Abstract
Rationale & Objective
Acute kidney injury (AKI) frequently complicates acute heart failure (HF), yet predictors of HF-associated AKI remain poorly defined. We examined whether urinary biomarkers of tubular health, measured years before HF hospitalization, predict subsequent AKI risk.
Study Design
Case-control study.
Setting & Participants
Adults aged ≥45 years in the reasons for geographic and racial differences in stroke study hospitalized for acute HF. We compared 400 AKI cases with 419 frequency-matched non-AKI controls.
Exposures
Biomarkers of tubular function (alpha-1 microglobulin, uromodulin, epidermal growth factor levels) and injury (kidney injury molecule-1 levels).
Outcomes
AKI, defined as a serum creatinine level rise ≥ 0.3 mg/dL during HF admission.
Analytical Approach
Logistic regression adjusted for demographics, comorbidities, urine creatinine, time to HF, estimated glomerular filtration rate, and albuminuria.
Results
Median time was 5.6 years from baseline to HF. AKI cases exhibited higher levels of alpha-1 microglobulin and lower levels of uromodulin and EGF than non-AKI controls. Higher EGF level was strongly associated with lower AKI risk (OR for fourth vs first quartile: 0.27; 95% CI, 0.16-0.46), persisting after adjustment for estimated glomerular filtration rate and albuminuria (OR, 0.49; 95% CI, 0.27-0.91). Continuous EGF level analysis showed a similar trend but was not significant in fully adjusted models (odds ratio per 2-fold increase: 0.89, 95% CI, 0.77-1.04). No other biomarkers independently predicted AKI.
Limitations
Extended time from baseline to HF hospitalization.
Conclusions
Higher urinary EGF levels, measured years before HF, are associated with a lower risk of AKI during HF hospitalization. Tubular health may be associated with resilience against HF-related kidney injury.
Index Words: Acute kidney injury (AKI), biomarkers, alpha-1 microglobulin (A1M), uromodulin (UMOD), epidermal growth factor (EGF), kidney injury molecule-1 (KIM-1), acute heart failure, epidemiology
Plain-Language Summary
Although a frequent complication of acute heart failure, it is unclear who is vulnerable to acute kidney injury. We studied whether urine markers that reflect tubule health could identify future vulnerability. Using data from the REasons for Geographic And Racial Differences in Stroke (REGARDS) study, we analyzed urine samples collected up to 10 years before heart failure hospitalization, measuring epidermal growth factor, uromodulin, alpha-1 microglobulin, and kidney injury molecule-1 levels. Higher levels of epidermal growth factor, a marker of healthier tubule function and repair capacity, were associated with a lower risk of acute kidney injury during heart failure hospitalization. These findings suggest that markers of underlying tubule health may be used to identify individuals at higher long-term risk of acute kidney injury.
Introduction
Acute kidney injury (AKI) affects up to 10%-15% of hospitalized patients and is associated with increased morbidity and mortality.1,2 Acute decompensated heart failure (HF) commonly occurs concurrently with AKI. Intact kidney function is essential for recovery from acute HF. However, acute decompensated HF can cause or worsen AKI through kidney hypoperfusion and sodium and fluid retention. These pathways create a vicious cycle of worsening cardiac and renal function.3,4 It is unclear why certain individuals with comparable clinical trajectories of acute HF develop AKI. There are few reliable tools to identify individuals vulnerable to cardiorenal syndrome. Early identification of these individuals could lead to strategies to mitigate or prevent AKI in the setting of HF.
Current assessments of kidney function mainly focus on glomerular metrics like serum creatinine (SCr) levels and estimated glomerular filtration rate (eGFR). However, these measures often overlook kidney tubular health, which is vital for overall kidney function. Emerging biomarkers such as the levels of uromodulin (UMOD), kidney injury molecule-1 (KIM-1), alpha-1 microglobulin (A1M), and epidermal growth factor (EGF) show potential in the detection of kidney tubular injury.5, 6, 7, 8 Recent studies indicate these biomarkers can predict AKI risk in patients with acute HF, especially under pre-renal stressors.5
In this study, we sought to evaluate the associations between markers of kidney tubule dysfunction and injury and future risk of AKI after acute HF.
Methods
Study Design and Data Collection – The Reasons for Geographic and Racial Differences in Stroke Study
This study was approved by the institutional review board of the University of Alabama at Birmingham. We performed a case-control study using data from the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study. The REGARDS study is a longitudinal cohort encompassing 30,239 community-dwelling adults aged ≥45 years from across the continental United States.9 The study enrolled participants between January 2003 and October 2007. Baseline assessments included an interview that collected demographics, health behaviors, diagnosed medical conditions, and medication usage. Research personnel collected physiologic measurements and obtained blood and urine samples from each participant. The REGARDS study identified health and hospitalization events through telephone follow-up conducted with participants at 6-month intervals. The REGARDS study has followed participants since baseline.
