Obesity, End-stage Renal Disease, and Survival in an Elderly Cohort With Cardiovascular Disease

Article (PDF Available)inObesity 17(12):2216-22 · April 2009with5 Reads
DOI: 10.1038/oby.2009.70 · Source: PubMed
Abstract
Obesity is highly prevalent in African Americans and is associated with increased risk of End-Stage Renal Disease (ESRD) and death. It is not known if the effect of obesity is similar among blacks and whites. The aim of this study is to examine racial differences in the association of obesity with ESRD and survival in elderly patients (age >65). Data were obtained for 74,167 Medicare patients with acute myocardial infarction (AMI) between February 1994 and July 1995. BMI was calculated as weight (kg) divided by height (m(2)). We evaluated the association of BMI class with ESRD incidence and death using multivariable Cox proportional hazards models, testing for race-BMI interactions. Compared to whites, African Americans had higher BMI (26.9 vs. 26.0, P < 0.0001) and estimated glomerular filtration rate (72.4 ml/min/1.73 m(2) vs. 66.6 ml/min/1.73 m(2), P < 0.0001). Crude ESRD rates increased with increasing obesity among whites but not among blacks. However, after adjusting for age, sex, and other comorbidities, obesity was not associated with increased ESRD rate among blacks or whites and the interaction between race and BMI was not significant. Furthermore, for both races, patients classified as overweight, class 1 obese, or class 2 obese had similar, significantly better survival abilities compared to normal weight patients and the race BMI interaction was not significant. In conclusion, obesity does not increase risk of ESRD among black or white elderly subjects with cardiovascular disease (CVD). However, both obese blacks and whites, in this population, experience a survival benefit. Further studies need to explore this obesity paradox.
2216 VOLUME 17 NUMBER 12 | DECEMBER 2009 | www.obesityjournal.org
articles nature publishing group
epidemiology
INTRODUCTION
Obesity is a serious health care issue worldwide. In the United
States, more than one half of all adults are overweight and
obesity ranks second to smoking as a cause of preventable
death. African Americans, particularly black women, tend to
have a higher prevalence of obesity when compared to other
various racial and ethnic groups (1). ere is evidence that
obesity is associated with loss of kidney function (2–4), and
one study concluded that a high BMI is potentially a modiable
risk factor for End-Stage Renal Disease (ESRD) (3). In addi-
tion, incident dialysis patients with a family history of ESRD
are reported to have a higher prevalence of obesity, and older
adults with a rst degree relative with ESRD are more obese
than others (5). African Americans have a higher incidence of
kidney failure and disproportionally comprise up to 30% of the
ESRD population. It is uncertain whether obesity contributes
to racial dierences in risk of ESRD; and, thus, the rst aim
of this study is to examine whether obesity is associated with
racial dierences in the risk for progression to ESRD.
ere are reports suggesting a paradoxical association with
obesity and improved survival in the ESRD and heart fail-
ure populations (6,7). It has also been reported that African
Americans with ESRD survive longer than whites despite hav-
ing a higher burden of cardiovascular disease (CVD) (8,9).
Until recently it was not known if the same racial survival
benet applied to a pre-ESRD population or to other CVD
populations as well. We recently reported that elderly African-
American patients with more severe chronic kidney disease
(CKD) and incident acute myocardial infarction (AMI) expe-
rienced better survival (8,10). us, our second aim is to exam-
ine the impact of obesity on survival among this same elderly
cohort with CVD and to determine if the eect of obesity on
survival diered among whites and African Americans.
METHODS AND PROCEDURES
Study population
We analyzed data from the Cooperative Cardiovascular Project, which
features data from 234,754 Medicare patients seen at 6,684 hospitals
Obesity, End-stage Renal Disease,
and Survival in an Elderly Cohort With
Cardiovascular Disease
Janice P. Lea1, Daryl O. Crenshaw1, Stephen J. Onufrak2, Britt B. Newsome2,3 and William M. McClellan4
Obesity is highly prevalent in African Americans and is associated with increased risk of End-Stage Renal Disease
(ESRD) and death. It is not known if the effect of obesity is similar among blacks and whites. The aim of this study
is to examine racial differences in the association of obesity with ESRD and survival in elderly patients (age >65).
Data were obtained for 74,167 Medicare patients with acute myocardial infarction (AMI) between February 1994 and
July 1995. BMI was calculated as weight (kg) divided by height (m2). We evaluated the association of BMI class with
ESRD incidence and death using multivariable Cox proportional hazards models, testing for race-BMI interactions.
Compared to whites, African Americans had higher BMI (26.9 vs. 26.0, P < 0.0001) and estimated glomerular filtration
rate (72.4 ml/min/1.73 m2 vs. 66.6 ml/min/1.73 m2, P < 0.0001). Crude ESRD rates increased with increasing obesity
among whites but not among blacks. However, after adjusting for age, sex, and other comorbidities, obesity was
not associated with increased ESRD rate among blacks or whites and the interaction between race and BMI was
not significant. Furthermore, for both races, patients classified as overweight, class 1 obese, or class 2 obese had
similar, significantly better survival abilities compared to normal weight patients and the race BMI interaction was
not significant. In conclusion, obesity does not increase risk of ESRD among black or white elderly subjects with
cardiovascular disease (CVD). However, both obese blacks and whites, in this population, experience a survival
benefit. Further studies need to explore this obesity paradox.
