Plasma resistin levels are associated with acute kidney injury in hospitalized COVID-19 patients

ABSTRACT

Rationale Obesity is a strong risk factor for acute kidney injury (AKI) in patients with COVID-19, but underlying mechanisms are unknown. Resistin is an immunomodulatory adipokine with elevated circulating levels in obese outpatients that could contribute to inflammatory kidney injury. We hypothesized that plasma resistin levels would be associated with AKI and BMI, and correlated with the inflammatory markers IL6 and MCP1 in hospitalized COVID-19 patients. Methods: We conducted a prospective cohort study of 134 patients admitted to the Hospital of the University of Pennsylvania with a primary diagnosis of COVID-19. Plasma samples were collected within 48 hours of admission and analyzed using the Olink Proximity Extension Assay, with biomarker levels expressed using normalized protein expression (NPX) values relative to common pooled control plasma. We tested the association of each biomarker with AKI, defined by Kidney Disease Improving Global Outcomes creatinine and dialysis criteria, using the Wilcoxon rank-sum test as well as multivariable logistic regression to adjust for confounders. Spearman's rho and correlation coefficients were calculated for the correlation of biomarker levels with each other. We used causal mediation models to investigate effects of BMI on AKI mediated by plasma resistin. Results: Of 134 patients enrolled, 43 (32.1%) developed AKI 25 with stage 1, 5 with stage 2, and 13 with stage 3. Plasma resistin levels ranged from 5.26-13.01 NPX units and were strongly associated with AKI odds ratio 2.13 (95% CI 1.43-3.17) per NPX unit. This association was diminished but remained significant after adjustment for age and APACHE III score (OR 1.69 (1.09-2.63)). Body mass index was higher in patients with AKI than without (median 31.4 (IQR 27.1-37.6) kg/m2 v. 28.3 (25.1-34.9) kg/m2, respectively), but the difference was not statistically significant (p=0.082). There was no significant correlation of BMI with resistin levels (rho 0.05, p=0.562), and causal mediation models failed to detect significant mediation of BMI-AKI association through resistin. Plasma IL6 and MCP1 were associated with AKI (p=0.044 and p=0.003, respectively) and correlated with resistin levels (rho=0.32, p<0.001 and rho=0.40, p<0.001, respectively). Conclusion: In patients hospitalized with COVID-19, plasma levels of the adipokine resistin were strongly associated with the development of AKI, and correlated with circulating inflammatory markers IL6 and MCP1. We did not detect a mediation effect of the obesity-AKI association by plasma resistin but had limited sample size to adequately power this analysis.