Abstract
Rationale & Objective:
Little information exists on incidence of and risk factors for chronic kidney disease (CKD) in contemporary US cohorts, and whether risk factors differ by race, sex or region in the US.
Study Design:
Observational cohort study.
Setting & Participants:
4,198 Black and 7,799 White participants age ≥45, recruited 2003-2007 across the continental US with baseline eGFR >60 ml/min/1.73m2 and eGFR measured again ~9 years later.
Exposures:
Age, sex, race (Black/White), region (stroke belt/other), education, income, systolic blood pressure, body mass index (BMI), diabetes, coronary heart disease, hyperlipidemia, smoking, albuminuria.
Outcomes:
1) eGFR change 2) incident CKD, defined as eGFR <60 ml/min/1.73m2 and ≥40% decline from baseline or kidney failure.
Analytic Approach:
Linear regression and modified Poisson regression were used to determine the association of risk factors with eGFR change and incident CKD, and stratified by race, sex and region.
Results:
Participants were 63±8 years old, 54% female, and 35% Black. After 9.4±1.0 years follow-up, 9% developed CKD. In an age-, sex- and race-adjusted model, Black race (β −0.13, p <0.0001) was associated with higher risk of eGFR change, but this was attenuated in the fully-adjusted model (β 0.02, p=0.54). Stroke belt residence was independently associated with eGFR change (β −0.10, p=0.0007) and incident CKD (RR 1.14 95%CI 1.01,1.30). Albuminuria was more strongly associated with eGFR change (β −0.26 vs β −0.17, pinteraction 0.01) in Black compared to White participants. Results were similar for incident CKD.
Limitations:
excluded Hispanics, unknown duration and/or severity of risk factors.
Conclusions:
Established CKD risk factors accounted for higher risk of incident CKD in Black vs. White individuals. Albuminuria was a stronger risk factor for eGFR decline and incident CKD in Black compared to White individuals. Living in the US stroke belt is a novel risk factor for CKD.
Keywords: Chronic kidney disease incidence, risk factors, Race differences, Disparities, Stroke belt, regional differences
Introduction
Chronic kidney disease (CKD) affects 37 million Americans and is the ninth leading cause of death.1 People with CKD experience high rates of disability,2 morbidity and mortality.3 The costs of care associated with CKD are also substantial; Medicare spending alone for all beneficiaries with CKD exceeded $84 billion in 2017.4 Substantial Black-White disparities exist in prevalence of CKD and associated costs.4 Black people with CKD experience greater morbidity, mortality and kidney failure than White people.4 Examination of CKD risk factors could guide strategies to prevent CKD and eliminate Black-White disparities in CKD burden.
Prior studies examined risk factors for incident CKD and/or longitudinal eGFR decline in large, population-based cohorts in the United States (US).5,6 However, few studies had substantial representation of Black individuals and none had geographic diversity, limiting their ability to examine the extent to which standard CKD risk factors may differ by race, the degree to which race/ethnic differences in CKD incidence are explained by traditional CKD risk factors, or the degree to which the risk of incident CKD differs by geographic region. Race is a social construct that has traditionally been used as a proxy for “genetic” biological variability, while it may actually reflect socioeconomic and health consequences of racism(a system of structuring opportunity and assigning value based on the social interpretation of how one looks, referred to as “race”).7 Black-White race disparities in CKD have long been documented, but the key drivers of these disparities have not been fully elucidated. Previous reports were also limited by studying cohorts recruited in the 1980s, and the risk factors for incident CKD may have changed over time. Obesity, for instance, has become much more prevalent and is a risk factor for CKD.8,9 Finally, since one cohort of persons oversampled on presence of albuminuria showed that there were some differences in risk factors for incident CKD, and another large consortium reported that the relationship of kidney measures to mortality were stronger in women than men, we evaluated whether there were differences in risk factors for incident CKD by sex in this US population sample.10,11
The aims of this study were to determine the incidence of and risk factors for eGFR decline and CKD in a large, contemporary, biracial US cohort, and to determine whether risk factors varied by sex, race or region.
Methods
Study Design
The REasons for Geographic and Racial Differences in Stroke (REGARDS) study is a prospective, cohort study of 30,239 Black and White adults age 45 and older in the US, enrolled in 2003-2007. The design was described previously.12 Briefly, participants were selected from commercial lists and recruited via phone and mail contact. Of eligible participants contacted, the cooperation rate was 49%, which is similar to other cohort studies.13 Due to the focus on assessing Black-White and regional disparities in stroke and cognitive decline, only self-identified non-Hispanic, White and Black individuals were included, and residents of the southeastern US stroke belt were oversampled. The stroke belt includes Alabama, Arkansas, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, and Tennessee, which are US states with a high burden of strokes.14,15 Additional exclusion criteria included medical conditions preventing long-term participation such as active cancer, residence in or awaiting placement in a nursing home, or inability to communicate in English.
Participants first completed a computer-assisted telephone interview for self-reported demographics, cardiovascular risk factors and medication data. They then underwent an in-home visit during which written informed consent, anthropomorphic measurements and medication inventory were obtained. During the in-home examination, standardized protocols were followed to obtain two blood pressure measurements that were averaged for analysis, an electrocardiogram, height and weight. Blood and urine samples were collected after a 10 to 12 hour fast. Blood was centrifuged locally and shipped with the urine samples on ice packs to the Laboratory for Clinical Biochemistry Research at the University of Vermont for reprocessing and analysis or storage.16
Participants were telephoned every 6 months for stroke and hospitalization ascertainment, cognitive testing, and vital status determination. Starting in April 2013, participants were invited to undergo a second in-person visit during which the baseline procedures were repeated. Among the 16,150 participants who participated in the second assessment, 15,938 completed the second telephone-administered assessment and 14,449 had in-person assessments, which were completed in December 2016.
For this analysis we excluded participants who were missing measures of kidney disease (creatinine or cystatin C) at the baseline or follow-up visit, had a baseline CKD (estimated glomerular filtration rate (eGFR) < 60 ml/min/1.73m2) or who were receiving kidney replacement therapy at baseline (Figure S1, Table S1).
The institutional review boards at all participating universities approved the study.
