 585 (50.6) 253 (21.9) 735 (63.6) 54.2 §14.2 52.7 §19.6 168.1 [71.6, 352.8] 3.6 §0.5 12.1 §2.0 96 § 30 1.14 §0.21 128 (11.1)
159 (42.4) 131 (34.9) 144 (38.4) 292 (77.9) 178 (47.5) 239 (63.7) 80 (21.3) 241 (64.3) 55.2 §13.4 53.7§18.2 140.2 [66.5, 277.8] 3.7§0.4 12.5§1.9 91 §30 1.13§0.20 21 ( 5.6)
74 (66.1) 41 (36.6) 32 (28.6) 66 (58.9) 23 (20.5) 32 (28.6) 43 (38.4) 49 (43.8) 52.0 §15.9 51.2 §25.4 367.9 [123.3, 831.9] 3.2 §0.5 11.1 §1.9 95 §34 0.60 §0.15 21 (18.8)
183 (60.8) 143 (47.5) 122 (40.5) 221 (73.4) 121 (40.2) 135 (44.9) 94 (31.2) 137 (45.5) 53.4§15.8 45.6§22.0 326.6 [156.0, 724.5] 3.3§0.5 11.1§1.8 89 §33 0.60§0.15 66 (21.9)
48 (67.6) 46 (64.8) 28 (39.4) 56 (78.9) 36 (50.7) 38 (53.5) 23 (32.4) 39 (54.9) 55.4§13.3 41.6§16.1 229.3 [122.1, 468.0] 3.3§0.5 11.5§1.8 89 §30 0.59§0.14 10 (14.1)
0.422 0.001 0.078 0.004 <0.001 0.001 0.385 0.289 0.322 0.011 0.034
15 (0.1) 12 (0.1) 248 (11.2) 0 (0.0) 0 (0.0)
124 (59.3) 71 (33.9) 92 (44.0) 106 (50.7) 42 (20.1) 58 (27.8) 58 (27.8) 102 (48.8) 52.4 §15.3 58.2§21.9 211.7 [107.3, 509.6] 3.4§0.5 11.5§1.7 98 §32 1.10§0.20 37 (17.7)
< 0.001 < 0.001 0.135 < 0.001 < 0.001 < 0.001 0.146 < 0.001 0.061 0.001 < 0.001
Albumin (g/dL) Hemoglobin (g/dL) LDL cholesterol (mg/dL) Gait speed (m/sec) All-cause mortality, n (%)
1081 (48.6) 1204 (54.1) 929 (41.8) 1542 (69.3) 792 (35.6) 1087 (48.9) 551 (24.8) 1303 (58.6) 53.8 §14.3 52.0§20.5 210.1 [92.9, 409.1] 3.5§0.5 11.9§2.0 94§31 1.02§0.29 283 (12.7)
< 0.001 < 0.001 0.024 0.043 < 0.001
0.110 0.148 0.342 0.947 0.308
Values are the means § SD, median [interquartile range], or number (%). BMI = body mass index; eGFR = estimated glomerular filtration rate; LDL = lowdensity lipoprotein; LVEF = left ventricular ejection fraction; SBP = systolic blood pressure.
BMI to the clinical model did not improve the AUC, cNRI, or IDI in the group with slow gait speed.