Obesity (2009) 17, 2216–2222. doi:10.1038/oby.2009.70
1Department of Medicine, Emory University School of Medicine, Atlanta, Georgia, USA; 2Agricultural Research Service, US Department of Agriculture, Stoneville,
Mississippi, USA; 3Department of Medicine, University of Alabama School of Medicine, Birmingham, Alabama, USA; 4Department of Public Health, Emory University
School of Medicine, Atlanta, Georgia, USA. Correspondence: Janice P. Lea (jlea@emory.edu)
Received 6 February 2008; accepted 23 February 2009; published online 26 March 2009. doi:10.1038/oby.2009.70
OBESITY | VOLUME 17 NUMBER 12 | DECEMBER 2009 2217
articles
epidemiology
in all 50 states for AMI between February 1994 and July 1995 (ref. 11).
Patients with an International Classication of Disease (12) principal
discharge diagnosis code of 410 (AMI) were sampled from 6,684 hos-
pitals, comprising virtually all acute care hospitals in the United States.
Date of death for patients not diagnosed with ESRD was obtained from
the Medicare Enrollment Database. Follow-up for mortality included
the time period from the inception of the Cooperative Cardiovascular
Project data collection (February 1994 through June 2004), which has
been previously described (10).
Exclusions
For the present analysis, we excluded patients with the following crite-
ria: death prior to hospital discharge (n = 32,655), AMI unconrmed
by medical criteria (n = 30,485), if the admission was a second hospi-
talization for MI during the study period (n = 23,773), age <65 years
(n = 17,591), patient transferred to an index hospital (n = 39,025),
or transferred from an index hospital within 24 h of admission
(n = 42,176), race other than white or African American (n = 3,455),
and rst dialysis during hospitalization (n = 595). In addition, 14,099
patients were excluded due to missing data for height or weight and
4,827 subjects were excluded due to inability to determine death status.
is le 74,167 study subjects available for analysis.
Study variables
We used the rst serum creatinine level collected within 24 h of hospital
admission to estimate baseline kidney function. Estimated glomeru-
lar ltration rate, calculated using the abbreviated Modication of Diet
in Renal Disease Study equation, was used as the measure of baseline
kidney function (13). We divided patients into categories of increas-
ing kidney disease severity based on the National Kidney Disease
Outcomes Quality Initiatives stages of kidney disease (14). We dened
anemia according to World Health Organization criteria as a hemat-
ocrit <39% in men and <36% in women. Race classication, height and
weight, use of angiotensin-converting enzyme inhibitor, or β-blocking
drugs at discharge, and information on comorbidities were obtained
through medical chart review. We created four mutually exclusive levels
of hospital technology mix, as described previously (15). We calculated
BMI as weight in kg divided by height in m2. Using the National Heart
Lung and Blood Institute guidelines, we dened overweight as a BMI of
25.0–29.9 kg/m2, class I obesity as a BMI of 30.0–34.9 kg/m2, class II as
a BMI of 35.0–39.9 kg/m2, and class III (extreme) as a BMI of 40.0 kg/
m2. Underweight was dened as a BMI <18.5 kg/m2. We conducted all
analyses using normal BMI (18.5–24.9 kg/m2) as the reference group.
Outcome variables
ESRD was dened as the requirement of renal replacement therapy in
the form of maintenance hemodialysis, peritoneal dialysis, or as a recip-
ient of renal transplantation. We identied cases of ESRD by matching
our cohort against the US Renal Data System registry data. Because
all subjects were Medicare patients, death information was obtained
through Medicare records. Patients were followed for ~10 years until
development of ESRD, death, or until the end of the follow-up period.
Follow-up data were available for the time period from initial enroll-
ment through June 2004.
Statistical analysis
All Statistical Analysis were conducted using SAS version 9.1 soware
(SAS, Cary, NC). Missing values for covariates were provided using
multiple imputation methods. We compared baseline covariates accord-
ing to race and body-weight class using χ2-tests, t-tests, and ANOVA.
We calculated a crude rate of ESRD according to race and body-weight
class by dividing the number of incident ESRD cases by the number
of person/years of observation in each category. We then t the data
to multivariable Cox proportional hazards models, ensuring that the
variables satised the proportional hazards assumptions. We used indi-
cator variables for body-weight class and included covariables for age,
sex, hospital technology index, baseline kidney function, hematocrit,
albumin, glucose, diabetes, hypertension, blood pressure, smoking,
prior MI, congestive heart failure, stroke, coronary artery bypass gra-
ing, percutaneous transluminal coronary angioplasty, angiotensin-
converting enzyme inhibitor and β-blocker at discharge, and Acute
Physiology and Chronic Health Evaluation II Score. We, then, calcu-
lated the death rates according to race and body-weight class and t the
data to age, sex, and multivariable adjusted Cox proportional hazards
models controlling for the same variables, as previously mentioned. For
both ESRD and death outcomes, we tested the interaction of race with
body-weight class using the likelihood-ratio test. Finally, we performed
supplemental analysis to determine the potential eect of dierential
hospital survival among obese subjects, we investigated whether sur-
vival to discharge diered according to body-weight status among the
93,112 subjects, who did not necessarily survive to discharge but oth-
erwise satised all other exclusion criteria. is analysis was performed
using a logistic regression model adjusted for age, sex, and race.