Outcome Definition:
Kidney function was measured at two time points: at the baseline and second in-person visits. Serum creatinine was measured using an isotope-dilution mass spectrometry–traceable method17 using the Vitros 950IRC instrument (Johnson & Johnson Clinical Diagnostics, Rochester, NY). Serum cystatin C was measured with high sensitivity particle-enhanced immunonephelometry (N Latex Cystatin C on the BNII, Dade Behring, Deerfield, IL). Serum creatinine and cystatin C measurements were calibrated for drift by re-assaying samples from 50 randomly selected participants who had available baseline and follow-up samples. eGFR change was defined as difference in eGFR divided by time in years for each participant. Incident CKD was defined as an eGFR<60 ml/min/1.73m2 at follow-up and at least a 40% decline in eGFR from baseline or initiated kidney replacement therapy through REGARDS linkage to the United States Renal Data System through 2018. The CKD-EPI collaborative equation based on serum creatinine and cystatin C refit without the race variable was used to calculate eGFR.18
Covariates:
The following characteristics were included as potential risk factors for eGFR change or incident CKD: age, self-reported race (Black/White), sex (male/female), region (residing in stroke belt or not), annual household income (< vs. ≥$35,000), education (< vs. ≥ high school graduate), diabetes (fasting glucose ≥126 mg/dL, non-fasting glucose ≥200 mg/dL, or use of diabetes medication), history of cardiovascular disease (CVD), smoking (past/never vs current), systolic and diastolic blood pressure, hyperlipidemia (cholesterol >240 mg/dL or low density lipoprotein >160 mg/dL or use of medications for high cholesterol) and body mass index (BMI, calculated as kg/m2).
Urine albumin and creatinine were measured using a random spot urine specimen by nephelometry (BN ProsSpec nephelometer, Dade Behring, Marburg, Germany) and Modular-P chemistry analyzer (Rocher/Hitachi, Indianapolis, IN), respectively. Spot urine albumin-creatinine ratio (UACR) was calculated in mg/g and log-transformed.
Statistical Analyses:
Baseline participant characteristics were compared by eGFR change and incident CKD status. We used binomial regression to calculate the probability of incident CKD and 95% confidence intervals (CI) for the overall cohort and by age (45-54, 55-64, 65-74, 75 and older at baseline), sex and race strata. Multivariable linear regression was used to calculate the association of each CKD risk factor and eGFR change in three sequential models: a demographic model (model 1) included age, sex and race; a sociodemographic model (model 2) additionally included education, income, smoking and residence in the stroke belt (yes/no); a comorbidity/medication adjusted model (model 3) that additionally included blood pressure, BMI and history of diabetes, hyperlipidemia, heart disease, UACR, angiotensin receptor blocker (ARB)/angiotensin converting enzyme inhibitor(ACEI) use, or non-steroidal anti-inflammatory drug (NSAID) use. We calculated an interaction term for each CKD risk factor and race, sex and region, using the fully adjusted model. We analyzed fully adjusted models stratified by Black/White, men/women and stroke belt/non-stroke belt.
We repeated this analysis using modified Poisson regression to calculate the risk of incident CKD in the aforementioned sequential models. For this analysis, follow-up time was assumed equal at 13.25 years (the maximum follow-up time) for all patients. Patients who did not develop CKD, including those who withdrew and died, were assumed CKD-free until 13.25 years. If participants were missing a covariate for that model, they were not included in the calculation for relative risk.
Because race disparities have been reported in CKD and race could be associated with attrition due to death or withdrawal from study, as a sensitivity analysis, we performed an inverse probability attrition weighting analysis that has been previously used in REGARDS.19 In brief, all participants were coded to indicate a follow-up visit, study withdrawal or death. Covariate adjusted logistic regression models were used to predict probabilities of attrition using all cohort participants. Stabilized attrition weights were calculated and applied to the Poisson regression primary analysis specified above to account for potential differential attrition. Confidence intervals for the attrition models were obtained through bootstrapping procedures that incorporated both weight estimation and outcome model estimation, as has been previously reported and is standard for inverse probability attrition weight analysis in REGARDS.19–21 Analyses were conducted in SAS v9.4.
Results
A total of 14,449 REGARDS participants attended the baseline and follow-up visits. After exclusions for missing creatinine or cystatin C (n=1,030) or requiring kidney replacement therapy or eGFR <60 ml/min/1.73m2 at baseline (n=1,422), there were 11,997 participants in the final analytic cohort (Figure S1).
Over a mean 9.4 ± 1.0 (5.7-13.3) years of follow-up, a total of 1,067 participants developed incident CKD, 62 of whom had kidney failure. Participants who developed incident CKD were older, more likely female, Black, residing in the stroke belt, and had lower income and education level at baseline. They were also more likely to have prevalent diabetes, hyperlipidemia and heart disease, and higher systolic and diastolic blood pressure (Table 1). We examined baseline characteristics across the analytic cohort, participants who died prior to the second REGARDS visit and participants missing kidney function (Table S1). Compared to participants in the analytic cohort, those who died were older, more likely to be male, Black, smokers, have lower income and educational attainment, had higher SBP and albuminuria and were more likely to have diabetes, hyperlipidemia and cardiovascular disease. Participants who were excluded from our analysis because they were missing kidney function measures, were more likely to be female, Black, residents of non-stroke belt, lower income and educational attainment, have diabetes, hyperlipidemia and heart disease. They were similar age, had similar BP, BMI and albuminuria.
Table 1.
Baseline characteristics of REGARDS participants with and without incident CKD at second visit
| Original cohort | Current analysis | No CKD | Incident CKD | |
|---|---|---|---|---|
| No. (%) | 30,239 | 11,997 | 10,930 (91) | 1,067 (9) |
| Age, mean (SD) | 65 (9) | 63 (8) | 62 (8) | 66 (8) |
| Female, No. (%) | 16,632 (55) | 6,469 (54) | 5,881 (54) | 588 (55) |
| Black, No. (%) | 12,514 (41) | 4,198 (35) | 3,757 (34) | 441 (41) |
| Residence in stroke belt, No. (%) | 16,754 (55) | 6,682 (56) | 6,041 (55) | 641 (60) |
| Household Income ≥ $35,000/year, No. (%) | 13,668 (45) | 6,764 (56) | 6,306 (58) | 458 (43) |
| Education ≥ high school, No. (%) | 26,366 (87) | 11,146 (93) | 10,217 (94) | 929 (87) |
| Current smoking, No. (%) | 4,396 (15) | 1,336 (11) | 1,201 (11) | 135 (13) |
| Systolic blood pressure, mean (SD), mmHg | 128 (17) | 125 (15) | 125 (15) | 132 (16) |
| Diastolic blood pressure, mean (SD), mmHg | 77 (10) | 76 (9) | 76 (9) | 78 (10) |
| Body mass index, mean (SD), kg/m2 | 29.3 (6.2) | 29.1 (5.8) | 28.9 (5.7) | 30.8 (6.4) |
| Urine albumin-to-creatinine ratio Log, mean (SD) | 2.4 (1.3) | 2.1 (0.9) | 2.0 (0.9) | 2.6 (1.4) |
| Diabetes, No. (%) | 6,398 (21) | 1,793 (15) | 1,423 (13) | 370 (35) |
| Hyperlipidemia, No. (%) | 17,228 (57) | 6,711 (56) | 6,022 (55) | 689 (65) |
| History of cardiovascular disease, No. (%) | 5,349 (18) | 1,458 (12) | 1,242 (11) | 216 (20) |
| ACEI/ARB use | 10,863 (36) | 3,513 (29) | 3,031 (28) | 482 (45) |
| NSAIDs use | 5,997 (20) | 2,350 (20) | 2,088 (19) | 262 (25) |
| Baseline eGFR (ml/min/1.73m2), mean (SD) | 85 (20) | 90 (15) | 90 (15) | 88 (15) |
The original cohort refers to REGARDS participants who participated in the baseline visit. The current analysis includes REGARDS participants who also attended a second visit approximately 9 years later. Incident CKD was defined as new eGFR < 60 ml/min/1.73m2 and ≥ 40% decline in eGFR, or kidney replacement therapy, in participants with eGFR ≥ 60 ml/min/1.73m2 at baseline.