DISCUSSION The results of the present study indicated that higher BMI was consistently associated with favorable prognosis, and BMI showed complementary prognostic predictive capability to preexisting prognostic risk factors in patients with cardiovascular disease and preserved gait speed but not in those with slow gait speed. These results suggest that there is no obesity paradox in patients with cardiovascular disease and slow gait speed. Most previous studies suggested that there was obesity paradox in patients with cardiovascular disease, including heart failure and coronary artery disease.21−23 Although some reports suggested that the obesity paradox could be explained by several factors, such as body composition,24 exercise capacity,8,10 chronic inflammation,25 or nutritional status,26 it was unclear whether there was an obesity paradox in older patients with cardiovascular disease and lower levels of physical function. Gait speed is well established as a reliable, valid, sensitive, and specific measurement, and slow gait is a core component of frailty and sarcopenia.13,27 Therefore, measurement of gait speed has been widely used
to evaluate physical function in older adults. In this study, we measured gait speed as a means of determining physical function and investigated whether gait speed affected the obesity paradox in older patients with cardiovascular disease. To our knowledge, this is first report indicating differences in prognostic value of BMI between patients with cardiovascular disease and preserved gait speed compared to those with slow gait speed. Gait speed could be a useful indicator of mortality risk and is a potential therapeutic target because gait speed is an independent predictor of operative mortality,28,29 and improvement in gait speed is associated with reduced rates of rehospitalization and mortality in patients with cardiovascular disease.30 There are several possible underlying mechanisms for the association between BMI and prognosis in patients with cardiovascular disease and slow gait speed. Body composition may affect prognosis in patients with cardiovascular disease and slow gait speed. Some studies showed that gait speed was positively correlated to muscle mass and negatively correlated to fat mass in community-dwelling older adults31−34 and in patients with heart failure.35 Recently, the presence of both muscle wasting and high fat mass has been defined as sarcopenic obesity, which is related to lower limb muscle function, inflammation, or physical activity. Therefore, it is related to poorer outcome
Nakamura et al
ARTICLE IN PRESS
Gait Speed and Obesity Paradox in Older Cardiac Patients
5
Figure 1 Kaplan-Meier curves for all-cause mortality according to body mass index (BMI) categories in the preserved and slow gait speed groups.
compared to muscle wasting alone. Moreover, several studies suggested that mortality risk in subjects with higher BMI was likely the result of the high risk associated with fat mass, and increased fat mass did not have a protective effect against mortality risk.24,36 These studies suggested that patients with cardiovascular disease and a high BMI, especially those with slow gait speed, may have both low muscle mass and high fat mass, which would adversely affect their prognosis. Physical inactivity represents another possible mechanism underlying the relation between BMI and prognosis in patients with slow gait speed. The act of walking uses energy and places demands on multiple organ systems, including the heart, lungs, circulatory system, nervous system, and musculoskeletal system. A slow gait may reflect the high energy costs of walking and indicate damage to these systems, which may be responsible for reduced levels of physical inactivity in the older patient. Reduced physical Table 2
activity itself is known to be a strong prognostic factor,37,38 and therefore, slow gait speed may be associated with poor prognosis despite BMI in patients with cardiovascular disease. On the other hand, the results of the present study showed that higher BMI was associated better prognosis in patients with cardiovascular disease and preserved gait speed. Muscle mass showed a consistent protective effect in community-dwelling subjects and patients with cardiovascular disease. Previous studies suggested that muscle mass had a protective effect against mortality risk during prolonged illness by providing a reliable protein reserve39 or the metabolic effect of changes in skeletal muscle.40 These studies suggested that patients with cardiovascular disease, high BMI, and preserved gait speed may have high muscle mass and may, therefore, have good prognosis. Although several recent studies regarding the obesity paradox suggested that increasing or preserving body
Univariariate and Multivariate Cox Regression Analysis for All-Cause Mortality
Variable
Preserved gait speed BMI < 18.5 BMI 18.5−24.9 BMI ≥ 25.0 Slow gait speed BMI < 18.5 BMI 18.5−24.9 BMI ≥ 25.0
Univariate Cox Analysis
Multivariate Cox Analysis*
HR
95%CI
P Value
HR
95%CI
P Value
2.01 1.00 0.47
1.39−2.90 (Reference) 0.29−0.74
< 0.001 (Reference) 0.001
2.06 1.00 0.55
1.40−3.03 (Reference) 0.35−0.88
< 0.001 (Reference) 0.013
1.12 1.00 0.54
0.69−1.84 (Reference) 0.28−1.04
0.648 (Reference) 0.067
1.04 1.00 0.68
0.62−1.75 (Reference) 0.34−1.35
0.875 (Reference) 0.271
P values represent pairwise relationships relative to reference group. BMI = body mass index; CI = confidence interval; HR = hazard ratio. *Adjusted HR after adjustment for age, sex, estimated glomerular filtration rate, log B-type natirutic peptide, albumin, hemoglobin, left ventricular ejection fraction, smoking, and prior heart failure
6
ARTICLE IN PRESS
The American Journal of Medicine, Vol 000, No 000, && 2019
Figure 2 Associations of body mass index (BMI) with all-cause mortality in the preserved and slow gait speed groups. Notes: Dotted lines represent the 95% confidence intervals. Rug plots are shown along the x-axes of the graphs to depict the distributions of BMI. The adjusted models were adjusted for age, sex, estimated glomerular filtration rate, log B-type natriuretic peptide, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure.