Selection of Participants
The identification and adjudication of acute HF hospitalizations were performed by a separate REGARDS study effort and team distinct from our renal research team. Based on information from the 6-month follow-ups, the REGARDS study personnel retrieved medical records for hospitalizations related to acute heart disease.10 Identification of HF as a cause of hospitalization was based on signs and symptoms (eg, orthopnea, exertional dyspnea, paroxysmal nocturnal dyspnea, nocturnal cough, rales on pulmonary examination, cardiomegaly, jugular vein distension, central venous pressure > 16 mm Hg, peripheral edema, pleural effusion, hepatomegaly, and heart rate > 120 bpm), laboratory results (eg, B-type natriuretic peptide levels), imaging (eg, echocardiogram or contrast ventriculography), and ≥ 4.5 kg of weight loss with diuresis. Trained clinician-investigators reviewed the medical records in duplicate, with disagreements resolved by discussion.
Biomarker Analysis
Study personnel obtained blood and urine samples during the initial in-home visit. The samples were centrifuged within 2 hours of collection, placed on ice, and shipped overnight to the REGARDS study central laboratory at the University of Vermont. Upon arrival, laboratory researchers centrifuged the samples at 30,000 g and 4 °C.
The biomarker level measurements for the current study were performed by Brigham and Women’s Hospital BioCon Central Laboratory. KIM-1 and EGF levels were measured using a microbead-based enzyme-linked immunosorbent assay from the Bio-Rad Bio-Plex Luminex 200 reader (Hercules). The lower limit of detection (LLOD) and interassay coefficients of variation (CV) for the KIM-1 level were 1.98 pg/mL and 7.8%, respectively. For the EGF level, the LLOD value was 0.313 pg/mL, and the CV value was 9.2%. We measured A1M levels using the SIEMENS BNII Nephelometer (Tarrytown) with an LLOD value of 5.63 mg/L and an intra-assay CV value of 1.2-4.0%. We measured UMOD levels using the R-Plex Protocol on the Meso Scale Discovery platform (Rockville) with an LLOD value of 244.14 pg/mL and an interassay CV value of 6.4%. We measured all KIM-1, EGF, and UMOD biomarker levels twice and averaged the results to improve precision. We measured A1M levels only once because of its low CV value.
Outcomes—AKI
The primary outcome was AKI during acute HF hospitalization. Trained research personnel reviewed medical records for each HF hospitalization, abstracting the date and time of hospitalization, history of end-stage kidney disease before hospitalization, performance of coronary artery bypass graft surgery, all values of SCr levels, hemoglobin concentration, blood urea nitrogen levels, and albumin values when available throughout the admission. We conducted quality checks of all records, including verification of timing and values of the SCr level. SCr level measurements occurred according to local practice. Based on the Acute Kidney Injury Network (AKIN) criteria, we defined AKI as either (1) a change in SCr levels of ≥ 0.3 mg/dL between the lowest and highest SCr levels or (2) treatment with acute hemodialysis.11,12 We used the Acute Kidney Injury Network rather than other AKI systems because the analysis focused on change from the first inpatient creatinine levels, not the prehospitalization baseline. To ensure a more robust assessment of AKI, we did not limit ascertainment to SCr levels obtained within a 48-hour window.13 Because they were not consistently documented, we did not define AKI based upon urine output.
Covariates
Questionnaires were used for participants to report age, sex, and race. Research personnel calculated body mass index (BMI) using height and weight obtained at the baseline in-home visit. Diabetes history consisted of self-reported use of insulin or oral hypoglycemic agents, fasting serum glucose concentration ≥ 126 mg/dL, or a nonfasting serum glucose concentration ≥ 200 mg/dL. Hypertension history consisted of self-reported use of blood pressure-lowering medications or a systolic blood pressure ≥ 140 mmHg or a diastolic blood pressure ≥ 90 mmHg at the baseline visit. History of coronary heart disease included evidence of myocardial infarction on the baseline electrocardiograph, self-report of a prior history of a cardiac procedure, or self-reported history of myocardial infarction.
