Abstract
Rationale & Objective
Uromodulin (UMOD) is a kidney tubule biomarker that can be measured in blood and urine and may help identify persons with chronic kidney disease (CKD) at greater risk of progression. Whether blood or urine UMOD is more strongly associated with CKD progression is unknown.
Study Design
This was a secondary analysis of the Systolic Blood Pressure Intervention Trial (SPRINT), in which an intensive systolic blood pressure target reduced risk of cardiovascular events and mortality.
Setting & Population
A total of 2,302 SPRINT participants with an estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m2.
Exposures
Plasma and urine UMOD measured at baseline.
Outcomes
Acute kidney injury, 30% decline in eGFR, end-stage kidney disease, and annualized percentage change in eGFR.
Analytical Approach
Cox proportional hazards (acute kidney injury, 30% eGFR decline, and end-stage kidney disease) and linear mixed models (annual percent eGFR change) were adjusted for baseline characteristics, including eGFR and albuminuria.
Results
Among included participants, mean age was 73 years, 41% were female, and median eGFR was 50 (interquartile range: 40-57) mL/min/1.73 m2. Median plasma and urine UMOD levels were 17.2 μg/L and 6.5 mg/L, respectively. Plasma and urine UMOD had a Spearman correlation of 0.40. In fully adjusted models, each standard deviation higher plasma UMOD level was associated with lower risk of 30% eGFR decline (hazard ratio=0.84, 95% confidence interval 0.75-0.94) and slower percent annual eGFR decline (+0.58, 95% confidence interval +0.22 to +0.94). In contrast, urine UMOD levels had weaker associations with these outcomes that did not remain after adjustment for baseline eGFR, albuminuria, and other risk factors.
Conclusions
Among persons with hypertension and CKD, plasma UMOD levels were associated with slower decline in eGFR and reduced risk of 30% eGFR decline after adjustment for demographic and clinical characteristics, whereas urine UMOD levels were not. When evaluating risk of kidney endpoints, plasma rather than urine may be the preferable specimen for UMOD measurement.
Index Words: acute kidney injury, chronic kidney disease, end-stage kidney disease, hypertension, uromodulin
Plain-Language Summary
Uromodulin is a protein produced in the kidney, which can be measured in the blood and urine. Lower levels of uromodulin in both the blood and urine are known to be associated with worse kidney health; however, it is not known whether the association with blood or urine uromodulin is stronger. We evaluated both blood and urine uromodulin with subsequent kidney health and observed more consistent associations between lower blood uromodulin and worse kidney health. These findings suggest measuring uromodulin in the blood, as opposed to the urine, may be the ideal approach for evaluating kidney health with uromodulin.
Chronic kidney disease (CKD) affects an estimated 850 million people worldwide.1 CKD progression is associated with substantial morbidity and mortality, including end-stage kidney disease (ESKD); however, it can be difficult to identify which patients with CKD will progress.2,3 Noninvasive, clinically available diagnostic tests to identify persons with CKD at high risk of progression are currently limited to estimated glomerular filtration rate (eGFR) and albuminuria, which reflect glomerular function and integrity. However, it is known that tubule atrophy and interstitial fibrosis are strongly associated with CKD progression and poorly captured by these clinical biomarkers.4 Novel biomarkers of tubule health may provide an opportunity to better identify persons with CKD at highest risk of kidney function decline, particularly among persons with nondiabetic CKD and low-grade or altogether absent albuminuria.5
Uromodulin (UMOD) is a protein produced exclusively in the thick ascending limb of the loop of Henle and found at high concentrations in the urine. Urinary UMOD is thought to have a role in the prevention of urinary stone formation and urinary tract infections.6 Higher urinary UMOD concentrations have also been associated with lower risk of acute kidney injury (AKI), incident CKD, and kidney function decline.7,8 However, there are analytical challenges inherent to clinical use of urine UMOD levels because of its propensity to aggregate in the urine.9 Although UMOD is exclusively produced in the tubules and primarily secreted into the tubule lumen, it is also detectable in the bloodstream, offering an alternative specimen for UMOD measurement to assess risk of kidney outcomes. As with urine UMOD levels, prior studies have reported that higher blood UMOD levels are also associated with lower risk of adverse kidney outcomes.10, 11, 12, 13, 14, 15 However, nearly all studies have evaluated the associations of blood uromodulin with kidney outcomes without accompanying urine uromodulin measures for comparison. Thus, it remains uncertain whether blood or urine UMOD levels better identify persons at risk for subsequent kidney disease progression.
We hypothesized that plasma uromodulin would have stronger associations with adverse kidney outcomes within a population with CKD. Among Systolic Blood Pressure Intervention Trial (SPRINT, ClinicalTrials.gov NCT01206062) participants with CKD, we have measured urine and plasma UMOD levels in the same individuals at baseline.16 We have recently reported that both plasma and urine UMOD have similar associations with cardiovascular disease (CVD) endpoints.17,18 Whether plasma or urine UMOD levels have stronger relationships with kidney endpoints is uncertain. Thus, we evaluated urine and plasma UMOD associations with kidney outcomes simultaneously among persons with hypertension and CKD who participated in SPRINT.