weight should be recommended for older patients with cardiovascular disease,41,42 frailty status was not evaluated in these studies. Therefore, it was unknown whether frailty status influenced the obesity paradox in patients with
cardiovascular disease. On the other hand, previous studies have indicated that slow gait speed is more closely associated with increased mortality than decreased body weight in patients with cardiovascular disease.43,44 Moreover,
Figure 3 Forest plot of hazard ratio for association of body mass index (BMI) with all-cause mortality in the preserved and slowgait-speed groups. *Adjusted for applying a propensity score calculated by age, sex, estimated glomerular filtration rate, log B-type natriuretic peptide, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure. CI = confidence interval; HR = hazard ratio; IHD = ischemic heart disease; LVEF = left ventricular ejection fraction.
Nakamura et al Table 3
ARTICLE IN PRESS
Gait Speed and Obesity Paradox in Older Cardiac Patients
7
Predictive Value Analyses for All-Cause Mortality AUC
Preserved gait speed Clinical model Clinical model + BMI Slow gait speed Clinical model Clinical model + BMI
95%CI
P Value
cNRI
0.726 0.684−0.765 (Reference) 0.744 0.703−0.781 0.028 0.263
95%CI
P Value
0.114−0.413
(Reference) < 0.001 0.012
IDI
95%CI
P Value
0.006−0.019
(Reference) < 0.001
0.716 0.656−0.769 (Reference) (Reference) (Reference) 0.716 0.656−0.770 0.789 -0.013 -0.235−0.208 0.907 -0.0001 -0.0006−0.0004 0.758
P values represent pairwise relationships relative to reference group. BMI was added to Clinical model containing age, sex, estimated glomerular filtration rate, log B-type natriuretic peptide, left ventricular ejection fraction, hemoglobin, albumin, smoking, and prior heart failure. AUC = areas under the curve; BMI = body mass index; CI = confidence interval; cNRI = continuous net reclassification improvement; IDI = integrated discrimination improvement.
another study indicated that slow gait speed outperformed other multicomponent frailty scales, including body weight assessment, for predicting mortality in patients with cardiovascular disease, even if it is a single measurement.45 These studies and our results indicated that BMI does not necessarily provide increased value in patients with cardiovascular disease in whom gait speed is already compromised.
Study Limitations This study had several limitations, the most important of which are its retrospective design, limited follow-up period, and small sample size in the group with slow gait speed. In addition, only patients in whom gait speed had been measured were included in the study. However, patients with missing data may have more severe illness than those with complete data, and missing data can cause biased estimates. Body composition was not measured directly by magnetic resonance imaging or dual-energy X-ray absorptiometry. Finally, all subjects included in this study were Asian, so further studies are required to determine the generalizability of the results to other populations.
CONCLUSIONS Higher BMI was consistently associated with favorable prognosis in patients with cardiovascular disease and preserved gait speed but not in those with slow gait speed. In addition, BMI showed complementary prognostic predictive capability to preexisting prognostic risk factors in patients with preserved gait speed but not in those with slow gait speed. These findings indicate that physical frailty may influence the obesity paradox in older patients with cardiovascular disease.
