Biomarkers of kidney tubule injury and dysfunction and risk of incident hypertension in community-living individuals: results from the multi-ethnic study of atherosclerosis

Hypertension (HTN) is the leading cause of premature death worldwide, affecting >1 billion people, and is a major risk factor for the onset and progression of chronic kidney disease (CKD) and cardiovascular disease (CVD) [1, 2]. While mechanisms leading to essential HTN remain elusive, many have hypothesized that abnormal salt and water regulation by the kidney tubules may be contributory [3, 4]. Contemporary clinical tests evaluate and assess kidney glomerular function but not tubule function.

Kidney damage and tubulointerstitial injury may lead to neurohormonal activation, inflammation and impaired sodium and potassium handling, which may contribute to the development of HTN through mechanisms independent of estimated glomerular filtration (eGFR) [3]. Recent investigations have identified several urinary biomarkers of kidney tubule injury and dysfunction and multiple prior studies have demonstrated that these biomarkers are associated with the loss of kidney function in ambulatory populations, independent of eGFR or albuminuria [5, 6]. However, few studies have related tubule function biomarkers to the risk of developing HTN [7].

The Multi-Ethnic Study of Atherosclerosis (MESA) is a prospective multicenter cohort of community-dwelling adults without prevalent cardiovascular disease (CVD) that was designed to understand risk factors for subclinical CVD [8]. Using the MESA cohort, we examined the associations between plasma and urinary kidney tubule biomarkers with incident HTN. We took a random sample of 500 MESA participants who did not have prevalent CKD or diabetes to minimize the confounding effects of glomerular kidney damage in this study. Among these, 463 participants had available plasma and urine specimens. An additional 166 participants had prevalent HTN at baseline and were excluded. Thus the final sample for this study included 297 participants (Supplementary Fig. 1). We measured five urinary biomarkers of kidney tubule health: alpha-1-microglobulin (A1M), epidermal growth factor (EGF), kidney injury molecule-1 (KIM-1), monocyte chemoattractant protein-1 (MCP-1) and anti-chitinase-3-like protein 1 (YKL-40). We compared associations to that of albumin:creatinine ratio, a clinical marker of glomerular injury, to provide a measure of strengths of associations for comparison with the tubule biomarkers. We also measured six plasma biomarkers of kidney tubule health: KIM-1, MCP-1, soluble urokinase-type plasminogen activator receptor (suPAR), tumor necrosis factor receptor 1 (TNFR1), tumor necrosis factor receptor 2 (TNFR2) and YKL-40 at baseline. These biomarkers were selected because they reflect aspects of kidney tubular biology including tubular injury, inflammation and repair [9, 10]. The plasma biomarkers were measured using a multiplex assay (Meso Scale Diagnostics, Rockville, MD, USA). The urine biomarkers were measured using a multiplex assay (Luminex, Austin, TX, USA) except for A1M, which was measured by nephelometry. All measurements were performed in duplicates except A1M and the mean values for each biomarker were used in the analysis to improve precision. The laboratory personnel who performed biomarker measurements were blinded to the clinical information.

The primary outcome of the study was the development of incident HTN, occurring anytime between the baseline visit and the year 15 follow-up visit. Blood pressure (BP) was measured at each MESA follow-up visit (visits 2, 3, 4, 5 and 6) using a standardized protocol [8]. Incident HTN was defined as a systolic BP (SBP) ≥140 mmHg, diastolic BP ≥90 mmHg or the use of any antihypertensive medication during the follow-up examinations [7]. Multivariable Poisson regression and linear mixed models were used to examine associations between baseline tubule markers and incident HTN and longitudinal relative change in SBP, respectively. We evaluated each plasma and urinary biomarker as a continuous variable [after logarithmic (base 2) transformation] and across quartiles, with the lowest quartile set as the referent. Model 1 was adjusted for age, sex, race/ethnicity, education, urine creatinine concentration (to account for urine tonicity), body mass index, baseline SBP, smoking status, low-density lipoprotein (LDL), high-density lipoprotein (HDL), triglyceride, baseline eGFR and urine albumin.

Among the 297 participants, the mean age was 58 ± 10 years, 53% were women, 47% were White, 20% were Black and the mean baseline eGFR was 94 ± 14 ml/min/1.73 m² (Supplementary Table 1). The median number of BP measurements per participant during follow-up was 6 with an interquartile range of 2–10. Nearly half (n = 142) had incident HTN over 15 years of follow-up. In unadjusted models, higher urinary A1M, KIM-1, MCP-1 and albumin, as well as higher plasma MCP-1 and YKL-40, were individually associated with the risk of incident HTN (Table 1). However, in the fully adjusted model, only urine albumin remained associated with incident HTN (incident rate ratio [IRR] 1.45 [95% confidence interval (CI) 1.12–1.88], P < .005). These findings were similar in quartile-based analyses, although the CIs were wide (Supplementary Tables 2 and 3). In linear mixed models, we observed no association of plasma or urinary kidney tubule markers or of urine albumin with longitudinal changes in SBP in fully adjusted models (Table 2).

Table 1:

Association of biomarkers of kidney tubule injury and dysfunction with incident HTN in community-living individuals without diabetes or CKD.

