Abstract
Rationale & Objective:
Cystatin C-based estimated glomerular filtration rate (eGFRcys) has stronger associations with adverse clinical outcomes than creatinine-based eGFR (eGFRcr). Obesity may be associated with higher cystatin C levels, independent of kidney function, but it is unknown whether obesity modifies associations of eGFRcys with kidney and cardiovascular outcomes.
Study Design:
Cohort study.
Setting & Participants:
27,249 US adults in the Reasons for Geographic and Racial Differences in Stroke Study.
Predictors:
eGFRcys, eGFRcr, waist circumference, and body mass index (BMI).
Outcomes:
All-cause mortality, kidney failure, incident atherosclerotic cardiovascular disease (ASCVD), and incident heart failure (HF).
Analytical Approach:
Multivariable Cox and Fine-Gray models with multiplicative interaction terms were constructed to investigate whether waist circumference quartiles or BMI categories modified associations of eGFRcys with risks of four clinical outcomes.
Results:
Participants had a mean age of 65 years; 54% were women, 41% were Black, and 21% had eGFRcys <60 mL/min/1.73 m2. Baseline prevalence of abdominal obesity (waist circumference ≥88 cm for women or ≥102 cm for men) was 48% and obesity was 38%. In multivariable adjusted analyses, each 15 mL/min/1.73 m2 lower eGFRcys was associated with higher hazard ratios [95% confidence intervals] of mortality in each waist circumference quartile (first quartile, 1.19 [1.15–1.24]; second quartile, 1.22 [1.18–1.26]; third quartile, 1.20 [1.16–1.24]; fourth quartile, 1.19 [1.15–1.23]) as well as within each BMI category (BMI<24.9: 1.21 [1.17–1.25]; BMI=25.0–29.9: 1.21 [1.18–1.25]; BMI=30.0–34.9: 1.20 [1.16–1.25]; BMI≥35: 1.17, [1.12–1.22]). Neither waist circumference nor BMI modified the associations of eGFRcys with mortality, kidney failure, incident ASCVD, and incident HF (all p-interactions >0.05).
Limitations:
Included only Black and White persons in the US.
Conclusion:
Obesity did not modify the associations of eGFRcys with all-cause mortality, kidney failure, incident ASCVD, and incident HF. Among individuals with obesity, cystatin C may be used to provide eGFR-based risk prognostication for adverse outcomes.
Keywords: estimated glomerular filtration rate, cystatin C, obesity, chronic kidney disease, cardiovascular disease
Plain Language Summary
Cystatin C is increasingly used in clinical practice to estimate kidney function and cystatin C-based eGFR (eGFRcys) may be used to determine risk for adverse clinical outcomes. Adiposity may increase serum levels of cystatin C, independent of kidney function. This cohort study investigated whether associations of eGFRcys with adverse kidney and cardiovascular outcomes are modified by measures of obesity, waist circumference and body mass index. We found that obesity does not modify associations of eGFRcys with four clinical outcomes and conclude that among individuals with obesity, cystatin C may be used to provide eGFR-based risk prognostication for adverse outcomes.
Introduction
To improve equity in the diagnosis and management of patients with kidney disease, the National Kidney Foundation and American Society of Nephrology (NKF/ASN) Task Force has recommended increasing adoption of cystatin C in clinical practice.1 Over the past 20 years, multiple studies have shown that cystatin C-based estimated glomerular filtration rate (eGFRcys) has stronger and more linear associations with mortality and cardiovascular disease (CVD) events than creatinine-based eGFR (eGFRcr).2–8 However, because adiposity may be associated with elevated cystatin C levels, independent of kidney function,9–12 some have questioned whether cystatin C may be used to estimate GFR among individuals who are obese.
Currently, over 42% of individuals living in the United States (US) are obese.13 Risk prognostication is an important component of chronic kidney disease (CKD) management, and the eGFR thresholds that define CKD stages are largely based upon risks for mortality, kidney failure, and CVD events.14,15 To our knowledge, whether or not obesity modifies the associations of eGFRcys with these clinical outcomes has not been well evaluated.
The objectives of this study were to: 1) investigate whether the associations of eGFRcys with all-cause mortality, kidney failure, incident atherosclerotic cardiovascular disease (ASCVD), and incident heart failure (HF) are modified by baseline indicators of obesity-waist circumference and body mass index (BMI); and 2) compare the associations of eGFRcys and eGFRcr with these outcomes across wide ranges of waist circumference and BMI values.
Methods
Study population
The REGARDS Study is a national prospective cohort of 30,239 participants that was designed to examine underlying causes for racial and regional differences in stroke mortality in the US. Details on the design and methods of REGARDS have been previously described (Figure S1).16 Briefly, between 2003 and 2007, REGARDS enrolled community-dwelling Black and White adults ≥45 years old. Participants were identified via mail and telephone using commercially available lists of US residents. By design, approximately 50% of the sample was recruited in the ‘stroke belt’ states: North Carolina, South Carolina, Georgia, Tennessee, Mississippi, Alabama, Louisiana, and Arkansas. Trained interviewers conducted computer-assisted telephone interviews to obtain information regarding sociodemographic characteristics, medical history, health behaviors, and medication use. Exclusion criteria included race other than Black or White, active treatment for cancer, chronic medical conditions precluding long-term participation, cognitive impairment, residence in or awaiting nursing home placement, or inability to communicate in English. An in-home study visit was then conducted to obtain blood pressure, height, weight, and waist circumference measurements; electrocardiogram; and blood and urine samples. Follow-up telephone calls were performed every 6 months to assess study endpoints.