References 1. Lavie CJ, Alpert MA, Arena R, Mehra MR, Milani RV, Ventura HO. Impact of obesity and the obesity paradox on prevalence and prognosis in heart failure. JACC Heart Fail 2013;1:93–102. 2. Krauss RM, Winston M, Fletcher BJ, Grundy SM. Obesity: impact on cardiovascular disease. Circulation 1998;98:1472–6. 3. Gruberg L, Weissman NJ, Waksman R, et al. The impact of obesity on the short-term and long-term outcomes after percutaneous coronary intervention: the obesity paradox? J Am Coll Cardiol 2002;39: 578–84.
4. Clark AL, Fonarow GC, Horwich TB. Obesity and the obesity paradox in heart failure. Prog Cardiovasc Dis 2014;56:409–14. 5. Mitchell RJ, Lord SR, Harvey LA, Close JC. Obesity and falls in older people: mediating effects of disease, sedentary behavior, mood, pain and medication use. Arch Gerontol Geriatr 2015;60:52–8. 6. Kosar CM, Thomas KS, Gozalo PL, Mor V. Higher level of obesity Is associated with intensive personal care assistance in the nursing home. J Am Med Dir Assoc 2018;19:1015–9. 7. Yamashita M, Kamiya K, Matsunaga A, et al. Prognostic value of sarcopenic obesity estimated by computed tomography in patients with cardiovascular disease and undergoing surgery. J Cardiol 2019. https://doi.org/10.1016/j.jjcc.2019.02.010. 8. Lavie CJ, Cahalin LP, Chase P, et al. Impact of cardiorespiratory fitness on the obesity paradox in patients with heart failure. Mayo Clin Proc 2013;88:251–8. 9. Piepoli MF, Corra U, Veglia F, et al. Exercise tolerance can explain the obesity paradox in patients with systolic heart failure: data from the MECKI Score Research Group. Eur J Heart Fail 2016;18:545–53. 10. McAuley PA, Keteyian SJ, Brawner CA, et al. Exercise capacity and the obesity paradox in heart failure: The FIT (Henry Ford Exercise Testing) Project. Mayo Clin Proc 2018;93:701–8. 11. Examination Committee of Criteria for ’Obesity Disease’ in Japan; Japan Society for the Study of Obesity. New criteria for ’obesity disease’ in Japan. Circ J 2002;66:987–92. 12. Ando Y, Ito S, Uemura O, et al. CKD Clinical Practice Guidebook. The essence of treatment for CKD patients. Clin Exp Nephrol 2009;13: 191–248. 13. Chen LK, Liu LK, Woo J, et al. Sarcopenia in Asia: consensus report of the Asian Working Group for Sarcopenia. J Am Med Dir Assoc 2014;15:95–101. 14. Cruz-Jentoft AJ, Baeyens JP, Bauer JM, et al. Sarcopenia: European consensus on definition and diagnosis: Report of the European Working Group on Sarcopenia in Older People. Age Ageing 2010;39: 412–23. 15. Schafer J, Strimmer K. A shrinkage approach to large-scale covariance matrix estimation and implications for functional genomics. Stat Appl Genet Mol Biol 2005;4:Article32. 16. Koch E, Finne K, Eikrem O, et al. Transcriptome-proteome integration of archival human renal cell carcinoma biopsies enables identification of molecular mechanisms. Am J Physiol Renal Physiol 2019;316:F1053–67. 17. Rubin DB. Estimating causal effects from large data sets using propensity scores. Ann Intern Med 1997;127:757–63. 18. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837–45. 19. Hosmer DW, Lemeshow S. Applied Logistic Regression. 2nd ed. New York, NY: Wiley; 2000. 20. Pencina MJ, D’Agostino R.B. Sr., D’Agostino R.B. Jr., Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008;27:157–72 [discussion 207-212].