RESULTS
Aer exclusions, there were 74,167 patients available for analy-
sis, of whom 4,937 (6.7%) were black. e overall prevalence of
obesity was 18.3% and similar to that in the general population
(16). e mean BMI was higher among African Americans
than whites (26.9 vs. 26.0, P < 0.0001) and black patients were
also more likely to be underweight or obese (Table 1). A com-
parison of baseline characteristic according to race has been
published previously (9). Briey, compared to whites, black
patients in the Cooperative Cardiovascular Project cohort
were younger, more frequently female, more oen currently
smoking, and had a higher prevalence of comorbidities that
included diabetes, hypertension, prior congestive heart failure,
previous stroke, and anemia. Furthermore, blacks also had a
higher blood pressure and a higher mean glomerular ltration
rate, but were more likely to have a glomerular ltration rate
<15 compared to whites, whereas whites were more likely to be
treated with β-blockers at discharge.
Table 1 lists baseline characteristics of black and white sub-
jects according to BMI class. ose in the higher BMI catego-
ries were more likely to be younger and female and less likely
to be anemic. Blood pressure, glucose levels, and prevalence
rates of hypertension and diabetes were higher as BMI class
increased.
During follow-up, 1,186 white and 230 black subjects devel-
oped ESRD. Among all body-weight classes, blacks had higher
crude rates of ESRD compared to whites (Ta b l e 2 ). Among
whites, the rate of ESRD increased from 255/100,000 per year
(PY) among underweight subjects to 420/100,000 PY among
subjects with class III obesity (Table 2). Among blacks, ESRD
rates were highest among those with normal body weight
and lowest among underweight and the most obese subjects
(Table 2). When the data were tted to an age and sex adjusted
Cox proportional hazards model (Table 2), hazard ratios (HRs)
increased from 1.08 (95% condence interval (CI): 0.94–1.23)
for overweight whites to 1.26 (95% CI: 0.82–1.94) for class III
obese whites compared to their normal weight white coun-
terparts. By contrast, among blacks all greater than normal
body-weight classes were associated with protective but non-
signicant HRs for ESRD compared to normal weight blacks
(Table 2). Underweight individuals experienced a lower rate of
ESRD among both whites and blacks.
2218 VOLUME 17 NUMBER 12 | DECEMBER 2009 | www.obesityjournal.org
articles
epidemiology
Despite this apparent racial dierence in the eect of obesity
on ESRD rate, the interaction in the age-adjusted model was
not statistically signicant (P value = 0.18).
When the data were tted to a multivariable Cox proportional
hazards model controlling for comorbidities and clinical charac-
teristics (Ta b l e 2 ), obesity was not signicantly associated with
ESRD among whites or blacks (Ta bl e 2 ) and the race-obesity
interaction remained nonsignicant (P value = 0.50). In the
noninteraction model, class II obesity was protectively associ-
ated with ESRD (HR: 0.77, 95% CI: 0.60–0.99) among both races
combined (Tab l e 3 ). A similar but nonsignicant association
was also observed among class III obese patients (Table 3).
Table 1 Baseline characteristics according to body-weight class
Characteristic Underweight Normal Overweight Class I obesity Class II obesity Class III obesity P value
N (%) 3,401
(4.6%)
30,128
(40.6%)
27,064
(36.5%)
9,807
(13.2%)
2,658
(3.6%)
1,109
(1.5%)
Age (mean) 80.7 78.3 75.6 74.3 73.4 72.7 <0.0001
Female (%) 69.9 50.4 40.8 51.6 61.6 69.0 <0.0001
Black (%) 4.9 5.8 6.3 8.7 10.1 12.3 <0.0001
Current smoking (%) 20.7 16.3 13.8 12.2 12.3 10.8 <0.0001
Systolic blood pressure (mean) 123.2 124.0 125.3 126.9 128.3 128.9 <0.0001
Diastolic blood pressure (mean) 66.6 67.1 68.6 69.2 69.5 69.5 <0.0001
Past medical history
Diabetes (%) 15.0 23.5 32.1 42.4 49.4 53.5 <0.0001
Hypertension (%) 54.2 58.4 63.8 71.0 74.8 76.9 <0.0001
Previous myocardial infarction (%) 28.7 31.5 32.1 31.0 31.3 29.6 0.003
CHF (%) 31.1 24.2 19.2 20.7 23.7 28.2 <0.0001
Stroke (%) 16.5 15.0 12.9 13.3 12.6 11.8 <0.0001
CABG (%) 7.7 12.9 15.3 14.3 11.0 7.9 <0.0001
PTCA (%) 4.0 6.6 8.2 8.0 7.2 6.1 <0.0001
Kidney function
Mean GFR (ml/min/1.73 m2) 66.6 66.1 67.8 67.6 65.8 66.4 <0.0001
GFR 60 ml/min/1.73 m2 (%) 55.0 57.3 60.9 60.4 56.3 56.7 <0.0001
GFR 45–60 ml/min/1.73 m2 (%) 21.5 22.7 23.0 22.8 23.7 24.2
GFR 30–45 ml/min/1.73 m2 (%) 14.9 13.4 11.6 12.1 14.4 13.8
GFR 15–30 ml/min/1.73 m2 (%) 7.4 5.7 3.9 4.1 4.9 4.2
GFR <15 ml/min/1.73 m2 (%) 1.3 1.0 0.7 0.6 0.6 1.1
Hematocrit (mean) 38.1 39.5 40.9 41.0 40.4 40.6 <0.0001
WHO anemia (%) 38.4 30.8 22.9 20.1 21.9 20.8 <0.0001
Hemoglobin (mean) 12.7 13.3 13.8 13.8 13.6 13.6 <0.0001
Albumin (mean) 3.60 3.72 3.78 3.78 3.73 3.71 <0.0001
Glucose (mean) 162.9 172.5 180.8 194.1 201.6 209.0 <0.0001
APACHE 2 (mean) 10.2 9.4 9.0 9.0 9.1 9.3 <0.0001
Treatment at discharge
ACE inhibitor (%) 32.5 34.3 34.3 38.5 38.8 42.6 <0.0001
β-blocker (%) 23.8 35.3 40.9 41.3 39.1 35.8 <0.0001
Hospital technology index (%) <0.0001
Level 0 37.2 34.3 31.7 33.2 33.4 30.9
Level 1 20.6 19.3 18.1 17.8 18.9 20.8
Level 2 5.0 4.8 4.5 4.4 4.4 6.3
Level 3 37.2 41.5 45.7 44.6 43.3 41.9
Level 0: no cardiac catherization (cath), PTCA, or CABG services, Level 1: cardiac cath, Level 2: cardiac cath + PTCA, Level 3: cath + PTCA + CABG.