For these analyses, we selected participants who had available baseline cystatin C, creatinine, waist circumference, height, and weight values. Individuals with a history of kidney failure at the baseline visit, defined as prevalent dialysis or receipt of a kidney transplant, were excluded. The study was approved by all institutional review boards at the participating centers and all participants provided written informed consent.
Primary Predictors
The primary predictors of interest were eGFRcys and waist circumference, a measure of abdominal obesity. Secondary predictors included eGFRcr and BMI. Blood was collected from participants after a 12-hour fast. Cystatin C was measured by particle-enhanced immunonephelometry (N Latex Cystatin C on the BNII, Siemens AG) and calibrated to the international standard.17 Serum creatinine was measured and calibrated to isotope dilution mass spectrometry-traceable methods.18 The 2012 CKD Epidemiology Collaboration (CKD-EPI) cystatin C equation19 was used to calculate eGFRcys, and the 2021 CKD-EPI creatinine equation20 was used to calculate eGFRcr.
Weight, height, and waist circumference were measured following a standardized protocol. BMI, a measure of obesity, was calculated as weight in kilograms divided by height in meters squared. Waist circumference, a measure of abdominal obesity, was measured using a tape measure midway between the lowest rib and the iliac crest while the participant was standing.
Outcomes
The primary outcome was all-cause mortality, which was ascertained by telephone contact with next-of-kin or through online searches of the Social Security Administration’s Death Master File or the National Death Index.16
Secondary outcomes included kidney failure, incident ASCVD, and incident HF. Kidney failure was defined as initiation of dialysis or receipt of a kidney transplant and was ascertained by linkage to the US Renal Data System, which collects information on persons initiating dialysis or receiving kidney transplants.21 ASCVD was defined as definite or probable myocardial infarction (MI), ischemic stroke, or CVD death; CVD death was defined as death within 28 days of an adjudicated definite or probable MI, or sudden death preceded by cardiac signs or symptoms without evidence of noncoronary causes. Incident HF was defined as first HF hospitalization. Report of non-fatal CVD events during follow-up triggered medical record retrieval and adjudication of study endpoints by clinician experts following published guidelines.22–24 Reports of fatal CVD events prompted two trained adjudicators to determine cause of death through interviews with next-of-kin or proxies and reviewing all available information in medical records, death certificates, and autopsy reports.
Outcomes were adjudicated from study entry until administrative censoring on December 13, 2018 for kidney failure, ASCVD, and HF; and December 22, 2020 for all-cause mortality.
Covariates
Age, sex, race, smoking history, and income were determined by self-report. Diabetes was defined by self-report of prior diagnosis and use of insulin or oral hypoglycemic agents, fasting blood glucose of 126 mg/dL or higher, or non-fasting blood glucose concentration of 200 mg/dL or higher. Hypertension was defined by self-report of prior diagnosis and use of antihypertensive medications, systolic blood pressure (SBP) ≥130 mm Hg, or diastolic blood pressure (DBP) ≥90 mm Hg at baseline. Baseline SBP and DBP were defined by the average of two blood pressure measurements obtained using an aneroid sphygmomanometer (American Diagnostic Corporation, Hauppauge, NY) after 5 minutes of rest. Hyperlipidemia was determined by self-reported diagnosis of “high cholesterol” by a health professional and statin medication use, or low-density lipoprotein (LDL) ≥130 mg/dL. Coronary artery disease (CAD) was defined by self-report of prior MI, coronary revascularization (percutaneous coronary intervention or coronary artery bypass surgery), or evidence of MI on study electrocardiogram. Peripheral artery disease (PAD) was defined by self-report of a procedure to fix the arteries in the leg or history of leg amputation. History of stroke was identified by self-report. Aspirin use was determined by pill bottle review during the in-home study visit.
Blood and urine specimens were collected, processed locally, and sent to a central laboratory for measurements. C-reactive protein (CRP) was measured by particle–enhanced immunonephelometry using the BNII nephelometer (N High Sensitivity CRP, Dade Behring Inc). Urine albumin concentration was measured with a BNII ProSpec nephelometer (Siemens), and urine creatinine concentration was measured by the Jaffé rate method (Roche/Hitachi), and the urine albumin-to-creatinine ratio (UACR) was calculated (mg/g).
Statistical analyses
We summarized baseline characteristics stratified by quartiles of waist circumference. Cox proportional hazards models evaluated associations of eGFRcys, scaled per 15 mL/min/1.73 m2 lower, with all-cause mortality, stratified by quartiles of waist circumference. Multivariable adjusted models included: age, sex, race, income, diabetes, hypertension, hyperlipidemia, history of CAD, PAD, or stroke, smoking, aspirin use, LDL, high-density lipoprotein (HDL), log CRP, and log UACR. To investigate the associations of eGFRcys with kidney failure, we adjusted for these same covariates in Fine-Gray proportional subhazards models, accounting for death as a competing risk.
For incident ASCVD and incident HF, we excluded participants with prevalent ASCVD and HF, respectively. We constructed Fine-Gray proportional subhazards models to evaluate the associations of eGFRcys with incident ASCVD and incident HF as separate outcomes, accounting for non-CVD death as a competing risk for ASCVD and all-cause death as a competing risk for incident HF. Multivariable adjusted models for incident ASCVD and incident HF included: age, sex, race, income, diabetes, hypertension, hyperlipidemia, PAD, smoking, aspirin use, LDL, HDL, log CRP, and log UACR. For the outcome of incident HF, we censored participants at kidney failure because discerning whether hospitalization for fluid overload is due to a true HF event or inadequate ultrafiltration during dialysis is challenging after kidney failure ensues.