Abbreviations: ACEi/ARB—angiotensin converting enzyme inhibitor/angiotensin II receptor blocker; NSAID—non-steroidal anti-inflammatory drug.
The probability of incident CKD was 9% (9%, 10%) for the overall cohort and showed minimal differences by sex and race groups. The probability of incident CKD (95%CI) was higher across age strata, ranging from 4% (3%, 5%) for the 45-54 age group to 18% (16%, 20%) for the ≥ 75 age group. Race- and sex-specific incidence of CKD demonstrated a similar pattern across age groups (Figure 1).
Figure 1.
Incidence of CKD across age, race and sex strata. Error bars represent 95% confidence intervals.
Risk factors for eGFR change and incident CKD were examined in sequential models (Table 2A & 2B). Age and Black race were associated with eGFR change in the demographics adjusted model, β −0.13 (p <0.0001) for both. In the fully adjusted model accounting for all risk factors, Black race was no longer associated with eGFR change, β 0.02 (P 0.54). Age, low income, residence in the stroke belt, systolic blood pressure, BMI, diabetes, hyperlipidemia, and albuminuria were independently associated with eGFR decline.
Table 2A.
Association of risk factors with eGFR change over time in the REGARDS study.
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| N=11,997 | N=10,737 | N=10,106 | |
| Beta coefficient (P) | Beta coefficient (P) | Beta coefficient (P) | |
| Age* | −0.13 (<0.001) | −0.11 (<0.001) | −0.08 (<0.001) |
| Male sex | 0.03 (0.33) | −0.009 (0.76) | −0.01 (0.67) |
| Black race | −0.13 (<0.001) | −0.10 (0.002) | 0.02 (0.54) |
| Education ≥ High School | 0.16 (0.007) | 0.11 (0.07) | |
| Household Income ≥ $35,000 | 0.16 (<0.001) | 0.08 (0.01) | |
| Stroke belt | −0.11 (0.001) | −0.10 (<0.001) | |
| Current smoking | −0.05 (0.26) | −0.05 (0.33) | |
| Systolic blood pressure* | −0.10 (<0.001) | ||
| Body mass index* | −0.04 (0.006) | ||
| Diabetes | −0.50 (<0.001) | ||
| Hyperlipidemia | 0.09 (0.006) | ||
| Cardiovascular Disease | −0.06 (0.19) | ||
| Log Urine ACR* | −0.21 (<0.001) |
per one standard deviation increase: for age 8 years, for SBP 15 mmHg, for BMI 5.8 kg/m2, for albumin-to-creatinine ratio 0.9 mg/g. Model 3 is additionally adjusted for ACEI/ARB and NSAID use.
Abbreviations: ACEi/ARB—angiotensin converting enzyme inhibitor/angiotensin II receptor blocker; NSAID—non-steroidal anti-inflammatory drug; ACR—albumin-to-creatinine ratio.
Table 2B.
Relative risk (95% Confidence Interval) of incident CKD accounting for risk factors in the REGARDS study
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| N=11,997 | N=10,737 | N=10,106 | |
| Age* | 1.51 (1.44, 1.59) | 1.46 (1.38, 1.55) | 1.46 (1.37, 1.55) |
| Male sex | 0.94 (0.84, 1.06) | 1.06 (0.94, 1.20) | 0.95 (0.84, 1.09) |
| Black race | 1.39 (1.24, 1.56) | 1.31 (1.15, 1.49) | 1.03 (0.90, 1.18) |
| Education ≥ High School | 0.70 (0.58, 0.84) | 0.80 (0.67, 0.96) | |
| Household Income ≥ $35,000 | 0.75 (0.65, 0.86) | 0.86 (0.75, 0.99) | |
| Stroke belt | 1.17 (1.04, 1.33) | 1.14 (1.01, 1.30) | |
| Current smoking | 1.23 (1.02, 1.48) | 1.30 (1.08, 1.57) | |
| Systolic blood pressure* | 1.16 (1.10, 1.22) | ||
| Body mass index* | 1.18 (1.11, 1.25) | ||
| Diabetes | 1.91 (1.65, 2.20) | ||
| Hyperlipidemia | 1.11 (0.97, 1.27) | ||
| Cardiovascular Disease | 1.31 (1.12, 1.52) | ||
| Log Urine ACR* | 1.30 (1.25, 1.36) |
per one standard deviation increase: for age 8 years, for SBP 15 mmHg, for BMI 5.8 kg/m2, for albumin-to-creatinine ratio 0.9 mg/g. Model 3 is additionally adjusted for ACEI/ARB and NSAID use.
Abbreviations: ACEi/ARB—angiotensin converting enzyme inhibitor/angiotensin II receptor blocker; NSAID—non-steroidal anti-inflammatory drug; ACR—albumin-to-creatinine ratio.
Similar results were noted for the models where incident CKD was the outcome. In the demographics adjusted model, age (RR 1.51 95%CI 1.44, 1.59) and Black race (RR 1.39 95%CI 1.24, 1.56) were independently associated with incident CKD. After income, education, region of residence and current smoking were added to the model, age, Black race, higher educational achievement, higher income, current smoking and living in the stroke belt were independently associated with incident CKD. After adding systolic blood pressure, BMI, albuminuria and a history of diabetes, hyperlipidemia, cardiovascular disease, albuminuria, ACE inhibitor or ARB use, and NSAID use, Black race was no longer associated with incident CKD, whereas greater age, lower income and education, current smoking, higher systolic blood pressure, BMI, albuminuria, diabetes, heart disease, and residence in the stroke belt remained independently associated with incident CKD.