The parent REGARDS study determined urine albumin excretion levels using nephelometry using the BNII ProSpec nephelometer (now Siemens AG); urine creatinine excretion levels using the rate Jaffé method using the Modular-P chemistry analyzer (Roche/Hitachi); and baseline eGFR from SCr level measured using isotope dilution mass spectrometry-traceable methods using the 2021 Chronic Kidney Disease Epidemiology Collaboration equation without race.14
Data Analysis
Among all the REGARDS study participants, we identified the first acute HF hospitalization through 2018. We then identified all acute HF hospitalizations with potential AKI, excluding those without biomarkers available and where the patient reported a history of kidney transplantation or chronic dialysis. We created a frequency-matched sample with approximately equal numbers of randomly selected AKI cases and non-AKI controls. We summarized baseline participant demographic characteristics, health behaviors, and kidney function using descriptive statistics.15 We evaluated interrelationships between biomarker levels using Spearman correlations. We determined the associations between each tubule dysfunction marker level and AKI using logistic regression. We modeled the marker levels as a continuous linear exposure (per log2 increase). We also categorized each biomarker level into quartiles. We repeated the models adjusted for age, sex, race, urine creatinine level (to account for differences in urine tonicity), time from the REGARDS study baseline visit to admission for heart failure, diabetes, hypertension, and BMI. We fitted another model with additional adjustments for eGFR and urine albumin level measured at the baseline. We set statistical significance at 0.05 (2-tailed). We performed all analyses using SAS version 9.4 (SAS Institute Inc).
Results
From 1,846 participants with hospitalizations for probable or possible acute HF, we identified a cohort of 600 AKI and 600 non-AKI (Fig 1). After excluding participants without available urine specimens, there were 422 with AKI and 422 without AKI. After excluding those with a history of chronic dialysis (n=23) or kidney transplant (n=2) before AHF, we retained 400 with AKI and 419 without AKI in the analysis.
Figure 1.
Identification of AKI cases and Non-AKI controls. AKI, acute kidney injury.
Baseline demographics and smoking status were similar between participants with and without AKI. (Table 1) There were no major differences in demographic characteristics between those with versus without AKI. Those experiencing AKI were more likely to have a history of diabetes or coronary heart disease and had lower eGFR and higher urine albumin-to-creatinine ratios than those without AKI.
Table 1.
Baseline characteristics of Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study Participants Hospitalized for Acute Heart Failure, Stratified by Acute Kidney Injury (AKI)
| Characteristic | AKI N=400 |
Non-AKI N=419 |
All N=819 |
|---|---|---|---|
| Age, y, mean (±SD) | 70 (9) | 70 (9) | 70 (9) |
| Female, n (%) | 184 (46) | 214 (51) | 398 (49) |
| Black race, n (%) | 164 (41) | 178 (43) | 342 (42) |
| BMI, kg/m2, mean (±SD) | 31 (7) | 31 (7) | 31 (7) |
| Current smoking, n (%) | 68 (17) | 56 (13) | 124 (15) |
| Diabetes, n (%) | 191 (48) | 158 (38) | 349 (43) |
| Coronary heart disease, n (%) | 192 (49) | 155 (38) | 347 (43) |
| Hypertension, n (%) | 319 (80) | 311 (74) | 630 (77) |
| SBP, mmHg, mean (±SD) | 132 (18) | 132 (18) | 132 (18) |
| DBP, mmHg, mean (±SD) | 75 (11) | 76 (11) | 75 (11) |
| Regular use of NSAIDs, n (%) | 62 (16) | 68 (16) | 130 (16) |
| eGFR, mL/min/1.73m2, mean (±SD) | 67 (22) | 76 (22) | 72 (23) |
| Urinary ACR, mg/g, median (IQR) | 17 (8-96) | 12 (6-31) | 14 (6- 47) |
Missing data: BMI (AKI=5 and Non-AKI=8), Smoking (AKI=1 and Non-AKI=4), Diabetes (AKI=3 and Non-AKI=6), Coronary heart disease (AKI=8 and Non-AKI=10), Hypertension (AKI=2 and Non-AKI=1), SBP (AKI=0, Non-AKI=4), DBP (AKI=0 and Non-AKI=4), NSAID (AKI=1 and Non-AKI=3), Urinary ACR (AKI=3 and Non-AKI=1).
Abbreviations: ACR, urine albumin-to-creatinine ratio; AKI, acute kidney injury; BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate (creatinine-cystatin); IQR, interquartile range; NSAIDs, nonsteroidal anti-inflammatory drugs; SBP, systolic blood pressure; SD, standard deviation.
The kidney tubule marker levels showed weak to moderate correlation (Spearman correlation coefficient −0.08 to 0.39) with each other and other key biomarker levels (Table 2). The highest correlation (0.39) was between A1M level and eGFR. A higher A1M level was inversely correlated with the eGFR. EGF and UMOD levels were directly correlated with eGFR, but the correlations were weak.
Table 2.
