Estimating the extent to which chronic kidney disease mediates the association between metabolic syndrome and heart failure: the Atherosclerosis Risk in Communities (ARIC) Study

Abstract Background/Introduction Metabolic Syndrome (MetS) is linked to increased risk of developing heart failure (HF), underscoring a need to implement more aggressive lifestyle and pharmacotherapeutic HF prevention strategies. Chronic kidney disease (CKD) is associated with MetS and is a risk factor for HF; however, therapies protective against both kidney function decline and HF risk such as ACEi/ARBs and SGLT2i are underutilized among individuals with MetS. A potential contributor to this underutilization is that creatinine (Cr)- based estimates of glomerular filtration rate (eGFR) may underestimate the role of CKD as a mediator of HF risk relative to Cystatin-C (CysC) based eGFR. Purpose To compare estimations of the relationship between MetS and kidney function, and of the extent to which eGFR mediates the link between MetS and incident HF, using CysC-based versus Cr-based eGFR. Methods Among ARIC Visit 4 (1996-1998) participants without baseline HF, we used linear regression to assess the cross-sectional association of MetS severity (measured using an established z-score, modeled in quartiles and continuously) with eGFR, with comparisons across 3 different eGFR equations (CKD-Cr 2021, CKD-Cr-CysC 2021 and CKD-CysC 2012). In prospective analyses, we used Cox regression and mediation analyses to estimate the extent to which kidney function mediated the association between MetS and incident HF after Visit 4 through 12/31/2019, with comparisons across the 3 eGFR equations. Results Among 8,582 participants (mean age 63.1 years, 56.9% women, 18.4% Black adults), 1,803 developed HF over a median follow-up of 21 years. MetS was more strongly associated with eGFR when using eGFR estimates that included CysC, with the 4th vs 1st quartile of MetS Z-score associated with an adjusted eGFR difference of -2.55 ml/min/1.73m2 (95% CI -3.36 to -1.73) for CKD-Cr, -8.07 ml/min/1.73m2 (95% CI -8.95 to -7.19) for CKD-Cr-CysC and -10.74 ml/min/1.73m2 (95% CI -11.67 to -9.80) for CKD-CysC. In prospective analyses, we also estimated greater mediation by kidney function of the association between MetS and incident HF when using eGFR equations that included CysC. The strength of the association between MetS and HF potentially mediated by kidney function increased from 3.0% to 15.2% to 24.2%, respectively, when using eGFR-Cr vs eGFR-Cr-CysC vs eGFR-CysC (Table and Figure). These estimates were not significantly attenuated after additional adjustment for coronary heart disease and C-reactive protein. Conclusion(s) The degree of kidney function reduction associated with MetS, and the subsequent implications of kidney dysfunction for HF risk in those with MetS, are underestimated using eGFR-Cr equations compared to eGFR equations including CysC. Greater appreciation by clinicians of the impact of kidney disease on HF risk among individuals with MetS could help increase utilization of kidney-protective therapies that can lower HF risk.