Methods
Study Design
SPRINT was a clinical trial in which 9,361 participants with hypertension and elevated risk of CVD were randomized (1:1) to either an intensive systolic blood pressure (SBP) target of <120 mm Hg or a standard SBP target of <140 mm Hg for prevention of a primary composite CVD outcome.19 Substantially lower risk of CVD and death were observed in participants randomized to the intensive SBP target, and the trial ended early on August 20, 2015 on the recommendation of the data and safety monitoring board. Institutional review boards of all participating institutions approved the study.
Participants
SPRINT inclusion criteria included SBP >130 mm Hg, age >50 years and increased CVD risk based on any of age ≥75 years, prevalent clinical or subclinical CVD, eGFR <60 mL/min/1.73 m2, or 10-year Framingham risk >15% at baseline. Major SPRINT exclusion criteria included diabetes, recent acute heart failure or known reduced ejection fraction, prior stroke or transient ischemic attack, eGFR <20 mL/min/1.73 m2, or proteinuria >1 g/day. This study used the SPRINT-CKD subset, which included individuals with baseline eGFR <60 mL/min/1.73 m2 by the 2012 CKD Epidemiology Collaboration combined creatinine and cystatin C equation.20 The analysis was considered institutional review board–exempt by the committees on human research at the University of California, San Francisco, the San Francisco Veterans Affairs Health Care System, and was approved by the Institutional Review Board at Veterans Affairs San Diego Healthcare System.
UMOD Measurement
Plasma and urine specimens were collected at baseline and stored at −80 °C until they were thawed for measurement. Plasma UMOD levels were measured using a single-plex electrochemiluminescent immunoassay by Meso Scale Discovery (Rockville, MD, USA). Urine UMOD was measured using a multiplex assay on a Meso Scale Discovery platform. The plasma UMOD lower limit of detection was 2.0 μg/L, and the upper limit of detection was 10,000 μg/L. The urine UMOD analytic range was 0.6-2510 μg/L.
Covariates
Demographic and clinical characteristics were selected as potential confounding factors based on hypothesized associations with kidney health. Covariates included randomization assignment, age, gender, race, Hispanic ethnicity, smoking status, creatinine- and cystatin C-based eGFR, urine albumin, urine creatinine, prevalent CVD, prevalent heart failure, SBP, body mass index, total cholesterol, high-density lipoprotein cholesterol, number of antihypertensive medications, and use of diuretics, angiotensin-converting enzyme-inhibitor and angiotensin-receptor blockers, beta-blockers, and statins.21
Outcomes
We evaluated 4 kidney-related outcomes. AKI was determined based on serious adverse event reporting recorded during the clinical trial, which was assessed through structured interviews quarterly during the trial and adjudicated by a serious adverse event committee. CKD progression was defined by a 30% decline in eGFR from baseline based on serum creatinine obtained at 3-month intervals during follow-up; a confirmatory measure was not required. Because cystatin C measurements were only available at the baseline study visit, 30% decline in eGFR was based only on creatinine-based eGFR estimates, using the race-free CKD Epidemiology Collaboration equation at both baseline and follow-up.21 We also evaluated annualized percentage change in eGFR from baseline. Finally, incident ESKD required the initiation of dialysis or receipt of a kidney transplant. Event assessment for AKI and eGFR decline continued after the end of the trial until July 29, 2016. ESKD events were identified through linkage with the United States Renal Data System and included follow-up through 2019.
Analysis
We summarized the baseline characteristics of participants. Characteristics were then summarized within quartiles of plasma UMOD levels. Event numbers and rates were summarized by quartiles of plasma UMOD levels and urine UMOD levels. Spearman correlation coefficients were derived for plasma UMOD levels, urine UMOD levels, and eGFR. Mean plasma and urine UMOD were also summarized across eGFR groups of 5 mL/min/1.73 m2.
For longitudinal analyses of each outcome, plasma and urine UMOD levels were log-transformed to correct for right skew and standardized to a mean of 0 and standard deviation (SD) of 1 to facilitate comparisons of effect sizes. Models evaluating urine UMOD levels were additionally adjusted for log-transformed urine creatinine to account for overall urine concentration at the time of urine collection. We modeled the log-scaled biomarkers as continuous linear predictors.
We used Cox proportional hazards models for our AKI and ESKD outcomes. Given the extended follow-up period for our analysis of ESKD, more deaths occurred, raising concerns about the competing risk of death. To address this, we employed the Fine and Gray method to evaluate the association between UMOD levels and incident ESKD, reporting the subdistribution hazard ratio.22 For the outcome of ≥30% eGFR decline, we employed interval censored Cox proportional hazards models with time to event measured up to the last study visit. The proportional hazards assumption was assessed using Schoenfeld residuals. Linear mixed models with random intercepts, random slopes, and an unstructured covariance structure were used to evaluate UMOD associations with annualized eGFR slope. eGFR was log-transformed in analyses of eGFR slope to allow interpretation of the slope as annualized percent change. To assess functional form, eGFR slope within quartiles of plasma and urine UMOD levels were estimated.
Additional analyses were stratified by baseline eGFR (<45 vs ≥45 mL/min/1.73 m2) and by randomization assignment (intensive vs usual arm) with testing for multiplicative interaction to determine whether associations between UMOD levels and the outcomes differed between strata.