ACE, angiotensin-converting enzyme; APACHE, Acute Physiology and Chronic Health Evaluation; CABG, coronary artery bypass graft; CHF, congestive heart failure;
GFR, glomerular filtration rate; PTCA, percutaneous transluminal coronary angioplasty; WHO, World Health Organization.
OBESITY | VOLUME 17 NUMBER 12 | DECEMBER 2009 2219
articles
epidemiology
During follow-up through June 2004, 78.8% of African
American and 76.0% of white patients died. Crude death
rates were greatest for underweight patients among both
African Americans (315 deaths/1,000 PY) and whites
(317 deaths/1,000 PY) followed by normal weight patients
(198/1,000 PY for blacks and 172/1,000 PY for whites)
(Table 4). Comparatively, lower mortality rates were observed
among overweight and obese patients for both races (Table 4).
When the data were tted to an age and sex adjusted Cox pro-
portional hazards model (Ta b le 4 ), signicantly protective
HRs were observed among whites who were overweight, class
I, or class II obese but not among those who were most obese
(Table 4). A similar trend was observed among black subjects
and the likelihood-ratio test for interaction was not signicant
(P = 0.08).
Results changed little with multivariable adjustment and
the test for interaction remained nonsignicant (P = 0.08)
(Ta bl e 4 ). In the nal multivariable adjusted model with the
race interaction terms removed (Ta bl e 3 ), underweight sub-
jects experienced 45% (95% CI: 40–51%) greater mortality rate
compared to normal weight subjects, whereas overweight and
class I obese subjects experienced 15% (95% CI: 17–13%) lower
mortality rates (Table 3). Mortality among class II obese sub-
jects was also signicantly lower (HR: 0.89, 95% CI: 0.84–0.93)
than normal weight subjects but class III obese subjects did not
dier signicantly in mortality from normal weight subjects.
In supplemental analysis of in-hospital mortality prior to dis-
charge, 15.3% of white subjects and 12.5% of black subjects did
not survive to discharge. Compared to normal weight patients,
underweight patients were more likely (odds ratio (OR): 1.37,
95% CI: 1.28–1.48) to die prior to discharge. Death prior to
discharge was less likely among both overweight (OR: 0.81,
95% CI: 0.77–0.84) and obese I (OR: 0.85, 95% CI: 0.80–0.90)
patients. While obese II patients did not dier from normal
weight patients in predischarge death (OR: 0.99, 95% CI: 0.89–
1.10), the most obese patients were more likely to die (OR:1.21,
95% CI: 1.05–1.40).
DISCUSSION
In contrast to previous studies concluding a positive associa-
tion of obesity with incident ESRD among healthy adults, we
report that obesity does not independently increase the risk
of ESRD among blacks or whites over 10 years in an elderly
cohort with CVD. In addition, similar to other reports in dif-
ferent patient populations, we report that obesity is independ-
ently associated with better survival among both black and
white patients.
Previous studies have shown that rates of ESRD increased
in a stepwise manner as BMI increased (2–4). However, these
prior studies were not done in predominantly elderly popula-
tions or in high cardiovascular risk populations. Perhaps, the
eect of BMI on risk of ESRD is modied by age and/or by
prevalent CVD. Possible modiers include the fact that in the
elderly and in CKD, lean body mass is reduced (17,18) and
thus BMI may not be the best clinical anthropometric measure
of fat in this population. Our observed ESRD rates in this pop-
ulation were much higher than those reported by the US Renal
Data System for similar ages, and may be due to the known
increase risk of CKD progression in individuals with CVD.