The interactions between waist circumference and eGFRcys for all-cause mortality, kidney failure, incident ASCVD, and incident HF were assessed by including a multiplicative interaction term (eGFRcys x waist circumference quartile) in the multivariable adjusted models. Interaction tests were also performed for a binary categorization of waist circumference, using the American Heart Association and National Heart, Lung, and Blood Institute (AHA/NHLBI) definition of abdominal obesity: <88 cm for women and <102 cm for men (non-obese), or ≥88 cm for women and ≥102 cm for men (obese).
We applied the same approach to assess interactions between BMI and eGFRcys for the risk of the four clinical outcomes. We used BMI categories defined by the World Health Organization (WHO): < 24.9 (underweight to normal weight), 25.0–29.9 (overweight), 30.0–34.9 (obesity class I), and ≥ 35.0 kg/m2 (obesity class II and III).
We repeated all analyses with eGFRcr as the predictor of interest. We compared associations of eGFRcys versus eGFRcr with clinical outcomes, stratified by waist or BMI category, by using Wald chi-squared tests.
Proportional hazards assumptions were assessed using Schoenfeld residuals. We identified non-proportional hazards in our models evaluating the associations of eGFR with mortality and ESKD. Nonetheless, we chose to present the overall hazard ratios, which represent the weighted average of the true hazard ratios over the entire follow-up period.25 All baseline variables had <4% missing. All tests were two-tailed with a statistical significance level of P <0.05. Statistical analyses were performed using SAS software, version 9.4 (SAS Institute Inc).
Results
Among 27,249 REGARDS study participants, mean (SD) age at enrollment was 65 (9), 14,811 (54%) were women, and 11,050 (41%) were Black. Median BMI was 28.3 mg/kg2, waist circumference was 95.3 cm, eGFRcys was 80 mL/min/1.73 m2, and eGFRcr was 88 mL/min/1.73 m2. Approximately 21% of participants had eGFRcys <60 mL/min/1.73 m2, 11% had eGFRcr <60 mL/min/1.73 m2, and 14% had UACR >30 mg/g (Table 1).
Table 1:
Summary of baseline characteristics in REGARDS participants by waist circumference quartile
| Baseline Characteristic | Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | Overall |
|---|---|---|---|---|---|
| Range of waist circumference | Females: 33–81 Males: 53–91 |
Females: 81–91 Males: 91–99 |
Females: 91–102 Males: 99–107 |
Females: 102–272 Males: 107–340 |
Females: 33–272 Males: 53–340 |
| N | 6,263 | 7,065 | 7,257 | 6,664 | 27,249 |
| Age, years | 65 ±10 | 65 ±10 | 65 ±9 | 64 ±9 | 65 ±9 |
| Female | 3,460 (55%) | 3,699 (52%) | 3,977 (55%) | 3,675 (55%) | 14,811 (54%) |
| Black race | 2,035 (32%) | 2,668 (38%) | 3,075 (42%) | 3,272 (49%) | 11,050 (41%) |
| Income | |||||
| <$20K | 929 (17%) | 1,083 (18%) | 1,253 (20%) | 1,494 (25%) | 4,759 (20%) |
| $20-$34K | 1,397 (26%) | 1,701 (28%) | 1,796 (28%) | 1,693 (29%) | 6,587 (28%) |
| $35-$74K | 1,882 (34%) | 2,202 (36%) | 2,224 (35%) | 1,888 (32%) | 8,196 (34%) |
| ≥$75K | 1,249 (23%) | 1,189 (19%) | 1,108 (17%) | 852 (14%) | 4,398 (18%) |
| Cigarette smoking | |||||
| Current | 1,143 (18%) | 988 (14%) | 980 (14%) | 851 (13%) | 3,962 (15%) |
| Past | 2,091 (34%) | 2,832 (40%) | 3,069 (42%) | 2,948 (44%) | 10,940 (40%) |
| Never | 3,003 (48%) | 3,225 (46%) | 3,182 (44%) | 2,838 (43%) | 12,248 (45%) |
| Hyperlipidemia | 3,000 (49%) | 4,192 (60%) | 4,415 (62%) | 3,965 (61%) | 15,572 (58%) |
| Diabetes | 480 (8%) | 944 (14%) | 1,610 (23%) | 2,520 (39%) | 5,554 (21%) |
| Hypertension | 3,075 (50%) | 4,464 (64%) | 5,184 (73%) | 5,445 (83%) | 18,168 (68%) |