8
ARTICLE IN PRESS
21. Mariscalco G, Wozniak MJ, Dawson AG, et al. Body mass index and mortality among adults undergoing cardiac surgery: a nationwide study with a systematic review and meta-analysis. Circulation 2017;135:850–63. 22. Moholdt T, Lavie CJ, Nauman J. Interaction of physical activity and body mass index on mortality in coronary heart disease: data from the Nord-Trondelag Health Study. Am J Med 2017;130:949–57. 23. Kamiya K, Masuda T, Matsue Y, et al. Complementary role of arm circumference to body mass index in risk stratification in heart failure. JACC Heart Fail 2016;4:265–73. 24. Lee DH, Keum N, Hu FB, et al. Predicted lean body mass, fat mass, and all cause and cause specific mortality in men: prospective US cohort study. BMJ 2018;362:k2575. 25. Antonopoulos AS, Tousoulis D. The molecular mechanisms of obesity paradox. Cardiovasc Res 2017;113:1074–86. 26. Casas-Vara A, Santolaria F, Fernandez-Bereciartua A, GonzalezReimers E, Garcia-Ochoa A, Martinez-Riera A. The obesity paradox in elderly patients with heart failure: analysis of nutritional status. Nutrition 2012;28:616–22. 27. Huded CP, Huded JM, Friedman JL, et al. Frailty status and outcomes after transcatheter aortic valve implantation. Am J Cardiol 2016;117: 1966–71. 28. Afilalo J, Kim S, O’Brien S, et al. Gait speed and operative mortality in older adults following cardiac surgery. JAMA Cardiol 2016;1:314–21. 29. Pulignano G, Del Sindaco D, Di Lenarda A, et al. Incremental value of gait speed in predicting prognosis of older adults with heart failure. insights from the IMAGE-HF Study JACC Heart Fail 2016;4:289–98. 30. Grazzi G, Mazzoni G, Myers J, et al. Improved walking speed is associated with lower hospitalisation rates in patients in an exercise-based secondary prevention programme. Heart 2016;102:1902–8. 31. Fiser WM, Hays NP, Rogers SC, et al. Energetics of walking in elderly people: factors related to gait speed. J Gerontol A Biol Sci Med Sci 2010;65:1332–7. 32. McLean RR, Shardell MD, Alley DE, et al. Criteria for clinically relevant weakness and low lean mass and their longitudinal association with incident mobility impairment and mortality: the foundation for the National Institutes of Health (FNIH) sarcopenia project. J Gerontol A Biol Sci Med Sci 2014;69:576–83. 33. Meng P, Hu YX, Fan L, et al. Sarcopenia and sarcopenic obesity among men aged 80 years and older in Beijing: prevalence and its association with functional performance. Geriatr Gerontol Int 2014;14:29–35 [Suppl 1]. 34. Stoever K, Heber A, Eichberg S, Brixius K. Sarcopenia and predictors of skeletal muscle mass in elderly men with and without obesity. Gerontol Geriatr Med 2017;3:2333721417713637.