Spearman Correlation Coefficients Between Tubule Injury and Dysfunction Biomarker Levels
| Biomarker | A1M/UCr | UMOD/UCr | EGF/UCr | KIM-1/UCr | eGFR |
|---|---|---|---|---|---|
| A1M/UCr | 1.0 | −0.11 | −0.01 | 0.31 | −0.19 |
| UMOD/UCr | 1.0 | 0.21 | −0.08 | 0.26 | |
| EGF/UCr | 1.0 | 0.25 | 0.39 | ||
| KIM-1/UCr | 1.0 | −0.02 | |||
| eGFR | 1.0 |
Abbreviations: A1M, alpha-1 microglobulin; EGF, epidermal growth factor; eGFR, estimated glomerular filtration rate; KIM-1, kidney injury molecule-1; UCr, urine creatinine level; UMOD, uromodulin.
Urine A1M levels were higher in AKI than in non-AKI participants (Table 3). The prevalence of AKI was higher with increasing A1M quartile (Fig 2). Conversely, urine UMOD and EGF levels were lower in AKI than in non-AKI participants; AKI prevalence was lower with increasing UMOD and EGF quartiles. KIM-1 level was similar between AKI and non-AKI participants. The relationship between KIM-1 quartiles and AKI prevalence was inconsistent.
Table 3.
Tubule Dysfunction and Injury Biomarker Levels, Stratified by Heart Failure-Associated AKI
| Biomarker Levels | AKI Median (IQR) |
No-AKI Median (IQR) |
Total Median (IQR) |
|---|---|---|---|
| A1M, mg/L | 10.6 (6.3-16.9) | 7.8 (5.6 – 12.8) | 11.0 (5.6-22.9) |
| UMOD, μg/mL | 7,351,542 (4,058,359-12,358,488) | 9,958,216 (5,181,700-15,162,523) | 8,553,721 (4,458,606-14,227,488) |
| EGF, ng/mL | 6,156 (2,822- 11,241) | 9,198 (4,824- 66,708) | 7,247 (3,496-13,444) |
| KIM-1, ng/mL | 816 (391-1760) | 851 (443-1563) | 832 (424-1641) |
Units standardized.
Abbreviations: AKI, acute kidney injury; A1M, alpha-1 microglobulin; EGF, epidermal growth factor; KIM-1, kidney injury molecule-1; IQR, interquartile range; UMOD, uromodulin.
Figure 2.
AKI development by biomarker quartile. AKI, acute kidney injury; A1M, alpha-1 microglobulin; EGF, epidermal growth factor; KIM-1, kidney injury molecule-1; UMOD, uromodulin.
Median time from the REGARDS study enrollment (when blood and urine were sampled for biomarker level measurements) to HF hospitalization was 5.6 (interquartile range [IQR], 2.6-9.4) years. Time to HF hospitalization was similar between participants with 5.7 (IQR, 2.9-9.1) years versus without AKI, 5.4 (IQR, 2.5-10.4) years, on the basis of frequency matching on this variable. The median minimum SCr level during hospitalization was similar between participants with AKI (1.2; IQR, 0.9-1.7 mg/dL) and without AKI (1.0; IQR, 0.8-1.3 mg/dL). Among those developing AKI, the median maximum SCr level was 1.9 (IQR, 1.4-2.7) mg/dL. Among participants with AKI, hemodialysis was performed in 10. Among the 40 AKI participants, AKI severities were stage I, n=290 (72.5%); stage II, n=61 (15.3%); and stage III, n=49 (12.3%). Approximately half (55.8%) of AKI occurred within the first 3 SCr level measurements during hospitalization.
Higher urine EGF level was associated with a lower risk of AKI in the unadjusted model (odds ratio [OR], 0.29; 95% CI, 0.19-0.44; Table 4). The results minimally changed after adjustment for age, sex, race, time from baseline to HF hospitalization, urine creatinine, diabetes, hypertension, and BMI (OR comparing the fourth to the first quartile: 0.27; 95% CI, 0.16-0.46) and after further adjustment for eGFR and urine albumin (OR, 0.49; 95% CI, 0.27-0.91). The results were similar when EGF level was analyzed as a continuous variable, though the results were no longer statistically significant in the fully adjusted model (OR per 2-fold higher EGF level, 0.89; 95% CI, 0.77-1.04). No other urine biomarker levels were consistently associated with incident AKI in multivariable-adjusted models, though a higher UMOD level was directionally associated with lower risk of AKI in fully adjusted models.
Table 4.