| Coronary artery disease | 862 (14%) | 1,138 (16%) | 1,369 (19%) | 1,342 (21%) | 4,711 (18%) |
| History of stroke | 303 (5%) | 386 (5%) | 455 (6%) | 496 (7%) | 1,640 (6%) |
| Peripheral artery disease | 114 (1.8%) | 147 (2.1%) | 173 (2.4%) | 136 (2.0%) | 570 (2.1%) |
| Aspirin | 23,60 (38%) | 3,017 (43%) | 3,268 (45%) | 3,129 (47%) | 11,774 (43%) |
| Systolic BP, mmHg | 121 (111, 131) | 124 (117, 136) | 127 (119, 138) | 130 (121, 140) | 126 (118, 137) |
| Diastolic BP, mmHg | 74 (68, 80) | 77 (70, 81) | 78 (71, 82) | 79 (72, 84) | 78 (70, 82) |
| LDL, mg/dL | 111 (90, 134) | 113 (92, 137) | 113 (90, 137) | 109 (87, 133) | 111 (90, 135) |
| HDL, mg/dL | 57 (46, 70) | 50 (41, 62) | 47 (39, 57) | 45 (38, 54) | 49 (40, 61) |
| BMI, kg/m2 | 23 (22, 25) | 27 (25, 29) | 30 (28, 32) | 35 (32, 39) | 28 (25, 32) |
| CRP, mg/L | 1.2 (0.6, 2.8) | 1.8 (0.9, 4.0) | 2.5 (1.2, 5.3) | 3.9 (1.8, 8.1) | 2.2 (1.0, 5.0) |
| UACR, mg/g | 6.7 (4.5, 12.6) | 6.9 (4.4, 13.3) | 7.4 (4.7, 15.5) | 8.6 (5.0, 22.2) | 7.3 (4.6, 15.3) |
| UACR >30 mg/g | 647 (11%) | 790 (12%) | 994 (14%) | 1,309 (20%) | 3,740 (14%) |
| UACR >300 mg/g | 102 (1.7%) | 121 (1.8%) | 165 (2.4%) | 260 (4.0%) | 648 (2.5%) |
| eGFRcys, mL/min/1.73 m2 | 87 (70, 101) | 81 (65, 97) | 78 (62, 94) | 74 (58, 89) | 80 (63, 96) |
| eGFRcys <60 mL/min/1.73 m2 | 896 (14%) | 1,274 (18%) | 1,585 (22%) | 1,858 (28%) | 5,613 (21%) |
| eGFRcr, mL/min/1.73 m2 | 90 (77, 99) | 88 (74, 98) | 87 (73, 98) | 88 (72, 99) | 88 (74, 99) |
| eGFRcr <60 mL/min/1.73 m2 | 470 (8%) | 710 (10%) | 831 (11%) | 870 (13%) | 2,881 (11%) |
Data displayed represent n (%) for categorical variables, and mean (standard deviation) or median (interquartile range) for continuous variables.
Abbreviations: BMI, body mass index; BP, blood pressure; LDL, low-density lipoprotein; HDL, high-density lipoprotein; CRP, C-reactive protein; UACR, urine albumin-to-creatinine ratio; eGFRcys, cystatin C-based estimated glomerular filtration rate; eGFRcr, creatinine-based estimated glomerular filtration rate
In the overall cohort, 48% of participants had abdominal obesity by waist circumference (≥88 cm for women and ≥102 cm for men), and 38% of participants were obese by BMI (≥30 kg/m2). Participants with larger waist circumferences were more likely to be Black and had higher prevalence of nearly all baseline comorbidities assessed, including hyperlipidemia, diabetes, hypertension, CAD, and history of stroke (Table 1). Participants in higher waist circumference quartiles also had lower baseline eGFRcys values and higher SBP and UACR.
During a median follow-up time of 12.1 years (interquartile range [IQR] 6.7–15.0), 9,350 (34%) participants died, 511 (2%) developed kidney failure, 2,271 (11%) had an incident ASCVD event, and 1,053 (4%) had an incident HF hospitalization. Among the 9,350 participants who died, median baseline eGFRcr was 81 mL/min/1.73 m2 (IQR 64–93) and eGFRcys was 66 mL/min/1.73 m2 (IQR 51–82). Among the 511 participants who progressed to kidney failure during follow-up, median baseline eGFRcr was 52 mL/min/1.73 m2 (IQR 34–75) and eGFRcys was 44 mL/min/1.73 m2 (IQR 29–62).
Effect modification of categorical measures of obesity on eGFR associations with clinical outcomes
In multivariable adjusted analyses, each 15 mL/min/1.73 m2 lower eGFRcys was associated with 20% higher risk of mortality, 126% higher risk of kidney failure, and 9% higher risks of both incident ASCVD and incident HF. These associations were maintained across all waist circumference quartiles (Table 2). Neither waist circumference quartile nor AHA/NHLBI waist circumference category modified associations of eGFRcys with any of the four clinical outcomes (Table 2, Table S1).
Table 2.