While it is known that the presence of CVD confers increased
risk of CKD progression to ESRD, our nding in a population
with CVD should not be viewed as a contradiction of other
reports, but as a hypothesis generator that needs to be explored
in prospective studies. We, however, explored the hypothesis
Table 2 ESRD incidence density rates and hazard ratios among African-American and white subjects according to body-weight
class
Body-weight class
Underweight Normal Overweight Obese I Obese II Obese III
African American
ESRD cases/person-years 4/711 79/7,054 85/8,756 44/4,527 15/1,420 3/687
Crude ESRD ratea (95% CI) 563 (179, 1,357) 1,120 (893, 1,388) 971 (780, 1,194) 972 (715, 1,293) 1,056 (614, 1,703) 437 (111, 1,188)
Age and sex adjusted
hazard ratio (95% CI)
0.52 (0.19, 1.41) 1.0 (Ref) 0.85 (0.63, 1.15) 0.86 (0.60, 1.25) 0.96 (0.55, 1.66) 0.37 (0.12, 1.18)
Multivariableb adjusted
hazard ratio (95% CI)
0.67 (0.25, 1.84) 1.0 (Ref) 0.91 (0.67, 1.25) 1.05 (0.72, 1.52) 0.89 (0.51, 1.56) 0.47 (0.15, 1.48)
White
ESRD cases/person-years 23/9,028 391/127,094 489/138,401 201/49,966 60/13,203 22/5,242
Crude ESRD ratea (95% CI) 255 (165, 376) 308 (278, 339) 353 (323, 386) 402 (350, 461) 454 (350, 581) 420 (270, 625)
Age and sex adjusted
hazard ratio (95% CI)
0.87 (0.57, 1.33) 1.0 (Ref) 1.08 (0.94, 1.23) 1.22 (1.02, 1.44) 1.37 (1.04, 1.80) 1.26 (0.82, 1.94)
Multivariableb adjusted
hazard ratio (95% CI)
0.88 (0.57, 1.34) 1.0 (Ref) 1.02 (0.89, 1.17) 1.05 (0.88, 1.25) 0.76 (0.58, 1.01) 0.81 (0.53, 1.26)
CI, confidence interval; ESRD, End-Stage Renal Disease.
aPer 100,000 person years. bAdjusted for age, sex, hospital technology index, baseline kidney function, hematocrit, albumin, glucose, diabetes, hypertension,
blood pressure, smoking, prior myocardial infarction, congestive heart failure, stroke, coronary artery bypass graft, percutaneous transluminal coronary angioplasty,
angiotensin-converting enzyme inhibitor use, β-blocker use, Acute Physiology and Chronic Health Evaluation II Score.
2220 VOLUME 17 NUMBER 12 | DECEMBER 2009 | www.obesityjournal.org
articles
epidemiology
that lack of an eect of obesity on ESRD risk in our cohort
may have been due to competing mortality in this high-risk
elderly population. However, we have shown that obesity did
not accelerate mortality but prolonged survival except for the
most obese African Americans.
Our ndings are consistent with those of others reported
in congestive heart failure and ESRD patients in that obesity
provides a survival benet (6,7,19). Previous studies in dialysis
patients suggest that blacks may benet more from obesity than
whites, but other studies are conicting (8). Pei et al. showed
that dierences in patient survival on dialysis exist between
racial groups, but they did not investigate racial dierences in
the eect of obesity on survival (8). Our current analysis does
not indicate that African Americans benet more from being
obese. Reasons for this paradoxical eect of obesity on survival
are unclear but do not appear to be related to other comorbidi-
ties. Perhaps, the subjects enrolled in this study represented a
healthy subset of the obese population who survived an AMI.
However, in our supplemental analysis of in-hospital mortal-
ity adjusted for age, gender, and race, overweight and mildly
obese patients were more likely to survive to discharge than
normal weight patients; and obese II patients were equally
likely to survive. erefore, the notion that a healthy subset
of overweight and obese subjects explains the observed results
in ESRD incidence and postdischarge survival seems unlikely.
Nonetheless, we have identied another population who
appears to not express the normal phenotype of obesity and
increased mortality, as seen in the general population. Another
recent paper reported that the eect of obesity on mortality is
variable according to the specic type of death, i.e., obesity is
associated with CVD mortality but not associated with cancer,
noncancer, and non-CVD mortality (20).