Associations Between Biomarker Levels of Tubule Dysfunction and Injury And Heart Failure-Associated Acute Kidney Injury (AKI)
| Biomarker Level | AKI (N / Total) | Unadjusted OR (95% CI) |
Model 1 OR (95% CI) |
Model 2 OR (95% CI) |
|---|---|---|---|---|
| A1M, mg/L | ||||
| Log2(A1M) | 400/819 | 1.20 (1.07-1.36) | 1.25 (1.09-1.44) | 1.01 (0.85-1.19) |
| Quartiles | ||||
| Q1 | 108/219 | 1.00 | 1.00 | 1.00 |
| Q2 | 77/189 | 0.71 (0.48-1.05) | 0.85 (0.56-1.29) | 0.68 (0.44-1.04) |
| Q3 | 95/206 | 0.88 (0.60-1.29) | 0.98 (0.65-1.49) | 0.67 (0.43-1.06) |
| Q4 | 120/205 | 1.45 (0.99-2.13) | 1.65 (1.05-2.59) | 0.82 (0.48-1.40) |
| UMOD, pg/mL | ||||
| Log2(UMOD) | 400/819 | 0.78 (0.69-0.88) | 0.81 (0.72-0.92) | 0.89 (0.79-1.02) |
| Quartiles | ||||
| Q1 | 117/204 | 1.00 | 1.00 | 1.00 |
| Q2 | 114/205 | 0.93 (0.63-1.38) | 0.95 (0.63-1.43) | 1.24 (0.81-1.90) |
| Q3 | 79/205 | 0.47 (0.31-0.69) | 0.51 (0.34-0.77) | 0.64 (0.41-0.99) |
| Q4 | 90/205 | 0.58 (0.39-0.86) | 0.64 (0.42-0.97) | 0.91 (0.58-1.43) |
| EGF, pg/mL | ||||
| Log2(EGF) | 400/819 | 0.78 (0.71-0.85) | 0.75 (0.66-0.86) | 0.89 (0.77-1.04) |
| Quartiles | ||||
| Q1 | 131/204 | 1.00 | 1.00 | 1.00 |
| Q2 | 101/205 | 0.54 (0.36-0.80) | 0.55 (0.36-0.84) | 0.68 (0.43-1.06) |
| Q3 | 97/205 | 0.50 (0.34-0.74) | 0.47 (0.30-0.75) | 0.73 (0.44-1.22) |
| Q4 | 71/205 | 0.29 (0.19-0.44) | 0.27 (0.16-0.46) | 0.49 (0.27-0.91) |
| KIM-1, pg/mL | ||||
| Log2(KIM-1) | 400/819 | 1.00 (0.92-1.09) | 1.08 (0.97-1.22) | 1.00 (0.88-1.14) |
| Quartiles | ||||
| Q1 | 110/204 | 1.00 | 1.00 | 1.00 |
| Q2 | 95/205 | 0.74 (0.50-1.09) | 0.88 (0.58-1.36) | 0.77 (0.49-1.20) |
| Q3 | 82/205 | 0.57 (0.39-0.84) | 0.73 (0.46-1.16) | 0.64 (0.39-1.04) |
| Q4 | 113/205 | 1.05 (0.71-1.55) | 1.39 (0.83-2.33) | 1.06 (0.61-1.84) |
Model 1: Adjusted for age, sex (male and female), race (Black and White), time from interview to heart failure hospitalization (<5 and ≥5 years), urine creatinine level, diabetes (yes and no), hypertension (yes and no), and body mass index.
Model 2: Model 1 + adjusted for eGFR and log2(urine albumin level).
Abbreviations: A1M, alpha-1 microglobulin; CI, confidence interval; EGF, epidermal growth factor; eGFR, estimated glomerular filtration rate; KIM-1, kidney injury molecule-1; OR, odds ratio; Q, quartile; UMOD, uromodulin.
Discussion
In this study, using data from the REGARDS cohort, higher EGF was associated with a lower risk of future AKI following hospitalization for HF. Other measures of kidney tubular function and injury did not exhibit consistent associations with future risk of AKI after adjustment for potential confounders. These results are in line with prior studies in the Systolic Blood Pressure Intervention Trial (SPRINT) blood pressure trial, which showed that higher levels of EGF and UMOD were associated with future development of AKI in a subgroup of individuals with CKD.5,16, 17, 18, 19 In aggregate, these data suggest that biomarker levels of kidney tubule function provide important insights into future vulnerability to AKI in those admitted for acute HF exacerbation.