Results
Baseline Characteristics
There were 2,302 SPRINT participants included (Fig S1). Compared with the 2,302 included participants, the 206 SPRINT participants with CKD who were excluded because of missing samples were more frequently male (71% vs 59%) and African American (38% vs 25%), with lower median eGFR (46 vs 50 mL/min/1.73 m2) and higher median urine albumin-to-creatinine ratio (UACR) (31 vs 14 mg/g) (Table S1). At baseline, the mean age of included participants was 73 (SD = 9) years, 41% were women, the median eGFR was 50 (interquartile range: 40-57) mL/min/1.73 m2, and the median UACR was 14 (interquartile range: 7-45) mg/g (Table 1). The median plasma and urine UMOD levels were 17.2 μg/L and 6.5 mg/L, respectively. On average, those with higher plasma UMOD levels were older, more often female, had higher eGFR, and lower UACR. We also examined UMOD levels across categories of eGFR, finding progressively higher levels of plasma and urine UMOD among those with higher eGFR (Fig 1). Plasma and urine UMOD levels had a Spearman correlation of 0.40. Spearman correlation with eGFR was 0.47 and 0.36 for plasma and urine UMOD levels, respectively.
Table 1.
Baseline Characteristics and Randomization by Plasma Uromodulin Quartile (N = 2302)
| Characteristic | Plasma UMOD, μg/L |
|||
|---|---|---|---|---|
| Q1: 2.0-12.2 (N = 575) | Q2: 12.3-17.2 (N = 576) | Q3: 17.2-23.2 (N = 576) | Q4: 23.2-117.7 (N = 575) | |
| Intensive arm | 310 (54%) | 283 (49%) | 281 (49%) | 298 (52%) |
| Age, y | 73 (63-79) | 74 (66-79) | 75 (67-80) | 76 (69-81) |
| Female | 223 (39%) | 206 (36%) | 219 (38%) | 304 (53%) |
| Race | ||||
| White | 408 (71%) | 408 (71%) | 438 (76%) | 445 (77%) |
| African American | 159 (28%) | 160 (28%) | 125 (22%) | 125 (22%) |
| Other | 8 (1%) | 8 (1%) | 13 (2%) | 5 (1%) |
| Hispanic | 47 (8%) | 31 (5%) | 41 (7%) | 40 (7%) |
| Smoking | ||||
| Current | 72 (13%) | 53 (9%) | 40 (7%) | 31 (5%) |
| Former | 253 (44%) | 271 (47%) | 268 (47%) | 239 (42%) |
| Never | 250 (43%) | 252 (44%) | 268 (47%) | 305 (53%) |
| eGFR, mL/min/1.73 m2 | 38 (30-49) | 49 (41-56) | 52 (45-58) | 56 (50-60) |
| Urine ACR, mg/g | 29 (11-145) | 17 (7-54) | 12 (6-31) | 10 (6-21) |
| BMI, kg/m2 | 29 (26-34) | 29 (26 - 33) | 29 (26-32) | 28 (25-32) |
| Systolic BP, mm Hg | 137 (128-148) | 140 (130 - 151) | 139 (130-147) | 138 (129-150) |
| No. antihypertensive meds | 2.0 (2.0 - 3.0) | 2.0 (1.0 - 3.0) | 2.0 (1.0-3.0) | 2.0 (1.0-3.0) |
| Antihypertensive med class | ||||
| Diuretic | 353 (61%) | 310 (54%) | 309 (54%) | 288 (50%) |
| ACE inhibitor/ ARB | 375 (65%) | 367 (64%) | 363 (63%) | 329 (57%) |
| Beta-blocker | 313 (54%) | 262 (45%) | 258 (45%) | 239 (42%) |
| Prevalent CVD | 182 (32%) | 171 (30%) | 143 (25%) | 151 (26%) |
| Prevalent heart failure | 59 (10%) | 35 (6%) | 26 (5%) | 22 (4%) |
| HDL cholesterol, mg/dL | 48 (41-58) | 49 (41-58) | 50 (42-60) | 53 (46-64) |
| Total cholesterol, mg/dL | 178 (153-206) | 174 (153-203) | 180 (154-208) | 181 (158-208) |
| Statin use | 299 (53%) | 295 (51%) | 302 (53%) | 290 (51%) |
| Urine UMOD, mg/L | 4.5 (2.8-6.5) | 6.4 (4.2-9.6) | 7.2 (4.9-10.7) | 8.8 (6.1-12.1) |
| Urine creatinine, mg/dL | 101 (65-150) | 114 (75-166) | 118 (77-162) | 110 (70-167) |
| Urine UMOD/cr, mg/g | 4.6 (3.0-6.7) | 5.9 (4.0-8.7) | 6.7 (4.4-9.6) | 8.4 (5.5-12.4) |
Data displayed as N (%) or median [interquartile range].
Abbreviations: ACR, albumin-to-creatinine ratio; BMI, body mass index; BP, blood pressure; CKD, chronic kidney disease; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate by creatinine and cystatin C; HDL, high-density lipoprotein cholesterol; HF, heart failure; SPRINT, Systolic Blood Pressure Intervention Trial.
Figure 1.
Mean plasma uromodulin and urine uromodulin-to-creatinine ratio across groupings of baseline eGFR by 5 mL/min/1.73 m2. Each symbol denotes mean plasma or urine UMOD level with 95% CI represented by spike. X-axis displays eGFR rounded up to the nearest 5 point category. Abbreviations: UMOD, uromodulin; eGFRcyscr, estimated glomerular filtration rate by creatinine and cystatin C; UMOD/cr, uromodulin-to-creatinine ratio.