ere are potential explanations for the apparently protective
eect of obesity on survival, including misclassication related
to estimates of obesity based solely on height and weight, (i.e.,
BMI) in populations known to have reduced lean body mass
(17,18). One recent study showed that obesity as dened by
waist-to-hip ratio was associated with increased cardiac events
and mortality in subjects with CKD, whereas BMI was not;
however, for the highest BMI group, there was a protective
Table 3 Multivariable hazard ratios for incident ESRD and all-cause mortality
Hazard ratio for incident
ESRD (95% CI) P value
Hazard ratio for all-cause
mortality (95% CI) P value
Underweight 0.86 (0.58, 1.27) 0.44 1.45 (1.40, 1.51) <0.0001
Normal weight 1.00 (Ref) N/A 1.00 (Ref) N/A
Overweight 1.00 (0.88, 1.13) 0.98 0.85 (0.83, 0.87) <0.0001
Obese I 1.01 (0.86, 1.19) 0.87 0.85 (0.83, 0.87) <0.0001
Obese II 0.77 (0.60, 0.99) 0.04 0.89 (0.84, 0.93) <0.0001
Obese III 0.74 (0.49, 1.11) 0.14 0.95 (0.88, 1.01) 0.11
Age (1 year) 0.93 (0.92, 0.94) <0.0001 1.06 (1.06, 1.06) <0.0001
Male 2.07 (1.87, 2.35) <0.0001 1.22 (1.20, 1.24) <0.0001
Black 2.03 (1.75, 2.37) <0.0001 1.01 (0.97, 1.04) 0.69
Current smoking 1.11 (0.96, 1.30) 0.17 1.35 (1.32, 1.39) <0.0001
Estimated GFR (1 unit increase) 0.93 (0.92, 0.93) <0.0001 0.95 (0.94, 0.95) <0.0001
APACHE score (1 unit increase) 1.01 (1.00, 1.03) 0.05 1.04 (1.04, 1.04) <0.0001
Hematocrit (1% increase) 0.96 (0.95, 0.97) <0.0001 0.98 (0.98, 0.98) <0.0001
Albumin (1 g/dl increase) 0.79 (0.71, 0.88) <0.0001 0.89 (0.87, 0.90) <0.0001
Diabetes 2.07 (1.82, 2.34) <0.0001 1.37 (1.34, 1.40) <0.0001
Glucose at admission (10 mg/dl increase) 0.99 (0.98, 1.00) 0.0006 1.01 (1.01, 1.01) <0.0001
Hypertension 1.47 (1.28, 1.70) <0.0001 1.02 (1.00, 1.04) 0.03
Systolic BP (10 mm Hg increase) 1.21 (1.17, 1.24) <0.0001 0.99 (0.99, 0.99) 0.03
Diastolic BP (per 10 mm Hg increase) 0.94 (0.89, 0.99) 0.03 1.01 (1.01, 1.02) 0.002
Prior stroke 0.84 (0.72, 0.98) 0.03 1.37 (1.34, 1.40) <0.0001
Prior MI 1.18 (1.06, 1.32) 0.005 1.21 (1.20, 1.24) <0.0001
Prior CABG 1.35 (1.18, 1.56) <0.0001 1.28 (1.24, 1.31) <0.0001
Prior PTCA 0.88 (0.74, 1.06) 0.19 0.89 (0.86, 0.92) <0.0001
Prior CHF 1.25 (1.10, 1.42) 0.0005 1.59 (1.56, 1.62) <0.0001
Hospital technology index 0.99 (0.95, 1.04) 0.80 0.95 (0.95, 0.96) <0.0001
APACHE, Acute Physiology and Chronic Health Evaluation; BP, blood pressure; CABG, coronary artery bypass graft; CHF, congestive heart failure; CI, confidence interval;
ESRD, End-Stage Renal Disease; GFR, glomerular filtration rate; MI, myocardial infarction; PTCA, percutaneous transluminal coronary angioplasty.
OBESITY | VOLUME 17 NUMBER 12 | DECEMBER 2009 2221
articles
epidemiology
eect (21). Another paper showed that correction for lean
body mass did not mitigate the eect of BMI on survival in a
dialysis population (22).
Strengths of our study include linkage to the US Renal Data
System to account for all cases of progression to ESRD as well
as a large event rate. Our study has a number of limitations that
should be addressed. First, the Cooperative Cardiovascular
Project data are ~10 years old and it is a retrospective analyses.
Another limitation is the use of serum creatinine value on the
initial day of hospitalization when subjects presented for AMI.
is creatinine measurement may not be reective of actual level
of underlying renal function, and may be aected by variability
of creatinine assays across laboratories. BMI has recognized lim-
itations as an indicator of obesity, particularly in the elderly and
in females; nevertheless, it has been widely used in the medical
literature. In addition, lack of complete nutritional status infor-
mation beyond serum albumin is a recognized limitation of our
study, which may have inuenced mortality. Finally, we did not
have information on cause of death that may have been informa-
tive and may have shown dierent eects according to cause-
specic mortality, as was seen in a recent paper (20).
Although we report that obesity did not predict ESRD
incidence when it prolonged survival in our cohort, the high
burden of hypertension and diabetes that attributed to obesity
still suggests that obese individuals are likely to be at higher
risk of ESRD and death compared to nonobese individuals.
Furthermore, overweight and obese individuals should still
be counseled on physical activity and weight control, due to
the benecial eect of these interventions on diabetes and
hypertension control. However, it is paramount to under-
stand the mechanisms of the obesity “survival paradox” in
specic populations before global recommendations are made
concerning weight control in the ESRD, congestive heart
failure, and currently the elderly with CVD. Further investiga-
tion into the reasons for the paradoxical eect of obesity on
survival in certain populations are urgently needed, given the
high morbidity and mortality as well as aggressive measures
used to eradicate obesity in the general population. Such stud-
ies need to explore how an elderly population with CVD might
modify the association of obesity and risk of ESRD and death.
ACKNOWLEDGMENTS
There was no funding associated with this project. Therefore, the data
collection, analysis, and interpretation of the data were completed solely
by the authors of the manuscript.
DISCLOSURE
The authors declared no conflict of interest.
© 2009 The Obesity Society
REFERENCES
1. McTigue K, Larson JC, Valoski A et al. Mortality and cardiac and vascular
outcomes in extremely obese women. JAMA 2006;296:79–86.
2. Gelber RP, Kurth T, Kausz AT et al. Association between body mass index
and CKD in apparently healthy men. Am J Kidney Dis 2005;46:871–880.
3. Hsu CY, McCulloch CE, Iribarren C, Darbinian J, Go AS. Body mass index
and risk for end-stage renal disease. Ann Intern Med 2006;144:21–28.
4. Iseki K, Ikemiya Y, Kinjo K et al. Body mass index and the risk of
development of end-stage renal disease in a screened cohort. Kidney Int
2004;65:1870–1876.