EGF level is a key mediator of renal tubular repair, promoting epithelial cell proliferation and survival after AKI.20 Urinary EGF level, primarily produced by distal tubular cells, declines during AKI and correlates with injury severity, making it a potential marker of tubular integrity and recovery capacity.20,21 Baseline EGF levels may plausibly reflect intrinsic tubular health and regenerative potential, with higher EGF suggesting preserved nephron integrity and a lower propensity for maladaptive repair. Prior studies have linked EGF levels with hypertension and chronic kidney disease progression.22, 23, 24 Studies linking higher urinary EGF excretion levels to favorable renal outcomes after cardiac surgery suggest its relevance in HF populations where recurrent AKI episodes are common.25
An interesting observation was that the odds of AKI differed inconsistently across biomarker quarters. Specifically, the third—but not the fourth—quartile of UMOD level was associated with lower odds of AKI. Quartile-based analyses are inherently subject to random fluctuation and are provided primarily to evaluate the functional form of the point estimates across quartiles. Even if the true relationship is monotonic, the observed ORs across categories may not decrease in a perfectly ordered fashion because of chance. The overlapping CIs between Q3 and Q4 would suggest that the difference between them may not be statistically meaningful. In addition, the association between UMOD level and AKI risk may not be strictly linear. It is possible that the strongest protective association occurs in the mid-to-high range (Q3), with attenuation at the extreme high end (Q4), potentially reflecting measurement variability at the high end of concentrations or residual confounding that differs across quartiles.
These results add to knowledge regarding risk factors detectable at times of health that can inform future risk of AKI. An understanding of the long-term ability of the kidney to withstand stressors could provide new opportunities for closer surveillance and prevention, particularly in vulnerable high-risk individuals. The identification of biomarkers associated with future AKI risk is an important first step toward profiling cumulative AKI risk. As has occurred with like acute conditions such as myocardial infarction and stroke, the early identification of individuals vulnerable to AKI could lead to novel measures to mitigate the effects of or prevent AKI.26
There are important strengths and limitations of this study. We used a novel design that enabled us to elucidate connections between baseline kidney health and AKI occurring in the distant future. We leveraged the REGARDS study, which is 1 of the nation’s largest cohorts with comprehensive health information over 2 decades of follow-up. Specially trained research personnel identified AKI using structured abstraction methods. Billing codes and administrative data have poor sensitivity for capturing AKI events in hospitalized patients.
Limitations of this study include a focus on adults over 45 years old. Although we identified AKI based upon the initial measured SCr level, AKI may have been in progress before hospital presentation. We did not limit AKI ascertainment to SCr levels obtained within a 48-hour window because shorter time windows may bias toward lower AKI detection.13 Urine output information was insufficient for classifying AKI. We had limited information on the course or acuity of HF hospitalizations. Although changes in SCr levels may result from fluid overload or over-diuresis, there are no validated methods for adjusting AKI classification to account for the course of HF treatment. We focused on printed medical records for specific hospitalizations; other hospitalizations or health data may be available with linkage through the electronic health records system. The REGARDS study used spot urine samples, which may not capture intraindividual or interindividual variability in the excretion of individual solutes and may have biased results toward the null. We selected modeling approaches that we felt were appropriate for the available data, including adjusting for baseline kidney function. Other studies have used alternate model approaches such as normalizing biomarker levels by urinary creatinine levels, but we do not expect markedly different results with alternate modeling strategies.27 Our study highlighted potential linkages between select tubule dysfunction marker levels and long-term AKI risk, but AKI may occur through separate pathways. We did not link to long-term mortality.
Renal functional reserve testing using protein load or stress ultrasonography has been shown to predict AKI risk. We were unable to compare renal functional reserve testing using protein loading or stress ultrasonography to the urine biomarker levels examined in this study. Renal functional reserve testing lacks standardized protocols and is difficult to implement retrospectively or at scale, limiting feasibility in both large epidemiologic cohorts and clinical practice, unlike urine biomarker levels.28
In conclusion, among community-dwelling adults, higher urine EGF levels were associated with future risk of AKI during acute HF hospitalization.
Article Information
Authors’ Full Names and Academic Degrees
Henry E. Wang, MD, MS, Jennifer A. Frey, PhD, Ashutosh Tamhane, MD, PhD, Lauren Shingler, MD, Alexander Keister, MD, Ching-Min Chu, BS, Emily B Levitan, ScD, Suzanne Judd, PhD, Alexander Bullen, MD, Edward D. Siew, MD, MSCI, Joseph V. Bonventre, MD, PhD, Michael Shlipak, MD, MPH, Byron Jaeger, PhD, Jesse Seegmiller, PhD, Joachim H. Ix, MD, Orlando M. Gutierrez, MD.