The American Journal of Medicine, Vol 000, No 000, && 2019 35. Fulster S, Tacke M, Sandek A, et al. Muscle wasting in patients with chronic heart failure: results from the studies investigating co-morbidities aggravating heart failure (SICA-HF). Eur Heart J 2013;34:512–9. 36. Iliodromiti S, Celis-Morales CA, Lyall DM, et al. The impact of confounding on the associations of different adiposity measures with the incidence of cardiovascular disease: a cohort study of 296 535 adults of white European descent. Eur Heart J 2018;39:1514–20. 37. Hegde SM, Claggett B, Shah AM, et al. Physical activity and prognosis in the TOPCAT Trial (Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist). Circulation 2017;136:982–92. 38. Stewart RAH, Held C, Hadziosmanovic N, et al. Physical activity and mortality in patients with stable coronary heart disease. J Am Coll Cardiol 2017;70:1689–700. 39. Stini WA. Body composition and longevity: is there a longevous morphotype? Med Anthropol 1991;13:215–29. 40. Briasoulis A, Bakris G. The future of interventional management of hypertension: threats and opportunities. Curr Vasc Pharmacol 2014;12:69–76. 41. Takiguchi M, Yoshihisa A, Miura S, et al. Impact of body mass index on mortality in heart failure patients. Eur J Clin Investig 2014;44: 1197–205. 42. Hamaguchi S, Tsuchihashi-Makaya M, Kinugawa S, et al. Body mass index is an independent predictor of long-term outcomes in patients hospitalized with heart failure in Japan. Circ J 2010;74:2605–11. 43. Rodriguez-Pascual C, Paredes-Galan E, Ferrero-Martinez AI, et al. The frailty syndrome is associated with adverse health outcomes in very old patients with stable heart failure: a prospective study in six Spanish hospitals. Int J Cardiol 2017;236:296–303. 44. Matsuzawa Y, Konishi M, Akiyama E, et al. Association between gait speed as a measure of frailty and risk of cardiovascular events after myocardial infarction. J Am Coll Cardiol 2013;61:1964–72. 45. Afilalo J, Mottillo S, Eisenberg MJ, et al. Addition of frailty and disability to cardiac surgery risk scores identifies elderly patients at high risk of mortality or major morbidity. Circ Cardiovasc Qual Outcomes 2012;5:222–8.
SUPPLEMENTARY DATA Supplementary data to this article can be found online at https://doi.org/10.1016/j.amjmed.2019.06.047.
Nakamura et al
ARTICLE IN PRESS
Gait Speed and Obesity Paradox in Older Cardiac Patients
8.e1
Supplementary Table S1 Patient characteristics Variable
Missing data
Preserved gait speed (> 0.8 m/s)
Slow gait speed ( 0.8 m/s)
Age (yrs) Male, n (%) Body weight (kg) Height (cm) BMI (kg/m2) Systolic blood pressure (mmHg) History of disease, n (%) Hheart failure Ischemic heart disease Cardiac surgery Hypertension Diabetes mellitus Dyslipidemia Prior heart failure Smoking, n (%) Left ventricular ejection fraction (%) eGFR (mL/min/1.73 m2) B-type natriuretic peptide (pg/mL) Albumin (g/dL) Hemoglobin (g/dL) LDL cholesterol (mg/dL) Gait speed (m/sec) All-cause mortality, n (%)
n (%)
(n=1740)
(n=484)
P Value
0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 57 (2.5)
71.5 ±6.8 1247 (71.7) 58.4 ± 11.0 161.1 ± 8.3 22.5 ± 3.5 122 ± 27
78.9 ± 8.0 237 (49.0) 51.6 ± 11.9 155.1 ± 9.5 21.4 ± 4.1 117 ± 26
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001
0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 69 (3.1) 188 (8.5) 11 (0.1) 215 (9.7) 15 (0.1) 12 (0.1) 248 (11.2) 0 (0.0) 0 (0.0)
776 (44.6) 974 (56.0) 747 (42.9) 1199 (68.9) 612 (35.2) 882 (50.7) 391 (22.5) 1078 (62.0) 53.9 ± 14.0 53.6 ± 19.7 168.55 [72.60, 349.03] 3.6 ± 0.5 12.1 ± 2.0 95 ± 31 1.13 ± 0.21 186 (10.7)
305 (63.0) 230 (47.5) 182 (37.6) 343 (70.9) 180 (37.2) 205 (42.4) 160 (33.1) 225 (46.5) 53.2 ± 15.4 46.3 ± 22.3 310.80 [145.05, 668.05] 3.3 ± 0.5 11.2 ± 1.8 90 ± 33 0.60 ± 0.15 97 (20.0)
0.001 0.001 0.037 0.435 0.421 0.001 <0.001 <0.001 0.300 <0.001 <0.001 <0.001 <0.001 0.006 <0.001 <0.001
Values are the means ± SD, median [interquartile range], or number (%). BMI: Body mass index; eGFR: estimated glomerular filtration rate