5. Speckman RA, McClellan WM, Volkova NV et al. Obesity is associated
with family history of ESRD in incident dialysis patients. Am J Kidney Dis
2006;48:50–58.
6. Kalantar-Zadeh K, Abbott KC, Salahudeen AK, Kilpatrick RD, Horwich
TB. Survival advantages of obesity in dialysis patients. Am J Clin Nutr
2005;81:543–554.
7. Kistorp C, Faber J, Galatius S et al. Plasma adiponectin, body mass
index, and mortality in patients with chronic heart failure. Circulation
2005;112:1756–1762.
8. Pei YP, Greenwood CM, Chery AL, Wu GG. Racial differences in survival of
patients on dialysis. Kidney Int 2000;58:1293–1299.
9. Reddan DN, Szczech LA, Klassen PS, Owen WF Jr. Racial inequity in
America’s ESRD program. Semin Dial 2000;13:399–403.
Table 4 Mortality rates and hazard ratios among African-American and white subjects according to body-weight class
Body-weight class
Underweight Normal Overweight Obese I Obese II Obese III
African American
Deaths/person-years 2,941/9,269 1,476/7,460 1,264/9,380 611/4,854 204/1,509 105/722
Crude death ratea (95% CI) 315 (276, 358) 198 (188, 208) 135 (127, 142) 126 (116, 136) 135 (118, 155) 146 (120, 175)
Age and sex adjusted
hazard ratio (95% CI)
1.21 (1.05, 1.39) 1.00 (Ref) 0.90 (0.83, 0.97) 0.93 (0.85, 1.02) 0.88 (0.76, 1.02) 1.05 (0.86, 1.28)
Multivariableb adjusted
hazard ratio (95% CI)
1.34 (1.17, 1.54) 1.00 (Ref) 0.88 (0.78, 0.90) 0.78 (0.71, 0.86) 0.88 (0.76, 1.02) 1.13 (0.93, 1.38)
White
Deaths/person-years 2,941/9,269 22,865/133,170 18,076/146,358 6,309/52,883 1,727/13,944 721/5,522
Crude death ratea (95% CI) 317 (306, 329) 172 (169, 174) 124 (122, 125) 119 (116, 122) 124 (118, 130) 131 (121, 140)
Age and sex adjusted
hazard ratio (95% CI)
1.37 (1.32, 1.42) 1.00 (Ref) 0.88 (0.87, 0.90) 0.91 (0.88, 0.93) 0.94 (0.89, 0.98) 0.99 (0.92, 1.06)
Multivariableb adjusted
hazard ratio (95% CI)
1.46 (1.40, 1.52) 1.00 (Ref) 0.85 (0.83, 0.87) 0.85 (0.83, 0.88) 0.89 (0.84, 0.93) 0.92 (0.86, 0.99)
CI, confidence interval; ESRD, End-Stage Renal Disease.
aPer 1,000 person years. bAdjusted for age, sex, hospital technology index, baseline kidney function, hematocrit, albumin, glucose, diabetes, hypertension, blood pressure,
smoking, prior myocardial infarction, congestive heart failure, stroke, coronary artery bypass graft, percutaneous transluminal coronary angioplasty, angiotensin-converting
enzyme inhibitor use, β-blocker use, Acute Physiology and Chronic Health Evaluation.
2222 VOLUME 17 NUMBER 12 | DECEMBER 2009 | www.obesityjournal.org
articles
epidemiology
10. Newsome BB, McClellan WM, Coffey CS et al. Survival advantage of black
patients with kidney disease after acute myocardial infarction. Clin J Am Soc
Nephrol 2006;1:993–999.
11. Marciniak TA, Ellerbeck EF, Radford MJ et al. Improving the quality of care
for Medicare patients with acute myocardial infarction: results from the
Cooperative Cardiovascular Project. [see comment]. JAMA 1998;279:
1351–1357.
12. International Classification of Diseases, Ninth Revision, Clinical Modification.
Public Health Service, US Department of Health and Human Services:
Washington, DC, 1997.
13. Levey AS, Bosch JP, Lewis JB et al. A more accurate method to estimate
glomerular filtration rate from serum creatinine: a new prediction equation.
Modification of Diet in Renal Disease Study Group. Ann Intern Med
1999;130:461–470.
14. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic
kidney disease: evaluation, classification and stratification. Am J Kidney Dis
2002;39(2 Suppl 1):S1–S266.
15. Allison JJ, Kiefe CI, Weissman NW et al. Relationship of hospital teaching
status with quality of care and mortality for Medicare patients with acute MI.
JAMA 2000;284:1256–1262.
16. Overweight, obesity, and health risk. National Task Force on the Prevention
and Treatment of Obesity. Arch Intern Med 2000;160:898–904.
17. Ohkawa S, Odamaki M, Ikegaya N et al. Association of age with muscle
mass, fat mass and fat distribution in non-diabetic haemodialysis patients.
Nephrol Dial Transplant 2005;20:945–951.
18. Mitch WE, Maroni BJ. Factors causing malnutrition in patients with chronic
uremia. Am J Kidney Dis 1999;33:176–179.
19. Kalantar-Zadeh K, Kopple JD, Kilpatrick RD et al. Association of
morbid obesity and weight change over time with cardiovascular
survival in hemodialysis population. Am J Kidney Dis 2005;46:
489–500.