Association of Plasma and Urine UMOD With Kidney Outcomes
Participants received a mean follow-up of 3.9 years for the outcomes of 30% and annualized eGFR decline, whereas mean follow-up for the outcome of end-stage kidney disease was 6.8 years. The incidence of all adverse kidney outcomes was highest in the lowest plasma and urine UMOD quartiles and lowest in the highest quartiles (Table 2). In models adjusted only for age, gender, and randomization, higher plasma and urine UMOD levels were both associated with lower incidence of AKI, 30% eGFR decline, ESKD, and with a slower annual eGFR decline (Fig 2 and Table 3). In each case, the point estimate was larger in magnitude for plasma versus urine UMOD. In fully adjusted models, associations of urine UMOD with each endpoint were no longer statistically significant (P > 0.05). Plasma UMOD levels remained significantly associated with 30% decline in eGFR (hazard ratio 0.84, 95% confidence interval [CI] 0.75-0.94) and annual percentage change in eGFR (+0.58, 95% CI +0.22 to +0.94) after adjustment (Table 3). The attenuation of the association between plasma UMOD levels and 30% eGFR decline appeared to be mostly attributable to adjustment for eGFR and urine albumin (Table S2). The association of plasma UMOD levels with incident AKI did not reach statistical significance in the final model (P = 0.06). Neither plasma nor urine UMOD levels were clearly associated with the ESKD endpoint (P > 0.5). When both plasma and urine UMOD were included in models concurrently, the findings remained similar: only plasma UMOD levels retained significant associations with both 30% eGFR decline (hazard ratio 0.82, 95% CI 0.73-0.92) and annual percentage change in eGFR (+0.61, 95% CI +0.24 to +0.99).
Table 2.
Incidence of Kidney Outcomes by Uromodulin Quartiles
| Outcome | Plasma UMOD |
|||||||
|---|---|---|---|---|---|---|---|---|
| Quartile 1, n = 575 |
Quartile 2, n = 576 |
Quartile 3, n = 576 |
Quartile 4, n = 575 |
|||||
| N | (%) | N | (%) | N | (%) | N | (%) | |
| Acute kidney injury | 72/575 | (12.5%) | 44/576 | (7.6%) | 33/576 | (5.7%) | 30/575 | (5.2%) |
| 30% eGFR decline | 162/563 | (28.8%) | 113/569 | (19.9%) | 101/569 | (17.8%) | 85/567 | (15.0%) |
| Outcome | Urine UMOD | |||||||
|
Quartile 1 (n = 575) |
Quartile 2 (n = 576) |
Quartile 3 (n = 576) |
Quartile 4 (n = 575) |
|||||
|
N |
(%) |
N |
(%) |
N |
(%) |
N |
(%) |
|
| Acute kidney injury | 69/575 | (12.0%) | 49/576 | (8.5%) | 32/576 | (5.6%) | 29/575 | (5.0%) |
| 30% eGFR decline | 150/565 | (26.5%) | 109/568 | (19.2%) | 103/567 | (18.2%) | 99/568 | (17.4%) |
End-stage kidney disease omitted because of less than 10 events in some quartiles.
Abbreviations: UMOD, uromodulin.
Figure 2.
Annual change in eGFR by quartile of plasma and urine uromodulin. Estimates with 95% CI from unadjusted linear mixed models of eGFR over time.
Table 3.
Associations of Plasma and Urine Uromodulin With Kidney Outcomes in Multivariable Models
| Outcome | Model 1 |
Model 2 |
||
|---|---|---|---|---|
| HR (95% CI) | P | HR (95% CI) | P | |
| Acute kidney injury (n = 179) | ||||
| Plasma UMOD | 0.64 (0.55, 0.73) | <0.001 | 0.84 (0.70, 1.01) | 0.06 |
| Urine UMOD | 0.80 (0.72, 0.88) | <0.001 | 0.90 (0.79, 1.03) | 0.14 |
| 30% eGFR decline (n = 461) | ||||
| Plasma UMOD | 0.69 (0.63, 0.75) | <0.001 | 0.84 (0.75, 0.94) | 0.002 |
| Urine UMOD | 0.88 (0.81, 0.96) | 0.004 | 1.04 (0.93, 1.15) | 0.53 |
| End-stage kidney disease (n = 72) | ||||
| Plasma UMOD | 0.41 (0.34, 0.50) | <0.01 | 0.91 (0.70, 1.19) | 0.51 |
| Urine UMOD | 0.68 (0.61, 0.76) | <0.01 | 0.94 (0.76, 1.16) | 0.56 |
| Annual % change in eGFR | Estimate (95% CI) | Estimate (95% CI) | ||
| Plasma UMOD | +1.07 (+0.69, +1.45) | <0.001 | +0.58 (+0.22, +0.94) | 0.002 |
| Urine UMOD | +0.45 (+0.08, +0.83) | 0.017 | +0.13 (−0.22, +0.48) | 0.47 |
Hazard ratios depicted per SD increase (log scale) of plasma uromodulin and urine uromodulin. All urine UMOD models additionally control for urine creatinine. Model 1: intervention, age and gender; Model 2: model 1 plus race, eGFR, log urine albumin, log urine creatinine, smoking, body mass index, systolic blood pressure, number and type of antihypertensive medication, prevalent cardiovascular disease, heart failure, high-density lipoprotein, total cholesterol, and statin use. Estimates with P < 0.05 are in bold.