Multivariable-adjusted associations of eGFRcys with four clinical outcomes (all-cause mortality, kidney failure, incident ASCVD, and incident HF), stratified by quartile of waist circumference
| All participants | Hazard and Subhazard ratios (95% CI) by quartile of waist circumference a | P interaction b | ||||
|---|---|---|---|---|---|---|
| Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | |||
| All-cause mortality | ||||||
| Proportion w/event (%) | 9,350 / 27,249 (34) | 1,999 / 6,263 (32) | 2,305 / 7,065 (33) | 2,474 / 7,257 (34) | 2,572 / 6,664 (39) | - |
| eGFRcys | 1.20 (1.18, 1.22), p<0.001 | 1.19 (1.15, 1.24), p<0.001 | 1.22 (1.18, 1.26), p<0.001 | 1.20 (1.16, 1.24), p<0.001 | 1.19 (1.15, 1.23), p<0.001 | p=0.7 |
| Kidney failure | ||||||
| Proportion w/event (%) | 511/ 27,249 (2) | 49 / 6,263 (0.8) | 107 / 7,065 (1.5) | 135 / 7,257 (1.9) | 220 / 6,664 (3.3) | - |
| eGFRcys | 2.26 (2.04, 2.49), p<0.001 | 2.34 (1.84, 2.99), p<0.001 | 2.21 (1.84, 2.65), p<0.001 | 2.53 (2.15, 2.99), p<0.001 | 2.08 (1.82, 2.37), p<0.001 | p=0.3 |
| Incident ASCVD | ||||||
| Proportion w/event (%) | 2,271 / 21,239 (11) | 425 / 5,137 (8) | 528 / 5,415 (10) | 580 / 5,392 (11) | 738 / 5,385 (14) | - |
| eGFRcys | 1.09 (1.05, 1.13), p<0.001 | 1.13 (1.05, 1.22), p=0.001 | 1.06 (0.99, 1.13), p=0.1 | 1.10 (1.03, 1.17), p=0.003 | 1.07 (1.01, 1.13), p=0.02 | P=0.4 |
| Incident HF | ||||||
| Proportion w/event (%) | 1,018/24,348 (4.2%) | 150/5,912 (2.5) | 217/6,232 (3.5) | 282/5,863 (4.8) | 369/6,341 (5.8) | |
| eGFRcys | 1.09 (1.03, 1.16), p=0.003 | 1.08 (0.96, 1.23), p=0.2 | 1.11 (1.00, 1.23), p=0.04 | 1.14 (1.04, 1.24), p=0.007 | 1.05 (0.97, 1.14), p=0.2 | p=0.6 |
Abbreviations: eGFRcys, cystatin C-based estimated glomerular filtration rate; ASCVD, atherosclerotic cardiovascular disease; HF, heart failure
eGFR is scaled per 15 mL/min/1.73 m2 lower
p for interaction of waist circumference with eGFRcys
Models for mortality and kidney failure adjusted for age, sex, race, income, diabetes, hypertension, hyperlipidemia, coronary artery disease, peripheral artery disease, history of stroke, smoking, aspirin use, low-density lipoprotein, high-density lipoprotein, C-reactive protein, urine albumin-to-creatinine ratio. Models for incident ASCVD and HF adjusted for age, sex, race, income, diabetes, hypertension, hyperlipidemia, peripheral artery disease, smoking, aspirin use, low-density lipoprotein, high-density lipoprotein, C-reactive protein, urine albumin-to-creatinine ratio
Among participants who were not overweight or obese (BMI < 24.9 kg/m2), each 15 mL/min/1.73 m2 lower eGFRcys was associated with 21% higher risk of mortality, 123% higher risk of kidney failure, 10% higher risk of incident ASCVD, and approximately 9% higher risk of incident HF (Table 3). BMI category did not modify associations of eGFRcys with and any the four clinical outcomes (Table 3). Waist circumference and BMI category also did not modify associations of eGFRcr with clinical outcomes (Tables S1–S3).
Table 3.
Multivariable-adjusted associations of eGFRcys with four clinical outcomes (all-cause mortality, kidney failure, incident ASCVD, and incident HF), stratified by BMI category
| Hazard and Subhazard ratios (95% CI) by BMI category a | P interaction b | ||||
|---|---|---|---|---|---|
| BMI < 24.9 | BMI 25.0–29.9 | BMI 30.0–34.9 | BMI ≥35 | ||
| All-cause mortality | |||||
| Proportion w/event (%) | 2,637 / 6,766 (39) | 3,429 / 10,171 (34) | 1,893 / 6,033 (31) | 1,391 / 4,279 (33) | |
| eGFRcys | 1.21 (1.17, 1.25), p<0.001 | 1.21 (1.18, 1.25), p<0.001 | 1.20 (1.16, 1.25), p<0.001 | 1.17 (1.12, 1.22), p<0.001 | p=0.5 |
| Kidney failure | |||||
| Proportion w/event (%) | 73 / 6,766 (1.1) | 154 / 10,171 (1.5) | 147 / 6,033 (2.4) | 137 / 4,279 (3.2) | |
| eGFRcys | 2.23 (1.80, 2.77), p<0.001 | 2.35 (2.01, 2.75), p<0.001 | 2.16 (1.86, 2.50), p<0.001 | 2.32 (1.96, 2.75), p<0.001 | p=0.8 |
| Incident ASCVD | |||||
| Proportion w/event (%) | 521 / 5,345 (10) | 861 / 7,892 (11) | 504 / 4,710 (11) | 385 / 3,382 (11) | |
| eGFRcys | 1.10 (1.03, 1.18), p=0.003 | 1.09 (1.03, 1.16), p=0.002 | 1.09 (1.02, 1.16), p=0.01 | 1.08 (1.01, 1.16), p=0.03 | p=0.9 |
| Incident HF | |||||
| Proportion w/event (%) | 202/6,319 (3.2) | 394/9,291 (4.2) | 226/5,319 (4.2) | 196/3,419 (5.7) | |
| eGFRcys | 1.09 (0.98, 1.22), p=0.1 | 1.08 (1.00, 1.17), p=0.04 | 1.09 (0.99, 1.20), p=0.08 | 1.11 (0.99, 1.24), p=0.08 | p=0.9 |
Abbreviations: BMI, body mass index in kg per m2; eGFRcys, cystatin C-based estimated glomerular filtration rate; ASCVD, atherosclerotic cardiovascular disease; HF, heart failure
eGFR is scaled per 15 mL/min/1.73 m2 lower
p for interaction of BMI category with eGFR
Models for mortality and kidney failure adjusted for age, sex, race, income, diabetes, hypertension, hyperlipidemia, coronary artery disease, peripheral artery disease, history of stroke, smoking, aspirin use, low-density lipoprotein, high-density lipoprotein, C-reactive protein, urine albumin-to-creatinine ratio. Models for incident ASCVD and HF adjusted for age, sex, race, income, diabetes, hypertension, hyperlipidemia, peripheral artery disease, smoking, aspirin use, low-density lipoprotein, high-density lipoprotein, C-reactive protein, urine albumin-to-creatinine ratio
Across all waist circumference quartiles and BMI categories, eGFRcys had stronger associations with mortality compared with eGFRcr (p<0.001 within each quartile or category). Below an eGFRcys of approximately 75 to 80 mL/min/1.73 m2, lower eGFRcys was associated with higher risk of all-cause mortality within each waist circumference quartile (Figure 1A). Conversely, the threshold at which lower eGFRcr was associated with higher risk of mortality varied widely by waist circumference quartile (Figure 1B). In all quartiles of waist circumference and BMI categories, there were no statistically significant differences between the associations of eGFRcys versus eGFRcr with kidney failure.