20. Flegal KM, Graubard BI, Williamson DF, Gail MH. Cause-specific excess
deaths associated with underweight, overweight, and obesity. JAMA
2007;298:2028–2037.
21. Elsayed EF, Tighiouart H, Weiner DE et al. Waist-to-hip ratio and body mass
index as risk factors for cardiovascular events in CKD. Am J Kidney Dis
2008;52:49–57.
22. Johansen KL, Young B, Kaysen GA, Chertow GM. Association of body
size with outcomes among patients beginning dialysis. Am J Clin Nutr
2004;80:324–332.
    • "On the other hand, DXA-measured abdominal fat describes more accurately fat distribution and hence may be more suitable to address the question on the effect of central fat versus peripheral fat. We and others have shown that adiposity (Adams et al. 2006; Flicker et al. 2010) and truncal adiposity (Auyeung et al. 2010) may be beneficial for survival in old age, even for those with history of cardiovascular disease (Lea et al. 2009). Our present results showed that in contrary to findings in middle-aged adults, higher proportion of abdominal fat may also be beneficial for survival in men older than 65 years of age. "
    [Show abstract] [Hide abstract] ABSTRACT: In contrast to that in the middle-aged, higher body mass index (BMI) in older people is associated with higher survival rates. Yet, BMI makes no distinction between fat elsewhere and abdominal fat, the latter being metabolically more harmful. We hypothesized that overall adiposity might be protective in old age, but that central fat might offset that benefit and remained harmful as in the middle-aged. Three thousand nine hundred seventy-eight Chinese elderly ≥65 years had demographics, medical conditions, physical activity, and body composition by DXA recorded at baseline. Overall adiposity was measured as whole body fat%, and abdominal adiposity as waist circumference, waist-hip ratio, and relative abdominal fat (RAF) (relative abdominal fat = abdominal fat according to anatomical landmarks/whole body fat). Deaths within 1 year from baseline were excluded from analysis. All-cause and cardiovascular mortality were analyzed using Cox regression, adjusted for covariates. The lowest quintile of adiposity measurements was used for comparison. After a mean follow-up of 72.3 months, 13.7% men and 4.5% women had died. In men, the highest two quintiles of whole body fat % and the upper four quintiles of RAF were associated with significantly lower all-cause mortality, and adjusted hazard ratio (95% CI) in ascending quintiles of RAF compared with the lowest quintile was 0.62 (0.43-0.89), 0.58 (0.4-0.85), 0.52 (0.36-0.77), and 0.67 (0.47-0.96). No relationship was found between abdominal adiposity and cardiovascular mortality in both genders. Higher whole body fat % as well as higher proportion of abdominal fat was associated with lower all-cause mortality in men. No such relation was found in women.
    Full-text · Article · Jun 2011
  • [Show abstract] [Hide abstract] ABSTRACT: Obesity is increasingly recognized as a global pandemic that threatens the health of millions of people. Obesity is considered to be an important cardiovascular risk factor, but there is increasing evidence that patients with elevated body mass index may be better off than others if they develop cardiovascular or renal disease. This phenomenon has been described as the "obesity paradox" or "reverse epidemiology." This article reviews some recent publications that have studied this phenomenon as it relates to heart failure, coronary artery disease, peripheral arterial disease, kidney disease, and a cohort of patients undergoing nonbariatric surgery.
    Article · Apr 2010
  • [Show abstract] [Hide abstract] ABSTRACT: The quality and effectiveness of resuscitation processes may be influenced by the patient's body mass index (BMI); however, the relationship between BMI and survival after in-hospital cardiac arrest has not been previously studied. We evaluated 21 237 adult patients with an in-hospital cardiac arrest within the National Registry for Cardiopulmonary Resuscitation (NRCPR). We examined the association between BMI (classified as underweight [<18.5 kg/m(2)], normal [18.5 to 24.9 kg/m(2)], overweight [25.0 to 29.9 kg/m(2)], obese [30.0 to 34.9 kg/m(2)], and very obese [≥35.0 kg/m(2)]) and survival to hospital discharge using multivariable logistic regression, after stratifying arrests by rhythm type and adjusting for patient characteristics. Of 4499 patients with ventricular fibrillation or pulseless ventricular tachycardia as initial rhythm, 1825 (40.6%) survived to discharge. After multivariable adjustment, compared with overweight patients, underweight (odds ratio [OR], 0.59; 95% confidence interval [CI], 0.41 to 0.84; P=0.003), normal weight (OR, 0.75; 95% CI, 0.63 to 0.89; P<0.001), and very obese (OR, 0.78; 95% CI, 0.63 to 0.96; P=0.02) had lower rates of survival, whereas obese patients had similar rates of survival (OR, 0.87; 95% CI, 0.72 to 1.06; P=0.17). In contrast, of 16 738 patients with arrests caused by asystole or pulseless electric activity, only 2501 (14.9%) survived. After multivariable adjustment, all BMI groups had similar rates of survival except underweight patients (OR, 0.67; 95% CI, 0.54 to 0.82; P<0.001). For cardiac arrest caused by shockable rhythms, underweight, normal weight, and very obese patients had lower rates of survival to discharge. In contrast, for cardiac arrest caused by nonshockable rhythms, survival to discharge was similar across BMI groups except for underweight patients. Future studies are needed to clarify the extent to which BMI affects the quality and effectiveness of resuscitation measures.
    Article · Sep 2010
Show more