Authors’ Contributions
Research idea and study design: HEW, JAF, AT, EBL, SJ, JVB, MS, JI, OG. Data acquisition: HEW, JAF, CMC, SJ, JVB, MS, JI, OG. Data analysis/interpretation: HEW, AT, JI, OG. Statistical analysis: AT. Supervision or mentorship: HEW, JAF, JI, OG. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved.
Support
Supported by award R01-DK128803 from the National Institute of Diabetes and Digestive and Kidney Diseases. The parent REGARDS study was supported by cooperative agreement U01-NS041588 from the National Institute of Neurological Disorders and Stroke (NINDS) and National Institute on Aging (NIA), National Institutes of Health (NIH), Department of Health and Human Services (DHHS). Additional funding was provided by R01-HL80477 and R01-HL165452 from the National Heart, Lung, and Blood Institute (NHLBI). JVB was funded by R37-DK39773, R01-DK072381, and U01-DK085660. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NINDS or the NIA. The funders had no role in the study design, collection, analysis, or interpretation of data, writing the report, or the decision to submit the report for publication.
Financial Disclosure
Dr. Wang received grant funding from the National Institutes of Health, Department of Defense, and funds as site principal investigator for a study by Vasomune, Inc. Dr. Levitan reports consulting for Pfizer and membership in a Data and Safety Monitoring Board for the University of Pittsburgh. The remaining authors declare that they have no relevant financial interests.
Acknowledgements
We would like to acknowledge the following individuals for contributions to medical record abstraction: Linda Cho, Derick Campbell, Eliane De Jong, Clemence Gatete, Cheryl Herchek, Emma Helkey, Emma Helwagen, Ireland Hester, Dina McGowan, Isabella Rigg, Riley Riggenbach and John Rider.
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/.
Peer Review
Received January 6, 2026, as a submission to the expedited consideration track with 3 external peer reviews. Direct editorial input from an Associate Editor and the Editor-in-Chief. Accepted in revised form March 30, 2026.
Footnotes
Complete author and article information provided before references.
References
- 1.Kellum J.A., Romagnani P., Ashuntantang G., Ronco C., Zarbock A., Anders H.J. Acute kidney injury. Nat Rev Dis Primers. 2021;7(1):52. doi: 10.1038/s41572-021-00284-z. [DOI] [PubMed] [Google Scholar]
- 2.Bellomo R., Kellum J.A., Ronco C. Acute kidney injury. Lancet. 2012;380(9843):756–766. doi: 10.1016/s0140-6736(11)61454-2. [DOI] [PubMed] [Google Scholar]
- 3.Chahal R.S., Chukwu C.A., Kalra P.R., Kalra P.A. Heart failure and acute renal dysfunction in the cardiorenal syndrome. Clin Med (Lond) 2020;20(2):146–150. doi: 10.7861/clinmed.2019-0422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Banerjee D., Ali M.A., Wang A.Y.-M., Jha V. Acute kidney injury in acute heart failure-when to worry and when not to worry? Nephrol Dial Transplant. 2024;40(1):10–18. doi: 10.1093/ndt/gfae146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ix J.H., Shlipak M.G. The promise of tubule biomarkers in kidney disease: a review. Am J Kidney Dis. 2021;78(5):719–727. doi: 10.1053/j.ajkd.2021.03.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Allinovi M., Sessa F., Villa G., et al. Novel biomarkers for early detection of acute kidney injury and prediction of long-term kidney function decline after partial nephrectomy. Biomedicines. 2023;11(4):1046. doi: 10.3390/biomedicines11041046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Amatruda J.G., Estrella M.M., Garg A.X., et al. Urine alpha-1-microglobulin levels and acute kidney injury, mortality, and cardiovascular events following cardiac surgery. Am J Nephrol. 2021;52(8):673–683. doi: 10.1159/000518240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gao L., Zhong X., Jin J., Li J., Meng X.M. Potential targeted therapy and diagnosis based on novel insight into growth factors, receptors, and downstream effectors in acute kidney injury and acute kidney injury-chronic kidney disease progression. Signal Transduct Target Ther. 2020;5(1):9. doi: 10.1038/s41392-020-0106-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Howard V.J., Cushman M., Pulley L., et al. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology. 2005;25(3):135–143. doi: 10.1159/000086678. [DOI] [PubMed] [Google Scholar]
- 10.Bailey L.N., Levitan E.B., Judd S.E., et al. Association of urine albumin excretion with incident heart failure hospitalization in community-dwelling adults. JACC Heart Fail. 2019;7(5):394–401. doi: 10.1016/j.jchf.2019.01.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120(4):c179–c184. doi: 10.1159/000339789. [DOI] [PubMed] [Google Scholar]