Interaction and Subgroup Analysis
We observed a stronger association between plasma UMOD levels and reduced annual eGFR decline in those with baseline eGFR <45 compared with those with eGFR ≥45 (mL/min/1.73 m2; P for interaction <0.01) (Table 4). No other significant interactions were observed between plasma UMOD levels and baseline eGFR category for the other kidney outcomes. We also observed a significant interaction between plasma UMOD levels and randomization arm for the outcome of >30% eGFR decline. Although higher plasma UMOD was associated with a lower risk of 30% eGFR decline across both arms, the association appeared to be stronger in those assigned to the standard SBP target than in those assigned to the intensive target (Table 4).
Table 4.
Associations of Plasma and Urine Uromodulin Levels With Kidney Outcomes by Baseline eGFR and SBP Target
| Outcome | eGFR < 45 mL/min/1.73 m2 |
eGFR ≥ 45 mL/min/1.73 m2 |
P for interaction |
|---|---|---|---|
| HR (95% CI) | HR (95% CI) | ||
| Acute kidney injury | |||
| Plasma UMOD | 0.71 (0.58, 0.88) | 0.93 (0.72, 1.22) | 0.10 |
| Urine UMOD | 0.89 (0.76, 1.04) | 0.88 (0.72, 1.06) | 0.88 |
| 30% eGFR decline | |||
| Plasma UMOD | 0.84 (0.74, 0.96) | 0.78 (0.67, 0.90) | 0.41 |
| Urine UMOD | 0.99 (0.88, 1.10) | 1.00 (0.86, 1.17) | 0.85 |
| End-stage kidney disease | |||
| Plasma UMOD | 0.60 (0.46, 0.77) | 0.95 (0.49, 1.83) | 0.51 |
| Urine UMOD | 0.84 (0.70, 1.02) | 1.68 (0.82, 3.43) | 0.02 |
| Annual % eGFR change | Estimate (95% CI) | Estimate (95% CI) | |
| Plasma UMOD | +1.48 (+0.70, +2.27) | −0.08 (−0.41, +0.26) | 0.001 |
| Urine UMOD | +0.59 (-0.12, +1.30) | −0.37 (−0.75, +0.02) | 0.01 |
| Outcome | Intensive SBP target | Usual SBP target | Pfor interaction |
| Acute kidney injury | |||
| Plasma UMOD | 0.81 (0.66, 0.99) | 0.91 (0.70, 1.17) | 0.44 |
| Urine UMOD | 0.87 (0.75, 1.01) | 0.99 (0.78, 1.27) | 0.31 |
| 30% eGFR decline | |||
| Plasma UMOD | 0.90 (0.79, 1.02) | 0.73 (0.62, 0.87) | 0.03 |
| Urine UMOD | 1.09 (0.96, 1.24) | 0.93 (0.80, 1.08) | 0.07 |
| End-stage kidney disease | |||
| Plasma UMOD | 0.73 (0.53, 1.03) | 1.25 (0.92, 1.70) | 0.12 |
| Urine UMOD | 0.87 (0.69, 1.10) | 1.21 (0.66, 2.22) | 0.46 |
| Annual % eGFR change | Estimate (95% CI) | Estimate (95% CI) | |
| Plasma UMOD | +0.62 (+0.06, +1.18) | +0.53 (+0.07, +0.99) | 0.81 |
| Urine UMOD | +0.05 (−0.44, +0.55) | +0.22 (−0.23, +0.66) | 0.62 |
Effect sizes depicted per SD increase (log scale) of plasma uromodulin and urine uromodulin, respectively. All urine UMOD models additionally control for urine creatinine. Variables included are intervention, age, gender, race, eGFR, log urine albumin, log urine creatinine, smoking, body mass index, systolic blood pressure, number and type of antihypertensive medications, prevalent cardiovascular disease, heart failure, high-density lipoprotein, total cholesterol, and statin use. Estimates with 95% CIs that excluded the null and P < 0.05 are in bold.
Discussion
We had hypothesized that plasma UMOD levels would be more strongly associated with adverse kidney outcomes among hypertensive trial participants with CKD than urine UMOD levels. We found that this was the case. Specifically, in fully adjusted models, higher plasma UMOD levels were associated with slower declines in eGFR and reduced risk of 30% eGFR decline, even after adjustment for demographic and clinical characteristics, whereas urine UMOD levels were not.
UMOD is the most abundant protein in the urine in healthy individuals, making it an easily measurable intrinsic product of normal tubule function. However, urinary uromodulin consists of 2 forms, one of which polymerizes in what is thought to be essential to its ability to protect the tubule lumen from bacterial infection and prevent kidney stone formation.6,23 Additionally the urinary concentration of UMOD, as with all urine biomarkers, is subject to change according to hydration status.24 Whereas UMOD concentrations in the blood are much smaller than in the urine, the absence of these potentially confounding factors may underlie the robustness of our observed plasma UMOD level associations with less eGFR decline.