Figure 1.
Spline plots of multivariable-adjusted associations of eGFRcys (A) and eGFRcr (B) with mortality by quartile of waist circumference.
Model diagnostics
We identified non-proportional hazards in our models evaluating the associations of eGFR with mortality and kidney failure. Lower eGFRcr and eGFRcys, were associated with progressively higher relative risks of mortality and kidney failure during follow-up. For example, each 15 mL/min/1.73 m2 lower eGFRcr was associated with a hazard ratio of 1.01 for mortality at one year of follow-up and a hazard ratio of 1.13 at 15 years of follow-up.
Discussion
In this large community-based population of Black and White individuals with a high burden of obesity, eGFRcys maintained consistent associations with mortality, kidney failure, incident ASCVD, and incident HF across a wide spectrum of waist circumference and BMI values. The association of eGFRcys with mortality was stronger than that of eGFRcr in all waist circumference quartiles and BMI categories. These findings indicate that while obesity may be a non-GFR determinant of cystatin C, it does not influence the prognostic strength of eGFRcys for determining risks of adverse clinical outcomes.
In 2021, the NKF/ASN Task Force recommended increased adoption of cystatin C to estimate kidney function. 1,26 However, because prior studies have identified adiposity as a non-GFR determinant of serum cystatin C levels, some have questioned whether cystatin C may be used to estimate GFR among individuals who are obese.9,27–29 Human preadipocytes release cystatin C,30 and one study found that obese individuals had 3-fold higher levels of cystatin C mRNA in adipose tissue compared with non-obese individuals.29 While cystatin C has repeatedly been shown to be a stronger prognostic biomarker than creatinine among both general and CKD populations,2–8 to our knowledge, no prior study has investigated whether obesity influences the prognostic value of cystatin C. Our findings that neither waist circumference nor BMI categories modify associations of eGFRcys with risks of mortality, kidney failure, incident ASCVD, and incident HF have important clinical implications given the growing prevalence of obesity globally.13,31
Prior prognosis studies demonstrating that eGFRcys has stronger associations with clinical outcomes than eGFRcr are discrepant from GFR measurement studies, which have found that eGFRcys and eGFRcr have similar performance in estimating measured GFR (mGFR).19,20 We hypothesize that these differences largely arise from the inclusion of older and more chronically ill participants in prognosis studies compared with GFR measurement studies. However, some have speculated that cystatin C is associated with cardiovascular risk factors, such as obesity, that may enhance the prognostic strength of eGFRcys, independent of kidney function.32,33 The results from our analyses do not support this latter concept, as we found that the prognostic strength of eGFRcys did not vary across clinically meaningful categories of waist circumference and BMI.
To our knowledge, this study is the first to demonstrate that obesity does not influence the prognostic strength of cystatin C for estimating risks of adverse events related to kidney disease. Our study includes a large, national population with standardized collection of measures of body habitus and data on multiple long-term clinical outcomes that are associated with kidney dysfunction. This study also has important limitations. First, as with most prospective cohort studies in the general population, mGFR was not available. Second, because the REGARDS study included self-identified Black and White persons in the US and excluded non-English speakers, results may not be generalizable to other racial or ethnic populations or to populations residing outside of the US. Third, we incorporated both waist circumference and BMI as measures of adiposity but did not have more precise measures of body fat. Fourth, several covariates used for statistical adjustment, such as diabetes and stroke were partially or fully determined by self-report.
Obesity does not modify the associations of eGFRcys with risks of mortality, kidney failure, incident ASCVD, and incident HF among community-living Black and White individuals. Across a wide spectrum of waist circumference and BMI values, eGFRcys had stronger associations with mortality compared with eGFRcr.