Risk factors for incident CKD and eGFR change were qualitatively similar across subgroups of sex, race and region with some exceptions (Table 3A & 3B).
Table 3A.
Association of risk factors with eGFR change over time by Race Sex and Region
| Black | White | P interaction | |
|---|---|---|---|
| n | 4,198 | 7,799 | |
| Age* | −0.12 (<0.001) | −0.06 (<0.001) | 0.10 |
| Male sex | −0.02 (0.70) | 0.00 (0.97) | 0.77 |
| Black race | n/a | n/a | - |
| Education ≥ High School | 0.07 (0.41) | 0.13 (0.14) | 0.64 |
| Income ≥$35,000 | 0.10 (0.07) | 0.08 (0.07) | 0.68 |
| Stroke belt | −0.09 (0.08) | −0.12 (0.001) | 0.73 |
| Current smoking | −0.13 (0.09) | 0.01 (0.90) | 0.15 |
| Systolic blood pressure* | −0.11 (<0.001) | −0.10 (<0.001) | 0.82 |
| Body mass index* | −0.09 (<0.001) | −0.02 (0.25) | 0.04 |
| Diabetes | −0.61 (<0.001) | −0.39 (<0.001) | 0.01 |
| Hyperlipidemia | 0.15 (0.007) | 0.04 (0.28) | 0.10 |
| Cardiovascular Disease | 0.06 (0.54) | −0.11 (0.04) | 0.10 |
| Log Urine ACR* | −0.26 (<0.001) | −0.17 (<0.001) | 0.01 |
| Male | Female | ||
| n | 5,528 | 6,469 | |
| Age* | −0.11 (<0.001) | −0.05 (0.02) | 0.06 |
| Male sex | n/a | n/a | - |
| Black race | 0.02 (0.61) | 0.04 (0.39) | 0.83 |
| Education ≥ High School | 0.01 (0.93) | 0.18 (0.03) | 0.17 |
| Income ≥$35,000 | 0.14 (0.007) | 0.05 (0.28) | 0.21 |
| Stroke belt | −0.11 (0.009) | −0.10 (0.02) | 0.87 |
| Current smoking | −0.05 (0.50) | −0.05 (0.41) | 0.95 |
| Systolic blood pressure* | −0.09 (<0.001) | −0.11 (<0.001) | 0.56 |
| Body mass index* | −0.02 (0.37) | −0.06 (0.001) | 0.22 |
| Diabetes | −0.48 (<0.001) | −0.53 (<0.001) | 0.53 |
| Hyperlipidemia | 0.07 (0.13) | 0.10 (0.03) | 0.67 |
| Cardiovascular Disease | 0.00 (0.99) | −0.14 (0.07) | 0.13 |
| Log Urine ACR* | −0.19 (<0.001) | −0.23 (<0.001) | 0.25 |
| Stroke belt | Non-stroke belt | ||
| n | 6,682 | 5,315 | |
| Age* | −0.08 (<0.001) | −0.08 (<0.001) | 0.95 |
| Male sex | −0.02 (0.63) | 0.01 (0.88) | 0.67 |
| Black race | 0.05 (0.32) | 0.01 (0.84) | 0.59 |
| Education ≥ High School | 0.11 (0.17) | 0.13 (0.20) | 0.85 |
| Income ≥$35,000 | 0.09 (0.05) | 0.08 (0.12) | 0.90 |
| Stroke belt | n/a | n/a | - |
| Current smoking | 0.04 (0.53) | −0.17 (0.02) | 0.03 |
| Systolic blood pressure* | −0.09 (<0.001) | −0.12 (<0.001) | 0.48 |
| Body mass index* | −0.06 (0.004) | −0.03 (0.18) | 0.35 |
| Diabetes | −0.52 (<0.001) | −0.47 (<0.01) | 0.62 |
| Hyperlipidemia | 0.11 (0.01) | 0.06 (0.20) | 0.45 |
| Cardiovascular Disease | −0.02 (0.77) | −0.11 (0.12) | 0.34 |
| Log Urine ACR* | −0.25 (<0.001) | −0.16 (<0.0001) | 0.01 |
relative risk per one standard deviation increase: for age 8 years, for SBP 15 mmHg, for BMI 5.8 kg/m2, for albumin-to-creatinine ratio (ACR) 0.9 mg/g.
Table 3B:
Association of risk factors and incident CKD (95% CI) by Race, Sex and Region
| Black | White | P interaction | |
|---|---|---|---|
| n | 4,198 | 7,799 | |
| Age* | 1.40 (1.27, 1.54) | 1.52 (1.40, 1.65) | 0.66 |
| Male sex | 0.96 (0.79, 1.18) | 0.91 (0.77, 1.09) | 0.71 |
| Black race | n/a | n/a | - |
| Education ≥ High School | 0.73 (0.58, 0.93) | 0.86 (0.63, 1.16) | 0.47 |
| Income ≥$35,000 | 0.95 (0.77, 1.18) | 0.80 (0.67, 0.95) | 0.15 |
| Stroke belt | 1.11 (0.92, 1.35) | 1.17 (0.99, 1.38) | 0.80 |
| Current smoking | 1.35 (1.05, 1.74) | 1.22 (0.92, 1.62) | 0.49 |
| Systolic blood pressure* | 1.12 (1.04, 1.20) | 1.21 (1.12, 1.31) | 0.12 |
| Body mass index* | 1.21 (1.11, 1.32) | 1.18 (1.09, 1.28) | 0.50 |
| Diabetes | 1.79 (1.46, 2.19) | 2.32 (1.92, 2.80) | 0.06 |
| Hyperlipidemia | 1.06 (0.88, 1.28) | 1.19 (0.98, 1.45) | 0.39 |
| Cardiovascular Disease | 1.15 (0.88, 1.49) | 1.45 (1.20, 1.75) | 0.15 |
| Log Urine ACR* | 1.38 (1.30, 1.45) | 1.24 (1.16, 1.32) | 0.02 |
| Male | Female | ||
| n | 5,528 | 6,469 | |
| Age* | 1.49 (1.36, 1.63) | 1.45 (1.33, 1.58) | 0.11 |
| Male sex | n/a | n/a | - |
| Black race | 1.06 (0.87, 1.29) | 1.04 (0.87, 1.25) | 0.84 |
| Education ≥ High School | 0.76 (0.57, 1.01) | 0.85 (0.67, 1.07) | 0.63 |
| Income ≥$35,000 | 0.99 (0.81, 1.20) | 0.74 (0.61, 0.91) | 0.03 |
| Stroke belt | 1.10 (0.92, 1.32) | 1.21 (1.01, 1.45) | 0.52 |
| Current smoking | 1.24 (0.94, 1.64) | 1.35 (1.05, 1.73) | 0.86 |
| Systolic blood pressure* | 1.19 (1.10, 1.30) | 1.13 (1.05, 1.22) | 0.42 |
| Body mass index* | 1.21 (1.10, 1.34) | 1.19 (1.11, 1.29) | 0.52 |
| Diabetes | 2.06 (1.70, 2.51) | 1.99 (1.63, 2.42) | 0.78 |
| Hyperlipidemia | 1.07 (0.87, 1.31) | 1.18 (0.99, 1.41) | 0.45 |
| Cardiovascular Disease | 1.27 (1.04, 1.55) | 1.47 (1.17, 1.86) | 0.30 |
| Log Urine ACR* | 1.31 (1.24, 1.39) | 1.30 (1.21, 1.39) | 0.81 |
| Stroke belt | Non-stroke belt | ||
| n | 6,682 | 5,315 | |
| Age* | 1.49 (1.38, 1.62) | 1.43 (1.30, 1.59) | 0.07 |
| Male sex | 0.92 (0.77, 1.09) | 1.00 (0.82, 1.23) | 0.48 |
| Black race | 1.01 (0.85, 1.21) | 1.07 (0.87, 1.32) | 0.76 |
| Education ≥ High School | 0.82 (0.66, 1.02) | 0.78 (0.56, 1.09) | 0.76 |
| Income ≥$35,000 | 0.83 (0.70, 1.00) | 0.90 (0.73, 1.11) | 0.76 |
| Stroke belt | n/a | n/a | - |
| Current smoking | 1.15 (0.90, 1.48) | 1.52 (1.15, 2.02) | 0.22 |
| Systolic blood pressure* | 1.13 (1.05, 1.21) | 1.22 (1.13, 1.33) | 0.11 |
| Body mass index* | 1.24 (1.16, 1.33) | 1.14 (1.03, 1.26) | 0.08 |
| Diabetes | 1.92 (1.60, 2.30) | 2.18 (1.75, 2.71) | 0.40 |
| Hyperlipidemia | 1.18 (0.99, 1.41) | 1.07 (0.87, 1.33) | 0.50 |