Two prior studies have simultaneously evaluated blood and urine UMOD levels with kidney outcomes, and results from these studies were similar to the results reported here. In a cohort of patients with CKD recruited in Germany, higher serum uromodulin levels, but not urine uromodulin levels were associated with lower risk of major adverse kidney outcomes over a median 6.5 years of follow-up, similar to our study findings. However, eGFR trajectory was not included as an outcome in the study.25 In another study of patients with biopsy-proven immunoglobulin A nephropathy and eGFR > 60 mL/min/1.73 m2, higher serum UMOD levels were associated with lower risk of 30% eGFR decline whereas urinary UMOD levels were not.26 We extend these findings to a unique setting, among persons with hypertension and nondiabetic CKD, and demonstrate that plasma UMOD levels provide a stronger and more consistent signal for risk of kidney disease progression in persons with nondiabetic CKD.
Because SPRINT participants in the intensive arm received more aggressive blood pressure lowering and had acute hemodynamic changes in eGFR early in the trial, it is possible that associations of plasma and urine UMOD levels with eGFR changes in this arm were obscured. In particular, our finding that the association between higher plasma UMOD levels and lower incidence of 30% eGFR decline was apparent in the usual SBP target group and not the intensive SBP target group may be the result of such an effect. Similar observations have been made in other studies examining biomarkers among persons subjected to intensive blood pressure lowering.27
It is well established that UMOD is an immunomodulatory glycoprotein produced exclusively in the thick ascending limb of the Loop of Henle and primarily secreted into the tubule lumen with small amounts being detectable in the blood.28, 29, 30, 31 With a molecular weight of 100 kDa, circulating UMOD is unlikely to pass through the glomerular filter. More recent work has suggested processing and regulation of a polymerizing form, comprising most UMOD found in urine, that is separate from the nonpolymerizing form which is the only form found in the blood.32,33 Further observations suggest that tubule injury results in an increase in UMOD production, basolateral localization, and higher concentrations in circulation with an accompanying decrease in kidney inflammation.34 Our findings of a stronger inverse association between plasma UMOD and kidney function decline may reflect a more specific relationship between plasma than urine UMOD levels with a protective response to kidney injury. This hypothesis warrants further investigation.
Our finding that plasma UMOD has a stronger protective association for annualized eGFR trajectory among those with eGFR <45 mL/min/1.73 m2 is novel, but warrants caution in interpretation. In previous studies evaluating serum uromodulin and kidney outcomes, including an analysis of the previously mentioned Germany-based study of persons with CKD and another analysis of the Cardiovascular Health Study, no interaction was observed between serum uromodulin levels and eGFR with respect to end-stage kidney disease, a relatively infrequent outcome. Those studies, however, did not evaluate the outcomes of eGFR decline or annual eGFR trajectory, which may provide greater statistical power.11,12 To this point, our analysis of plasma UMOD and ESKD did not observe statistically significant association in our study population within and across eGFR subgroups; but because of the relatively few ESKD events, this result should be interpreted cautiously. Other studies have also evaluated interaction between urine uromodulin and eGFR with respect to eGFR decline with results suggesting similar associations between urine UMOD levels and eGFR decline across baseline eGFR levels.35,36
Strengths of our study include concurrent evaluation of UMOD levels in urine and plasma, measurement of a variety of potential confounders, and follow-up of eGFR over time. Among notable limitations are the exclusion of persons with diabetes, proteinuria >1 g/day and eGFR ≥60 mL/min/1.73 m2, with consequent limitations in generalizability. Nonetheless, the similar findings in a population with IgA nephropathy and in the general population otherwise support our findings and their generalizability. Finally, the possibility of residual confounding remains present despite our adjustment for several covariates.
In summary, we found that higher plasma UMOD levels were associated with lower risk of 30% eGFR decline and slower declines in eGFR, associations that appeared more robust than those observed for urine UMOD levels after adjustment for potential confounders, including eGFR and albuminuria. Plasma UMOD levels may provide a more precise indicator of kidney tubule health and risk for kidney-related outcomes than urine UMOD levels. Given our findings and those of prior studies, plasma specimens may be preferrable to urine specimens for UMOD assessment when attempting to identify persons at risk for CKD progression.
Article Information
Authors’ Full Names and Academic Degrees
Jesse C. Ikeme, MD, MAS, Rebecca Scherzer, PhD, Pranav S. Garimella, MBBS, MPH, Stein I. Hallan, MD, PhD, Ronit Katz, DPhil, Michelle M. Estrella, MD, MHS, Joachim H. Ix, MD, MAS, Michael G. Shlipak, MD, MPH
Authors’ Contributions
Research idea and study design: JCI, JHI, MGS; data acquisition: JHI, MGS; data analysis/interpretation: JCI, RS, PSG, SIH, RK, MME, JHI, MGS; statistical analysis: RS, RK; supervision or mentorship: JHI, MGS. 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
RK, RS, PSG, JHI, and MGS receive research support from the National Institute of Diabetes and Digestive and Kidney Diseases (grant number R01-DK098234) for their effort in this research. PSG is supported by K23 Career Development award from National Institute of Diabetes and Digestive and Kidney Diseases (DK114556). The funders of this study had no role in study design; collection, analysis, and interpretation of data; writing the report; or the decision to submit the report for publication.
