Endothelial Dysfunction Biomarkers and CKD Incidence in the REGARDS Cohort

Abstract

Introduction: Chronic kidney disease (CKD) is only partly caused by traditional risk factors. Endothelial dysfunction is common in CKD and may contribute to CKD incidence. We studied the association of circulating biomarkers reflecting endothelial dysfunction with incident CKD.

Methods: The Reasons for Geographical and Racial Differences in Stroke (REGARDS) study is a prospective cohort of 30,239 Black or White adults aged ≥45 years. Baseline levels of intercellular cellular adhesion molecule 1 (ICAM-1), vascular cellular adhesion molecule 1 (VCAM-1), factor VIII (FVIII), and E-selectin were measured in 3300 participants without baseline CKD or albuminuria who attended a second visit 9.4 years later. Kidney outcomes were incident CKD (estimated glomerular filtration rate [eGFR] <60 ml/min per 1.73 m² and ≥40% decline or onset of new end-stage kidney disease), incident ≥30% eGFR decline, and incident albuminuria (albumin-to-creatinine ratio [ACR] ≥30 mg/g). Sequentially adjusted logistic regression models assessed the association of biomarkers with kidney outcomes.

Results: Median age of participants was 62 years, 49% were women, and 46% identified as Black. Of the participants, 228 (6.9%) developed CKD, 613 (18.9%) experienced ≥30% decline in eGFR, and 356 (11.4%) developed albuminuria. The adjusted odds ratios (ORs) for incident CKD per 1 SD increment biomarker was 1.12 for ICAM-1 (95% confidence interval [CI]: 1.02-1.22), 1.10 for VCAM-1 (95% CI: 1.01-1.20), 1.15 for FVIII (95% CI: 1.06-1.24), and 1.10 for E-selectin (95% CI: 1.01-1.20). Results were similar for incident ≥30% eGFR decline but not albuminuria, where only higher FVIII was positively associated.

Conclusion: Higher concentration of ICAM-1, VCAM-1, FVIII, and E-selectin were associated with incident CKD and ≥30% eGFR decline in a large cohort study. Higher FVIII was also associated with incident albuminuria.

Keywords: albuminuria; biomarkers; chronic kidney disease; endothelial dysfunction; risk factors.

Graphical abstract

Figure 1

Odds ratios and 95% CIs for incident CKD per 1 SD increment of biomarkers of endothelial dysfunction. ^aModel 1 was unadjusted. ^bModel 2 adjusted for age, sex, race, and geographic region of residence. ^cModel 3 was further adjusted for body mass index, hypertension, diabetes, coronary artery disease, smoking status, and dyslipidemia. ^dModel 4 was further adjusted for baseline eGFR and ACR. 1 log-transformed SD was, as follows: ICAM-1, 0.25 ln(ng/ml); VCAM-1, 0.36 ln(ng/ml); factor VIII, 0.32 ln (%); E-selectin, 0.46 ln(ng/ml). Regression models were weighted to reflect the entire cohort from which this subcohort was sampled. ACR, urinary albumin-to-creatinine ratio; CI, confidence interval; CKD, chronic kidney disease defined by eGFR <60 ml/min per 1.73 m² plus 40% decline from baseline or new kidney failure; eGFR, estimated glomerular filtration rate by 2021 CKD-EPI equation by serum creatinine and cystatin C; ICAM-1, intercellular cellular adhesion molecule 1.

Figure 2

Odds ratios and 95% CIs for incident ≥ 30% decline in eGFR per 1 SD increment of biomarkers of endothelial dysfunction. ^aModel 1 was unadjusted. ^bModel 2 adjusted for age, sex, race, and geographic region of residence. ^cModel 3 was further adjusted for body mass index, hypertension, diabetes, coronary artery disease, smoking status, and dyslipidemia. ^dModel 4 was further adjusted for baseline eGFR and ACR. 1 log-transformed SD was, as follows: ICAM-1, 0.25 ln(ng/ml); VCAM-1, 0.36 ln(ng/ml); factor VIII, 0.32 ln(%); E-selectin, 0.46 ln(ng/ml). Regression models were weighted to reflect the entire cohort from which this subcohort was sampled. ACR, urinary albumin-to-creatinine ratio; CI, confidence interval; eGFR, estimated glomerular filtration rate by 2021 CKD-EPI equation by serum creatinine and cystatin C; ICAM-1, intercellular cellular adhesion molecule 1.

Figure 3

Odds ratios and 95% CIs for incident albuminuria per 1 SD increment of biomarkers of endothelial dysfunction. Model 1 was unadjusted. ^aModel adjusted for age, sex, race, and geographic region of residence. ^bModel 3 was further adjusted for body mass index, hypertension, diabetes, coronary artery disease, smoking status, and dyslipidemia. ^cModel 4 was further adjusted for baseline eGFR and ACR. Regression models were weighted to reflect the entire cohort from which this subcohort was sampled. Albuminuria was defined as ACR ≥30 mg/g. 1 log-transformed SD was, as follows: ICAM-1, 0.25 ln(ng/ml); VCAM-1, 0.36 ln(ng/ml); factor VIII, 0.32 ln (%); E-selectin, 0.46 ln(ng/ml). ACR, urinary albumin-to-creatinine ratio; CI, confidence interval; eGFR, estimated glomerular filtration rate by 2021 CKD-EPI equation by serum creatinine and cystatin C; ICAM-1, intercellular cellular adhesion molecule 1.