| Cardiovascular Disease | 1.25 (1.03, 1.53) | 1.47 (1.16, 1.87) | 0.27 |
| Log Urine ACR* | 1.35 (1.28, 1.42) | 1.25 (1.16, 1.34) | 0.11 |
relative risk per one standard deviation increase: for age 8 years, for SBP 15 mmHg, for BMI 5.8 kg/m2, for albuminuria 0.9 mg/g.
Abbreviations: CI –confidence interval, ACR— albumin-to-creatinine ratio
In Black compared to White participants, albuminuria was a stronger risk factor for eGFR change (β −0.26 vs β −0.17, pinteraction 0.01) and incident CKD (RR 1.38 (95% CI 1.30, 1.45) vs 1.24 (95%CI 1.16, 1.32) pinteraction 0.02. BMI was a stronger risk factor for eGFR change in Black compared to White participants (β −0.09 vs β −0.02, pinteraction 0.04), but relative risk of incident CKD was not different. Diabetes was a stronger risk factor for eGFR in Black compared to White participants (β −0.61 vs β −0.39, pinteraction 0.01) whereas diabetes was a stronger risk factor for incident CKD in White compared to Black participants (RR 2.32 (95%CI 1.92, 2.80) vs RR 1.79 (95%CI 1.46, 2.19), pinteraction 0.06).
Albuminuria was a stronger risk factor for eGFR change and incident CKD in the stroke belt compared to non-stroke belt, whereas smoking was a stronger risk factor of eGFR decline in the non-stroke belt compared to stroke belt.
In a sensitivity analysis accounting for attrition due to death or withdrawal from the study, relative risk estimates of incident CKD were similar to the primary analyses (Table S2).
Discussion
We assessed the incidence of and risk factors for CKD in a large, contemporary cohort of Black and White Americans. Using two complementary definitions for CKD eGFR change and incident CKD (decline in eGFR of at least 40% to below 60 ml/min/1.73m2), older age, lower education and income, higher systolic blood pressure, albuminuria and BMI, diabetes, cardiovascular disease, current smoking and residence in the US stroke belt were independently associated with CKD. Black race was not associated with risk of eGFR change or incident CKD in the model that included other risk factors, suggesting that prevention or better treatment of conventional CKD risk factors is important for narrowing the racial disparity in CKD.
Race differences in CKD are of particular interest given the disproportionate numbers of Black compared to White individuals with kidney failure.22 Our finding that the higher incidence of CKD and eGFR decline in Black compared to White participants was attenuated when adjusting for other risk differs from other studies. In MESA, which recruited Chinese, Black, Hispanic and White individuals free from cardiovascular disease from five U.S. sites, incident CKD was higher in Blacks than Whites, and this was only modestly attenuated by adjustment for diabetes and hypertension.5 In CARDIA, a study of Black and White adults aged 18-30 from four cities, with approximately 30 years of follow-up, the trajectory of eGFR decline was steeper in Black than White participants after age 35, a difference that was partly attenuated when adjusting for systolic blood pressure, diabetes and albuminuria.23 Data from ARIC demonstrated Black adults had a 21% higher risk of CKD compared to White adults, even after accounting for demographics, socioeconomic and clinical factors.24 REGARDS findings confirm that CKD incidence is higher among Black than White adults, but here we show that the higher incidence of CKD and eGFR decline in Black participants in REGARDS is fully attenuated by adjusting for CKD risk factors. REGARDS findings may differ for a few reasons. First, REGARDS has much greater geographic representation than prior studies5,23–25 avoiding regional differences in risk factors or differential participation in the study by race. Second, it is the largest cohort study to date to evaluate this topic, so had better ability to account for potential factors that may impact racial differences. Third, we included a more stringent definition of incident CKD than in prior studies in order to more faithfully capture the development of clinically significant kidney disease.24,26 Finally, it is also important to note that results might differ since we estimated GFR using the new CKD-EPI equation with combined creatinine-cystatin C equation, in contrast to prior studies. that used either cystatin C or creatinine, but not both. Also, since we used the updated CKD-EPI equation without race, our incident CKD analysis excluded more Black participants and included more White participants compared to studies that used the 2012 CKD-EPI equation.18
This study highlights how risk factors for CKD compare in Black and White individuals. We found that albuminuria was a stronger risk factor for eGFR change and incident CKD among Black compared with White adults in our study. Others have demonstrated the higher prevalence of albuminuria among Black compared to White individuals.27 Diabetes was a stronger risk factor of eGFR decline in Black compared to White individuals but was a stronger risk factor for incident CKD in White compared to Black individuals. BMI was a stronger risk factor of eGFR decline in Black compared to White individuals but there was no apparent difference across race groups for incident CKD. The remainder of the risk factors we studied were qualitatively similar across Black and White participants.