Financial Disclosure
MME and MGS receive research funding from Bayer, Inc. MME has received an honorarium from Boehringer-Ingelheim, Inc. MGS has received honoraria from Bayer, Boehringer-Ingelheim, Inc, and Astra Zeneca, and has received consulting fees from Cricket Health and Intercept Pharmaceutical. MGS previously served as an advisor to and held stock in TAI Diagnostics. JHI has received honoraria from Astra Zeneca, Bayer, Jnana, and Marius Pharmaceuticals in the past 2 years. PSG receives speaking fees from Otsuka, Inc. The other authors declare that they have no relevant financial interests.
Data Sharing
The data underlying this article will be shared upon reasonable request to the corresponding author.
Peer Review
Received July 06, 2024. Evaluated by 2 external peer reviewers, with direct editorial input from an Associate Editor and the Editor-in-Chief. Accepted in revised form June 10, 2025.
Footnotes
Complete author and article information provided before references.
JHI and MGS contributed equally to this work.
Figure S1: Flowchart of included and excluded SPRINT participants.
Table S1: Baseline Characteristics of Included and Excluded Participants.
Table S2: Association of Plasma with 30% eGFR Decline Modified by Adjustment or Exclusion of Individual Variables.
Supplementary Material
Figure S1; Tables S1-S2.
References
- 1.Jager K.J., Kovesdy C., Langham R., Rosenberg M., Jha V., Zoccali C. A single number for advocacy and communication-worldwide more than 850 million individuals have kidney diseases. Nephrol Dial Transplant. 2019;34(11):1803–1805. doi: 10.1093/ndt/gfz174. [DOI] [PubMed] [Google Scholar]
- 2.Astor B.C., Matsushita K., Gansevoort R.T., et al. Chronic Kidney Disease Prognosis Consortium Lower estimated glomerular filtration rate and higher albuminuria are associated with mortality and end-stage renal disease. A collaborative meta-analysis of kidney disease population cohorts. Kidney Int. 2011;79(12):1331–1340. doi: 10.1038/ki.2010.550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Grams M.E., Sang Y., Ballew S.H., et al. Evaluating glomerular filtration rate slope as a surrogate end point for ESKD in clinical trials: an individual participant meta-analysis of observational data. J Am Soc Nephrol. 2019;30(9):1746–1755. doi: 10.1681/ASN.2019010008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Schelling J.R. Tubular atrophy in the pathogenesis of chronic kidney disease progression. Pediatr Nephrol. 2016;31(5):693–706. doi: 10.1007/s00467-015-3169-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Plantinga L.C., Crews D.C., Coresh J., et al. CDC CKD Surveillance Team Prevalence of chronic kidney disease in US adults with undiagnosed diabetes or prediabetes. Clin J Am Soc Nephrol. 2010;5(4):673–682. doi: 10.2215/CJN.07891109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Devuyst O., Olinger E., Rampoldi L. Uromodulin: from physiology to rare and complex kidney disorders. Nat Rev Nephrol. 2017;13(9):525–544. doi: 10.1038/nrneph.2017.101. [DOI] [PubMed] [Google Scholar]
- 7.LaFavers K., Garimella P.S. Uromodulin: more than a marker for chronic kidney disease progression. Curr Opin Nephrol Hypertens. 2023;32(3):271–277. doi: 10.1097/MNH.0000000000000885. [DOI] [PubMed] [Google Scholar]
- 8.Köttgen A., Hwang S.J., Larson M.G., et al. Uromodulin Levels Associate with a Common UMOD Variant and Risk for Incident CKD. J Am Soc Nephrol. 2010;21(2):337. doi: 10.1681/ASN.2009070725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Youhanna S., Weber J., Beaujean V., Glaudemans B., Sobek J., Devuyst O. Determination of uromodulin in human urine: influence of storage and processing. Nephrol Dial Transplant. 2014;29(1):136–145. doi: 10.1093/ndt/gft345. [DOI] [PubMed] [Google Scholar]
- 10.Then C., Then H.L., Lechner A., et al. Serum uromodulin and decline of kidney function in older participants of the population-based KORA F4/FF4 study. Clin Kidney J. 2021;14(1):205–211. doi: 10.1093/ckj/sfaa032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Steubl D., Schneider M.P., Meiselbach H., et al. GCKD Study Investigators Association of serum uromodulin with death, cardiovascular events, and kidney failure in CKD. Clin J Am Soc Nephrol. 2020;15(5):616–624. doi: 10.2215/CJN.11780919. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Steubl D., Buzkova P., Garimella P.S., et al. Association of serum uromodulin With ESKD and kidney function decline in the elderly: the cardiovascular health study. Am J Kidney Dis. 2019;74(4):501–509. doi: 10.1053/j.ajkd.2019.02.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lv L., Wang J., Gao B., et al. Serum uromodulin and progression of kidney disease in patients with chronic kidney disease. J Transl Med. 2018;16(1):316. doi: 10.1186/s12967-018-1693-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Scherberich J.E., Gruber R., Nockher W.A., et al. Serum uromodulin-a marker of kidney function and renal parenchymal integrity. Nephrol Dial Transplant. 2018;33(2):284–295. doi: 10.1093/ndt/gfw422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Leiherer A., Muendlein A., Saely C.H., et al. The value of uromodulin as a new serum marker to predict decline in renal function. J Hypertens. 2018;36(1):110–118. doi: 10.1097/HJH.0000000000001527. [DOI] [PubMed] [Google Scholar]