Supplementary Material
Figure S1. Flow diagram depicting the selection of REGARDS participants for each analytic dataset
Table S1. Multivariable-adjusted associations of eGFR with four clinical outcomes (all-cause mortality, kidney failure, incident ASCVD, and incident HF), stratified by AHA/NHLBI waist circumference category
Table S2. Multivariable-adjusted associations of eGFRcr with four clinical outcomes (all-cause mortality, kidney failure, incident ASCVD, and incident HF), stratified by quartile of waist circumference
Table S3. Multivariable-adjusted associations of eGFRcr with four clinical outcomes (all-cause mortality, kidney failure, incident ASCVD, and incident HF), stratified by BMI category
Support:
This research project is supported by cooperative agreement U01 NS041588 co-funded by the National Institute of Neurological Disorders and Stroke (NINDS) and the National Institute on Aging (NIA), National Institutes of Health, Department of Health and Human Service. MGS and MME are supported by SD-20-387 from the Department of Veterans Affairs. T. Lori is funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) K23DK119542 and the Department of Medicine, Boston Medical Center. The funders had no role in study design, data collection, analysis, reporting or the decision to submit for publication. The authors thank the other investigators, the staff, and the participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at: https://www.uab.edu/soph/regardsstudy/
DCC (Chen) is an employee at Genentech/Roche, MME and MGS receive research funding from Bayer, Inc. MME has received an honorarium from Boehringer-Ingelheim, Inc. MGS reports honoraria from Bayer, Inc., Boehringer-Ingelheim, and AztraZeneca, and served as a consultant to Cricket Health and Intercept Pharmaceuticals. MGS previously served as an advisor to and held stock in TAI Diagnostics. JHI receives research funding from Baxter International and the Juvenile Diabetes Research Foundation, and is a member of the Data Safety Monitoring Board for Sanifit International and the Advisory Board for Ardelyx Inc., AztraZeneca, Akebia, and Bayer. TI receives funding from Vertex Pharmaceuticals. OG reports receiving honoraria and grant support from Amgen, honoraria from Akebia and AztraZeneca, and serving on the Data Monitoring Committee for QED pharmaceuticals. DCC (Crews) reports receiving research funding from Baxter International and Somatus.
Footnotes
Financial Disclosure: The remaining authors declare that they have no relevant financial interests.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Delgado C, Baweja M, Crews D, et al. A Unifying Approach for GFR Estimation: Recommendations of the NKF-ASN Task Force on Reassessing the Inclusion of Race in Diagnosing Kidney Disease. Journal of the American Society of Nephrology. 2021:ASN.2021070988. doi: 10.1681/asn.2021070988 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Shlipak MG, Sarnak MJ, Katz R, et al. Cystatin C and the risk of death and cardiovascular events among elderly persons. N Engl J Med. May 19 2005;352(20):2049–60. doi: 10.1056/NEJMoa043161 [DOI] [PubMed] [Google Scholar]
- 3.Shlipak MG, Matsushita K, Ärnlöv J, et al. Cystatin C versus creatinine in determining risk based on kidney function. N Engl J Med. Sep 5 2013;369(10):932–43. doi: 10.1056/NEJMoa1214234 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Peralta CA, Shlipak MG, Judd S, et al. Detection of chronic kidney disease with creatinine, cystatin C, and urine albumin-to-creatinine ratio and association with progression to end-stage renal disease and mortality. Jama. Apr 20 2011;305(15):1545–52. doi: 10.1001/jama.2011.468 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Peralta CA, Katz R, Sarnak MJ, et al. Cystatin C identifies chronic kidney disease patients at higher risk for complications. J Am Soc Nephrol. Jan 2011;22(1):147–55. doi: 10.1681/asn.2010050483 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Menon V, Shlipak MG, Wang X, et al. Cystatin C as a risk factor for outcomes in chronic kidney disease. Ann Intern Med. Jul 3 2007;147(1):19–27. doi: 10.7326/0003-4819-147-1-200707030-00004 [DOI] [PubMed] [Google Scholar]
- 7.Sarnak MJ, Katz R, Stehman-Breen CO, et al. Cystatin C concentration as a risk factor for heart failure in older adults. Ann Intern Med. Apr 5 2005;142(7):497–505. doi: 10.7326/0003-4819-142-7-200504050-00008 [DOI] [PubMed] [Google Scholar]
- 8.Lees JS, Welsh CE, Celis-Morales CA, et al. Glomerular filtration rate by differing measures, albuminuria and prediction of cardiovascular disease, mortality and end-stage kidney disease. Nat Med. Nov 2019;25(11):1753–1760. doi: 10.1038/s41591-019-0627-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Stevens LA, Schmid CH, Greene T, et al. Factors other than glomerular filtration rate affect serum cystatin C levels. Kidney Int. Mar 2009;75(6):652–60. doi: 10.1038/ki.2008.638 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Knight EL, Verhave JC, Spiegelman D, et al. Factors influencing serum cystatin C levels other than renal function and the impact on renal function measurement. Kidney Int. Apr 2004;65(4):1416–21. doi: 10.1111/j.1523-1755.2004.00517.x [DOI] [PubMed] [Google Scholar]
- 11.Muntner P, Winston J, Uribarri J, Mann D, Fox CS. Overweight, obesity, and elevated serum cystatin C levels in adults in the United States. Am J Med. Apr 2008;121(4):341–8. doi: 10.1016/j.amjmed.2008.01.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kottgen A, Russell SD, Loehr LR, et al. Reduced kidney function as a risk factor for incident heart failure: the atherosclerosis risk in communities (ARIC) study. J Am Soc Nephrol. Apr 2007;18(4):1307–15. doi: 10.1681/asn.2006101159 [DOI] [PubMed] [Google Scholar]