Our results also add to the available literature by newly identifying residence in the US stroke belt as a risk factor for the development of incident CKD and eGFR decline. We also highlight that albuminuria is a more potent risk factor for CKD in the stroke belt compared to non-stroke belt, while the remainder of CKD risk factors were similar across region. Prior work in REGARDS similarly identified regional differences in incident kidney failure needing replacement therapy.28 It is noteworthy that the increased risk of CKD among residents of the US stroke belt was independent of established CKD risk factors that disproportionately impact the southeastern US, including smoking, diabetes, low socioeconomic status and cardiovascular disease. These findings suggest that other factors may contribute to the development of incident CKD in those residing in the US stroke belt. Compared to the rest of the US, residence in the southeast may experience differences in environmental exposures such as heat,29 air pollution30 or water quality,31 which have each been linked to kidney disease. Future research should work to determine these factors, as well as their impact on albuminuria, in order to reduce the excess burden of CKD and kidney failure needing replacement therapy in southeastern US states.
The reported incidence of CKD has varied considerably in population-based cohorts in the US from 6%-25% in studies with 7-10 years of follow-up.6,32–34 Some of the variability in incidence may be attributed to the different age groups included in the studies. Our study is in line with Cardiovascular Health Study (CHS) and Health, Aging and Body Composition (Health ABC) cohorts in that we found that the incidence of CKD was 18% in those 75 years of age and older. In addition, our results add to these prior reports by showing that, for the most part, the incidence of CKD did not substantially differ by race or sex in older age groups. It is important to note that the definition of incident CKD varied considerably from cohort to cohort, with some using the onset of an eGFR < 60 ml/min/1.73m2 and at least a 25% decrease from baseline,6,24 whereas others used at least a 1 ml/min/1.73m2 per year decline from baseline as the qualifier.5 Given the more stringent definition of CKD in the current study (eGFR < 60 ml/min/1.73m2 and at least a 40% decline from baseline), if we applied these older definitions the rate would be even higher in REGARDS.
In this study, incident CKD and eGFR change did not vary by sex in any of the multivariable models. This finding is in contrast to other studies that report women have a higher risk of incident CKD than men, but the literature is mixed.6,35 Previous studies suggested that women have slower progression of CKD due to potential protective effects of estrogen or that men have faster progression of CKD and more incident kidney failure needing replacement therapy due to fibrotic effects of testosterone.36,37 Use of hormone replacement therapy in menopausal women is associated with improved kidney function, whereas in post-menopausal woman it has been associated with decline in kidney function.38,39 In a Chinese CKD screening study, eGFR values were lower in men compared to women in ages 18-60 but not over at 60 years.40 It is possible that sex differences in incident CKD may have been obscured by the broader age inclusion in REGARDS compared to other studies. In our age stratified analyses, the risk of CKD in Black and White women was numerically higher compared to Black and White men, respectively at higher age strata.
Sex-stratified analyses of CKD risk factors were similar for most risk factors. Only higher income was associated with 26% reduced risk of incident CKD in women, whereas income was not a risk factor for incident CKD in men. This finding was not replicated in the eGFR change models. Age was a more potent risk factor of eGFR decline in men compared to women, but this was qualitatively similar across men and women in the incident CKD models.
There are limitations to consider in our analysis. Due to the design of the REGARDS study, self-identified Hispanic participants were excluded, and we do not know if our results are generalizable to other populations. We included only REGARDS participants who attended the second visit potentially introducing a survival bias. If non-participants were also sicker, they may have been more likely to develop CKD, and our estimates would underrepresent the true incidence of CKD. If Black participants were more likely to die before the second visit, then our estimates of race as a risk factor for CKD would also be biased. Importantly, when we accounted for differential attrition in a sensitivity analysis, similar to an approach used previously, we found that our results were similar to the main findings.19 We are unable to account for duration or strength of risk factors, which are likely to be important with regards to CKD risk. Finally, this study was not designed to evaluate whether differences in CKD risk factors are consequences of genetics, behavior, socioeconomic status, or structural racism.
We also note several key strengths to our study. We report one of the largest, contemporary cohort studies of Black and White participants, with a high follow-up rate, sampling across the United States and approximately ten-year follow up. We evaluated the incidence of and risk factors for CKD using both eGFR change and a stringent definition of incident CKD.
In summary, in this large cohort study of Black and White adults, traditional CKD risk factors accounted for the higher risk of incident CKD and eGFR decline in Black as compared to White adults, supporting the focus on addressing modifiable risk factors such as diabetes, hypertension and obesity in reducing disparities in CKD. Albuminuria was a more potent risk factor incident CKD and eGFR decline for Black participants and stroke belt residents, compared to White participants and those living outside the stroke belt. In addition, residence in the US stroke belt was an independent risk factor for incident CKD, an observation that will require further study to determine the key factors which underlie this finding.
Supplementary Material
Figure S1: Flow diagram of REGARDS study participants included in this analysis
Table S1: Baseline Characteristics of Participants included and excluded (due to death or missing kidney function)
Table S2: Relative risk (95% Confidence Interval) of incident CKD accounting for risk factors in the REGARDS study accounting for attrition (death or withdrawal from study)
Plain language summary.
In this longitudinal cohort study of Black and White adults at least 45 years of age followed for approximately 10 years, 9% developed chronic kidney disease (CKD), ranging from 4% in those 45-54 years old to 18% in those ≥75 years old. Systolic blood pressure, diabetes, heart disease, body-mass index, and residence in the Southeastern stroke belt were independent risk factors for CKD. Black individuals had higher incidence of CKD compared to White individuals, but this was attenuated after adjusting for traditional CKD risk factors. Albuminuria was a stronger risk factor for kidney disease in Black compared to White individuals.
Acknowledgements:
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/.