- 12.Mehta R.L., Kellum J.A., Shah S.V., et al. Acute kidney injury network: report of an initiative to improve outcomes in acute kidney injury. Crit Care. 2007;11(2):R31. doi: 10.1186/cc5713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Swan J.T., Moore L.W., Sparrow H.G., Frost A.E., Gaber A.O., Suki W.N. Optimization of acute kidney injury (AKI) time definitions using the electronic health record: a first step in automating in-hospital AKI detection. J Clin Med. 2021;10(15):3304. doi: 10.3390/jcm10153304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Inker L.A., Eneanya N.D., Coresh J., et al. Chronic Kidney Disease Epidemiology Collaboration New creatinine- and cystatin C-based equations to estimate GFR without race. N Engl J Med. 2021;385(19):1737–1749. doi: 10.1056/NEJMoa2102953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Vandenbroucke J.P., von Elm E., Altman D.G., et al. Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. Ann Intern Med. 2007;147(8):W163–W194. doi: 10.7326/0003-4819-147-8-200710160-00010-w1. [DOI] [PubMed] [Google Scholar]
- 16.Bullen A.L., Katz R., Jotwani V., et al. Biomarkers of kidney tubule health, CKD progression, and acute kidney injury in SPRINT (systolic blood pressure intervention trial) participants. Am J Kidney Dis. 2021;78(3):361–368.e1. doi: 10.1053/j.ajkd.2021.01.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bullen A.L., Katz R., Lee A.K., et al. The SPRINT trial suggests that markers of tubule cell function in the urine associate with risk of subsequent acute kidney injury while injury markers elevate after the injury. Kidney Int. 2019;96(2):470–479. doi: 10.1016/j.kint.2019.03.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Malhotra R., Katz R., Jotwani V., et al. Urine markers of kidney tubule cell injury and kidney function decline in SPRINT trial participants with CKD. Clin J Am Soc Nephrol. 2020;15(3):349–358. doi: 10.2215/cjn.02780319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Malhotra R., Craven T., Ambrosius W.T., et al. Effects of intensive blood pressure lowering on kidney tubule injury in CKD: a longitudinal subgroup analysis in SPRINT. Am J Kidney Dis. 2019;73(1):21–30. doi: 10.1053/j.ajkd.2018.07.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ono K., Maeshima A., Nagayama I., Kubo T., Yagisawa T., Nagata D. Urinary epidermal growth factor level as a noninvasive indicator of tubular repair in patients with acute kidney injury. Diagnostics (Basel) 2024;14(9):947. doi: 10.3390/diagnostics14090947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tang J., Liu N., Zhuang S. Role of epidermal growth factor receptor in acute and chronic kidney injury. Kidney Int. 2013;83(5):804–810. doi: 10.1038/ki.2012.435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Geurts F., van Heugten M.H., Blijdorp C.J., Fenton R.A., Chaker L., Hoorn E.J. Urinary EGF reflects distal tubular mass and is associated with hypertension, serum magnesium, and kidney outcomes. Kidney360. 2025;6(3):451–460. doi: 10.34067/kid.0000000687. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Menez S., Wen Y., Xu L., et al. The ASSESS-AKI Study found urinary epidermal growth factor is associated with reduced risk of major adverse kidney events. Kidney Int. 2023;104(6):1194–1205. doi: 10.1016/j.kint.2023.08.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ledeganck K.J., Van Eyck A., Wouters K., et al. Urinary epidermal growth factor reflects vascular health in boys with either obesity or type 1 diabetes. A role for renin, or beyond? PLoS One. 2023;18(3) doi: 10.1371/journal.pone.0283716. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Menez S., Ju W., Menon R., et al. Urinary EGF and MCP-1 and risk of CKD after cardiac surgery. JCI Insight. 2021;6(11) doi: 10.1172/jci.insight.147464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Centers for Disease Control and Prevention (CDC) Ten great public health achievements--United States, 2001-2010. MMWR Morb Mortal Wkly Rep. 2011;60(19):619–623. [PubMed] [Google Scholar]
- 27.Hsu C-y, Chinchilli V.M., Coca S., et al. ASSESS-AKI Investigators Post–acute kidney injury proteinuria and subsequent kidney disease progression: the assessment, serial evaluation, and subsequent sequelae in acute kidney injury (ASSESS-AKI) study. JAMA Intern Med. 2020;180(3):402–410. doi: 10.1001/jamainternmed.2019.6390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Husain-Syed F., Ferrari F., Sharma A., et al. Preoperative renal functional reserve predicts risk of acute kidney injury after cardiac operation. Ann Thorac Surg. 2018;105(4):1094–1101. doi: 10.1016/j.athoracsur.2017.12.034. [DOI] [PubMed] [Google Scholar]