- 13.Stierman B National Health and Nutrition Examination Survey 2017–March 2020 Prepandemic Data. National Health Statistics Reports. 2021;(NHSR No. 158) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int Suppl. 2013;3:1–150. [Google Scholar]
- 15.Levey AS, de Jong PE, Coresh J, et al. The definition, classification, and prognosis of chronic kidney disease: a KDIGO Controversies Conference report. Kidney Int. Jul 2011;80(1):17–28. doi: 10.1038/ki.2010.483 [DOI] [PubMed] [Google Scholar]
- 16.Howard VJ, Cushman M, Pulley L, et al. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology. 2005;25(3):135–43. doi: 10.1159/000086678 [DOI] [PubMed] [Google Scholar]
- 17.Erlandsen EJ, Randers E, Kristensen JH. Evaluation of the Dade Behring N Latex Cystatin C assay on the Dade Behring Nephelometer II System. Scand J Clin Lab Invest. Feb 1999;59(1):1–8. doi: 10.1080/00365519950185940 [DOI] [PubMed] [Google Scholar]
- 18.Kurella Tamura M, Wadley V, Yaffe K, et al. Kidney function and cognitive impairment in US adults: the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study. Am J Kidney Dis. 2008;52(2):227–34. doi: 10.1053/j.ajkd.2008.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Inker LA, Schmid CH, Tighiouart H, et al. Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med. Jul 5 2012;367(1):20–9. doi: 10.1056/NEJMoa1114248 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Inker LA, Eneanya ND, Coresh J, et al. New Creatinine- and Cystatin C–Based Equations to Estimate GFR without Race. New England Journal of Medicine. 2021;doi: 10.1056/NEJMoa2102953 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Saran R, Robinson B, Abbott KC, et al. US Renal Data System 2019 Annual Data Report: epidemiology of kidney disease in the United States. Am J Kidney Dis. 2020;75(1)(suppl 1):S1–S64. doi: 10.1053/j.ajkd.2019.09.003 [DOI] [PubMed] [Google Scholar]
- 22.Luepker RV, Apple FS, Christenson RH, et al. Case Definitions for Acute Coronary Heart Disease in Epidemiology and Clinical Research Studies. Circulation. 2003;108(20):2543–2549. doi:doi: 10.1161/01.CIR.0000100560.46946.EA [DOI] [PubMed] [Google Scholar]
- 23.Thygesen K, Alpert JS, Jaffe AS, et al. Fourth Universal Definition of Myocardial Infarction (2018). Circulation. 2018;138(20):e618–e651. doi:doi: 10.1161/CIR.0000000000000617 [DOI] [PubMed] [Google Scholar]
- 24.Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD. Third Universal Definition of Myocardial Infarction. Circulation. 2012;126(16):2020–2035. doi:doi: 10.1161/CIR.0b013e31826e1058 [DOI] [PubMed] [Google Scholar]
- 25.Stensrud MJ, Hernán MA. Why Test for Proportional Hazards? JAMA. 2020;323(14):1401–1402. doi: 10.1001/jama.2020.1267 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hsu C-y, Yang W, Parikh RV, et al. Race, Genetic Ancestry, and Estimating Kidney Function in CKD. New England Journal of Medicine. 2021;doi: 10.1056/NEJMoa2103753 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Muntner P, Winston J, Uribarri J, Mann D, Fox CS. Overweight, Obesity, and Elevated Serum Cystatin C Levels in Adults in the United States. Article. American Journal of Medicine. 2008;121(4):341–348. doi: 10.1016/j.amjmed.2008.01.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Shankar A, Teppala S. Relationship between body mass index and high cystatin levels among US adults. J Clin Hypertens (Greenwich). Dec 2011;13(12):925–30. doi: 10.1111/j.1751-7176.2011.00548.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Naour N, Fellahi S, Renucci JF, et al. Potential contribution of adipose tissue to elevated serum cystatin C in human obesity. Obesity (Silver Spring). Dec 2009;17(12):2121–6. doi: 10.1038/oby.2009.96 [DOI] [PubMed] [Google Scholar]
- 30.Taleb S, Lacasa D, Bastard JP, et al. Cathepsin S, a novel biomarker of adiposity: relevance to atherogenesis. Faseb j. Sep 2005;19(11):1540–2. doi: 10.1096/fj.05-3673fje [DOI] [PubMed] [Google Scholar]
- 31.World Health Organization: Obesity and overweight. Accessed January 22, 2023, https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight
- 32.Rule AD, Bailey KR, Lieske JC, Peyser PA, Turner ST. Estimating the glomerular filtration rate from serum creatinine is better than from cystatin C for evaluating risk factors associated with chronic kidney disease. Kidney Int. Jun 2013;83(6):1169–76. doi: 10.1038/ki.2013.7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Mathisen UD, Melsom T, Ingebretsen OC, et al. Estimated GFR Associates with Cardiovascular Risk Factors Independently of Measured GFR. Journal of the American Society of Nephrology. 2011;22(5):927–937. doi: 10.1681/asn.2010050479 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1. Flow diagram depicting the selection of REGARDS participants for each analytic dataset
Table S1. Multivariable-adjusted associations of eGFR with four clinical outcomes (all-cause mortality, kidney failure, incident ASCVD, and incident HF), stratified by AHA/NHLBI waist circumference category
Table S2. Multivariable-adjusted associations of eGFRcr with four clinical outcomes (all-cause mortality, kidney failure, incident ASCVD, and incident HF), stratified by quartile of waist circumference
Table S3. Multivariable-adjusted associations of eGFRcr with four clinical outcomes (all-cause mortality, kidney failure, incident ASCVD, and incident HF), stratified by BMI category