Support:
This research project is supported by cooperative agreement U01 NS041588 co-funded by the National Institute of Neurological Disorders and Stroke (NINDS), the National Institute on Aging (NIA), National Institutes of Health, Department of Health and Human Service. Representatives of the NINDS were involved in the review of the manuscript but were not directly involved in the collection, management, analysis or interpretation of the data. KLC was also supported by the National Palliative Care Research Center and the National Institute of General Medical Sciences (NIGMS) P20 GM135007. OG was also supported by grant K24DK116180 from the National Institute of Diabetes, Digestive and Kidney Diseases.
Financial Disclosure:
DCC reports research funding from Somatus, Inc. , is co-chair for Bayer Healthcare Pharmaceuticals Inc. Patient and Physician Advisory Board Steering Committee for Disparities in Chronic Kidney Disease Project, is a member of the Advisory Group for the Health Equity Collaborative, and is a member of Partner Research for Equitable System Transformation after COVID-19 (PRESTAC), Optum Labs. MGS is a consultant for Cricket Health and Intercept Pharmaceuticals. OG reports receiving grant support and honoraria from Amgen and Akebia, grant support form GSK, honoraria from AstraZeneca, Reata, and Ardelyx, and serving on a Data Monitoring Committee for QED Therapeutis. The other authors declare that they have no relevant financial interests.
Footnotes
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.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the NINDS or the NIA.
Contributor Information
Katharine L. Cheung, Larner College of Medicine at The University of Vermont, Division of Nephrology.
Deidra C. Crews, Johns Hopkins University, Division of Nephrology.
Mary Cushman, Larner College of Medicine at The University of Vermont, Department of Medicine.
Ya Yuan, University of Alabama at Birmingham, School of Public Health.
Katherine Wilkinson, Larner College of Medicine at The University of Vermont.
D. Leann Long, University of Alabama at Birmingham, School of Public Health.
Suzanne E. Judd, University of Alabama at Birmingham, School of Public Health.
Michael G. Shlipak, University of California San Francisco, Division of Nephrology.
Joachim H. Ix, University of California San Diego, Division of Nephrology.
Alexander L. Bullen, University of California San Diego, Division of Nephrology.
David G. Warnock, University of Alabama at Birmingham, Division of Nephrology.
Orlando M. Gutiérrez, University of Alabama at Birmingham, Division of Nephrology.
References:
- 1.Heron M Deaths: Leading causes for 2017. Vol. 68. 2019. National vital statistics reports. https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_06-508.pdf [PubMed] [Google Scholar]
- 2.Plantinga LC, Johansen K, Crews DC, et al. Association of CKD with disability in the United States. American journal of kidney diseases: the official journal of the National Kidney Foundation. Feb 2011;57(2):212–27. doi: 10.1053/j.ajkd.2010.08.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. The New England journal of medicine. Sep 23 2004;351(13):1296–305. doi: 10.1056/NEJMoa041031 [DOI] [PubMed] [Google Scholar]
- 4.Johansen KL, Chertow GM, Gilbertson DT, et al. US Renal Data System 2021 Annual Data Report: epidemiology of kidney disease in the United States. Am J Kidney Dis. 2022;79(4)(suppl 1):A8–A12. doi: 10.1053/j.ajkd.2022.02.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Peralta CA, Katz R, DeBoer I, et al. Racial and ethnic differences in kidney function decline among persons without chronic kidney disease. Journal of the American Society of Nephrology : JASN. Jul 2011;22(7):1327–34. doi: 10.1681/ASN.2010090960 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bash LD, Coresh J, Kottgen A, et al. Defining incident chronic kidney disease in the research setting: The ARIC Study. American journal of epidemiology. Aug 15 2009;170(4):414–24. doi: 10.1093/aje/kwp151 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Jones CP. Levels of racism: a theoretic framework and a gardener’s tale. American journal of public health. Aug 2000;90(8):1212–5. doi: 10.2105/ajph.90.8.1212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hales CM, Carroll MD, Fryar CD, Ogden CL. Prevalence of obesity among adults and youth: United States, 2015-2016. 2017. NCHS data brief, no 288. [PubMed] [Google Scholar]
- 9.Lu JL, Molnar MZ, Naseer A, Mikkelsen MK, Kalantar-Zadeh K, Kovesdy CP. Association of age and BMI with kidney function and mortality: a cohort study. Lancet Diabetes Endocrinol. Sep 2015;3(9):704–14. doi: 10.1016/S2213-8587(15)00128-X [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Halbesma N, Brantsma AH, Bakker SJ, et al. Gender differences in predictors of the decline of renal function in the general population. Kidney international. Aug 2008;74(4):505–12. doi: 10.1038/ki.2008.200 [DOI] [PubMed] [Google Scholar]
- 11.Nitsch D, Grams M, Sang Y, et al. Associations of estimated glomerular filtration rate and albuminuria with mortality and renal failure by sex: a meta-analysis. Bmj. Jan 29 2013;346:f324. doi: 10.1136/bmj.f324 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.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]
- 13.Tell GS, Fried LP, Hermanson B, Manolio TA, Newman AB, Borhani NO. Recruitment of adults 65 years and older as participants in the cardiovascular health study. Annals of epidemiology. 1993/07/01/ 1993;3(4):358–366. doi: 10.1016/1047-2797(93)90062-9 [DOI] [PubMed] [Google Scholar]
- 14.Lanska DJ. Geographic distribution of stroke mortality in the United States: 1939-1941 to 1979-1981. Neurology. Sep 1993;43(9):1839–51. doi: 10.1212/wnl.43.9.1839 [DOI] [PubMed] [Google Scholar]
- 15.Howard G, Evans GW, Pearce K, et al. Is the stroke belt disappearing? An analysis of racial, temporal, and age effects. Stroke; a journal of cerebral circulation. Jul 1995;26(7):1153–8. doi: 10.1161/01.str.26.7.1153 [DOI] [PubMed] [Google Scholar]
- 16.Gillett SR, Boyle RH, Zakai NA, McClure LA, Jenny NS, Cushman M. Validating laboratory results in a national observational cohort study without field centers: the Reasons for Geographic and Racial Differences in Stroke cohort. Clinical biochemistry. Nov 2014;47(16-17):243–6. doi: 10.1016/j.clinbiochem.2014.08.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.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. American journal of kidney diseases : the official journal of the National Kidney Foundation. Aug 2008;52(2):227–34. doi: 10.1053/j.ajkd.2008.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.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;385(19):1737–1749. doi: 10.1056/nejmoa2102953 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Long DL, Howard G, Long DM, et al. An Investigation of Selection Bias in Estimating Racial Disparity in Stroke Risk Factors. American journal of epidemiology. 2019;188(3):587–597. doi: 10.1093/aje/kwy253 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Weuve J, Tchetgen Tchetgen EJ, Glymour MM, et al. Accounting for bias due to selective attrition: the example of smoking and cognitive decline. Epidemiology. Jan 2012;23(1):119–28. doi: 10.1097/EDE.0b013e318230e861 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Robins JM, Rotnitzky A. Semiparametric Efficiency in Multivariate Regression-Models with Missing Data. Journal of the American Statistical Association. Mar 1995;90(429):122–129. doi: 10.2307/2291135 [DOI] [Google Scholar]