- 16.Verdecchia P., Angeli F., Reboldi G. The SPRINT trial. J Am Soc Hypertens. 2015;9(10):750–753. doi: 10.1016/j.jash.2015.09.001. [DOI] [PubMed] [Google Scholar]
- 17.Ikeme J.C., Scherzer R., Garimella P.S., Estrella M.M., Ix J.H., Shlipak M. Plasma Uromodulin and Cardiovascular Outcomes in Adults with Hypertension and CKD: SA-PO529. J Am Soc Nephrol. 2023;34(11S):870–871. [Google Scholar]
- 18.Ikeme J.C., Scherzer R., Garimella P.S., et al. The association of plasma and urine uromodulin with cardiovascular disease in persons with hypertension and CKD. Am J Kidney Dis. 2024;84(6):799–802. doi: 10.1053/j.ajkd.2024.05.012. [DOI] [PubMed] [Google Scholar]
- 19.SPRINT Research Group, Wright J.T., Jr., Williamson J.D., et al. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2015;373(22):2103–2116. doi: 10.1056/NEJMoa1511939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Inker L.A., Schmid C.H., Tighiouart H., et al. CKD-EPI Investigators Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med. 2012;367(1):20–29. doi: 10.1056/NEJMoa1114248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.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]
- 22.Fine J.P., Gray R.J. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94(446):496–509. [Google Scholar]
- 23.Serafini-Cessi F., Malagolini N., Cavallone D. Tamm-Horsfall glycoprotein: biology and clinical relevance. Am J Kidney Dis. 2003;42(4):658–676. doi: 10.1016/s0272-6386(03)00829-1. [DOI] [PubMed] [Google Scholar]
- 24.LaFavers K.A., Gaddy A.R., Micanovic R., et al. Water loading and uromodulin secretion in healthy individuals and idiopathic calcium stone formers. Clin J Am Soc Nephrol. 2023;18(8):1059–1067. doi: 10.2215/CJN.0000000000000202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bächle H., Sekula P., Schlosser P., et al. GCKD investigators Uromodulin and its association with urinary metabolites: the German Chronic Kidney Disease Study. Nephrol Dial Transplant. 2023;38(1):70–79. doi: 10.1093/ndt/gfac187. [DOI] [PubMed] [Google Scholar]
- 26.Tachibana S., Iyoda M., Suzuki T., et al. Serum uromodulin is associated with the severity of clinicopathological findings in ANCA-associated glomerulonephritis. PLoS One. 2019;14(11) doi: 10.1371/journal.pone.0224690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Jotwani V., Garimella P.S., Katz R., et al. SPRINT Research Group Tubular biomarkers and chronic kidney disease progression in SPRINT participants. Am J Nephrol. 2020;51(10):797–805. doi: 10.1159/000509978. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Peach R.J., Day W.A., Ellingsen P.J., McGiven A.R. Ultrastructural localization of Tamm-Horsfall protein in human kidney using immunogold electron microscopy. Histochem J. 1988;20(3):156–164. doi: 10.1007/BF01746679. [DOI] [PubMed] [Google Scholar]
- 29.Avis P.J. The development of a radioimmunoassay procedure for the estimation of Tamm-Horsfall glycoprotein in human serum. Clin Sci Mol Med. 1977;52(2):183–191. doi: 10.1042/cs0520183. [DOI] [PubMed] [Google Scholar]
- 30.Muchmore A.V., Decker J.M. Uromodulin: a unique 85-kilodalton immunosuppressive glycoprotein isolated from urine of pregnant women. Science. 1985;229(4712):479–481. doi: 10.1126/science.2409603. [DOI] [PubMed] [Google Scholar]
- 31.El-Achkar T.M., Wu X.R., Rauchman M., McCracken R., Kiefer S., Dagher P.C. Tamm-Horsfall protein protects the kidney from ischemic injury by decreasing inflammation and altering TLR4 expression. Am J Physiol Renal Physiol. 2008;295(2):F534–F544. doi: 10.1152/ajprenal.00083.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Micanovic R., LaFavers K.A., Patidar K.R., et al. The kidney releases a nonpolymerizing form of uromodulin in the urine and circulation that retains the external hydrophobic patch domain. Am J Physiol Renal Physiol. 2022;322(4):F403–F418. doi: 10.1152/ajprenal.00322.2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Nanamatsu A., Mori T., Ando F., et al. Vasopressin induces urinary uromodulin secretion by activating PKA (Protein Kinase A) Hypertension. 2021;77(6):1953–1963. doi: 10.1161/HYPERTENSIONAHA.121.17127. [DOI] [PubMed] [Google Scholar]
- 34.El-Achkar T.M., McCracken R., Liu Y., et al. Tamm-Horsfall protein translocates to the basolateral domain of thick ascending limbs, interstitium, and circulation during recovery from acute kidney injury. Am J Physiol Renal Physiol. 2013;304(8):F1066–F1075. doi: 10.1152/ajprenal.00543.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Garimella P.S., Biggs M.L., Katz R., et al. Urinary uromodulin, kidney function, and cardiovascular disease in elderly adults. Kidney Int. 2015;88(5):1126–1134. doi: 10.1038/ki.2015.192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Steubl D., Block M., Herbst V., et al. Urinary uromodulin independently predicts end-stage renal disease and rapid kidney function decline in a cohort of chronic kidney disease patients. Medicine (Baltimore) 2019;98(21) doi: 10.1097/MD.0000000000015808. [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; Tables S1-S2.