Biomarkers	Unadjusted IRR (95% CI)	Model 1 IRR (95% CI)
P KIM-1	1.16 (0.97 to 1.38)	1.07 (0.86 to 1.33)
P MCP-1	1.46 (1.15 to 1.86)*	1.31 (0.93 to 1.86)
P suPAR	1.25 (0.92 to 1.69)	1.36 (0.94 to 1.97)
P YKL-40	1.28 (1.10 to 1.50)*	1.03 (0.87 to 1.21)
P TNFR1	1.11 (0.83 to 1.49)	1.04 (0.76 to 1.43)
P TNFR2	1.16 (0.79 to 1.70)	1.03 (0.71 to 1.49)
U KIM-1	1.29 (1.12 to 1.48)**	1.04 (0.83 to 1.28)
U MCP-1	1.38 (1.20 to 1.58)**	0.92 (0.70 to 1.20)
U EGF	1.19 (0.99 to 1.41)	0.92 (0.71 to 1.20)
U YKL-40	1.10 (0.96 to 1.27)	1.01 (0.90 to 1.13)
U A1M	1.29 (1.01 to 1.65)*	0.82 (0.62 to 1.09)
U albumin	1.40 (1.26 to 1.56)**	1.25 (1.08 to 1.45)*

^* P < .005, **P < .001.

Subcohort: total n = 297 and incident HTN = 142. Model 1: age, sex, race/ethnicity, education, urine creatinine, BMI, SBP, smoking, LDL, HDL, triglyceride, urine albumin and eGFR.

BMI, body mass index; P, plasma; U, urine.

Table 2:

Association of (A) blood and (B) urine biomarkers of kidney tubule injury and dysfunction with longitudinal relative change in SBP in community-living individuals without diabetes or CKD.

Biomarkers per 2-fold higher biomarker	Unadjusted, β (95% CI)	Model 1, β (95% CI)
P KIM-1	−0.01 (−0.11 to 0.12)	0.02 (−0.10 to 0.14)
P MCP-1	−0.08 (−0.26 to 0.11)	−0.06 (−0.26 to 0.13)
P suPAR	0.10 (−0.09 to 0.30)	0.04 (−0.21 to 0.28)
P YKL-40	−0.08 (−0.18 to 0.01)	−0.06 (−0.17 to 0.06)
P TNFR-1	0.03 (−0.15 to 0.21)	0.01 (−0.21 to 0.23)
P TNFR-2	0.13 (−0.11 to 0.37)	0.08 (−0.19 to 0.35)
U KIM-1	−0.02 (−0.10 to 0.07)	−0.06 (−0.20 to 0.08)
U MCP-1	−0.02 (−0.11 to 0.06)	0.07 (−0.09 to 0.23)
U EGF	0.02 (−0.09 to 0.14)	−0.07 (−0.28 to 0.14)
U YKL-40	−0.06 (−0.13 to 0.01)	−0.07 (−0.15 to 0.01)
U A1M	0.10 (−0.08 to 0.27)	0.08 (−0.12 to 0.28)
U albumin	−0.02 (−0.09 to 0.05)	0 (−0.10 to 0.11)

Model 1: age, sex, race/ethnicity, education, urine creatinine, BMI, SBP, smoking, LDL, HDL, triglyceride, urine albumin and eGFR.

β represents the percent change in SBP/year.

BMI, body mass index; P, plasma; U, urine.

The strengths of this study include the availability of multiple plasma and urinary biomarkers of kidney tubule injury and function and long (15 years) follow-up for incident HTN by standardized protocols within the MESA cohort. The primary study limitation is its relatively small sample size (n = 297) and exclusion of persons with CVD, CKD or diabetes. Results may differ in these populations.

In summary, in this study of community-living individuals without CKD, diabetes or CVD there was no association of selected kidney tubule injury or dysfunction biomarkers with incident HTN. In comparison, urine albumin was independently associated with incident HTN. Albuminuria primarily marks glomerular injury and was included for comparison with the kidney tubule biomarkers. Thus associations of the broad panel of kidney tubule biomarkers assessed here are likely to have weak or no association with incident HTN among healthy community-living individuals. Future studies with larger samples and assessing other populations will be needed to confirm and extend these findings.

AUTHORS’ CONTRIBUTIONS

All authors were involved in drafting and critically revising the report. All authors had access to study results and the first author assumes responsibility for the integrity and accuracy of the data reported. All authors reviewed and approved the final submitted version of the article.

CONFLICT OF INTEREST STATEMENT

M.J.S. serves on the Steering Committee for a trial funded by Akebia and served as a consultant for Cardurian. M.G.S. receives research funding from Bayer; has received honoraria from Bayer, Boehringer Ingelheim and AstraZeneca; consulted for Cricket Health and Intercept Pharmaceuticals and previously served as an advisor to and held stock in TAI Diagnostics. J.H.I. holds an investigator-initiated research grant from Baxter International for an unrelated project. O.M.G. has received grant funding and honoraria from Akebia and Amgen; grant funding from GlaxoSmithKline; honoraria from AstraZeneca, Reata and Ardelyx and serves on a data monitoring committee for QED Therapeutics. The remaining authors declare that they have no relevant financial interests. The results presented in this article have not been published previously in whole or part, except in abstract format.