- 22.Albertus P, Morgenstern H, Robinson B, Saran R. Risk of ESRD in the United States. American journal of kidney diseases : the official journal of the National Kidney Foundation. Dec 2016;68(6):862–872. doi: 10.1053/j.ajkd.2016.05.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Peralta CA, Vittinghoff E, Bansal N, et al. Trajectories of kidney function decline in young black and white adults with preserved GFR: results from the Coronary Artery Risk Development in Young Adults (CARDIA) study. American journal of kidney diseases : the official journal of the National Kidney Foundation. Aug 2013;62(2):261–6. doi: 10.1053/j.ajkd.2013.01.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Evans K, Coresh J, Bash LD, et al. Race differences in access to health care and disparities in incident chronic kidney disease in the US. Nephrology, dialysis, transplantation : official publication of the European Dialysis and Transplant Association - European Renal Association. Mar 2011;26(3):899–908. doi: 10.1093/ndt/gfq473 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Shlipak MG, Katz R, Kestenbaum B, et al. Rapid decline of kidney function increases cardiovascular risk in the elderly. Journal of the American Society of Nephrology : JASN. Dec 2009;20(12):2625–30. doi: 10.1681/ASN.2009050546 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kshirsagar AV, Bang H, Bomback AS, et al. A Simple Algorithm to Predict Incident Kidney Disease. Archives of internal medicine. 2008;168(22):2466. doi: 10.1001/archinte.168.22.2466 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Jolly SE, Burrows NR, Chen SC, et al. Racial and ethnic differences in albuminuria in individuals with estimated GFR greater than 60 mL/min/1.73 m(2): results from the Kidney Early Evaluation Program (KEEP). American journal of kidney diseases: the official journal of the National Kidney Foundation. Mar 2010;55(3 Suppl 2):S15–22. doi: 10.1053/j.ajkd.2009.09.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Tanner RM, Gutierrez OM, Judd S, et al. Geographic variation in CKD prevalence and ESRD incidence in the United States: results from the reasons for geographic and racial differences in stroke (REGARDS) study. American journal of kidney diseases : the official journal of the National Kidney Foundation. Mar 2013;61(3):395–403. doi: 10.1053/j.ajkd.2012.10.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lee WS, Kim WS, Lim YH, Hong YC. High Temperatures and Kidney Disease Morbidity: A Systematic Review and Meta-analysis. J Prev Med Public Health. Jan 2019;52(1):1–13. doi: 10.3961/jpmph.18.149 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kuźma Ł, Małyszko J, Bachórzewska-Gajewska H, Kralisz P, Dobrzycki S. Exposure to air pollution and renal function. Scientific reports. 2021;11(1)doi: 10.1038/s41598-021-91000-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Chang K, Wu IW, Huang B-R, et al. Associations between Water Quality Measures and Chronic Kidney Disease Prevalence in Taiwan. International journal of environmental research and public health. 2018;15(12):2726. doi: 10.3390/ijerph15122726 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Anuwatmatee S, Allison MA, Shlipak MG, et al. Relationship of fibroblast growth factor 21 with kidney function and albuminuria: multi-ethnic study of atherosclerosis. Nephrology, dialysis, transplantation : official publication of the European Dialysis and Transplant Association - European Renal Association. Jun 1 2019;34(6):1009–1016. doi: 10.1093/ndt/gfy120 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Drew DA, Katz R, Kritchevsky S, et al. Association between Soluble Klotho and Change in Kidney Function: The Health Aging and Body Composition Study. Journal of the American Society of Nephrology : JASN. Jun 2017;28(6):1859–1866. doi: 10.1681/ASN.2016080828 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Shlipak MG, Katz R, Kestenbaum B, et al. Rate of kidney function decline in older adults: a comparison using creatinine and cystatin C. American journal of nephrology. 2009;30(3):171–8. doi: 10.1159/000212381 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Chesnaye NC, Dekker FW, Evans M, et al. Renal function decline in older men and women with advanced chronic kidney disease—results from the EQUAL study. Nephrol Dial Transpl. 2021;36(9):1656–1663. doi: 10.1093/ndt/gfaa095 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Neugarten J, Acharya A, Silbiger SR. Effect of gender on the progression of nondiabetic renal disease: a meta-analysis. Journal of the American Society of Nephrology : JASN. Feb 2000;11(2):319–329. doi: 10.1681/ASN.V112319 [DOI] [PubMed] [Google Scholar]
- 37.Carrero JJ, Hecking M, Chesnaye NC, Jager KJ. Sex and gender disparities in the epidemiology and outcomes of chronic kidney disease. Nature reviews Nephrology. Mar 2018;14(3):151–164. doi: 10.1038/nrneph.2017.181 [DOI] [PubMed] [Google Scholar]
- 38.Kaygusuz I, Gumus II, Yuvaci HU, Kasapoglu B, Carlioglu A. Does hormone replacement therapy have beneficial effects on renal functions in menopausal women? Archives of gynecology and obstetrics. Jun 2012;285(6):1643–6. doi: 10.1007/s00404-012-2215-8 [DOI] [PubMed] [Google Scholar]
- 39.Silbiger S The effects of hormone replacement therapy on renal function. Nature clinical practice Nephrology. Jan 2009;5(1):6–7. doi: 10.1038/ncpneph0993 [DOI] [PubMed] [Google Scholar]
- 40.Xu R, Zhang L-X, Zhang P-H, Wang F, Zuo L, Wang H-Y. Gender differences in age-related decline in glomerular filtration rates in healthy people and chronic kidney disease patients. BMC nephrology. 2010;11:20–20. doi: 10.1186/1471-2369-11-20 [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 of REGARDS study participants included in this analysis
Table S1: Baseline Characteristics of Participants included and excluded (due to death or missing kidney function)
Table S2: Relative risk (95% Confidence Interval) of incident CKD accounting for risk factors in the REGARDS study accounting for attrition (death or withdrawal from study)