Abstract
<b><i>Introduction:</i></b> Metabolic acidosis is associated with chronic kidney disease (CKD) progression and mortality, but the association of race/ethnicity with incident metabolic acidosis and/or its adverse outcomes in patients with CKD is unknown. <b><i>Methods:</i></b> We used deidentified medical records data (2007–2019) to generate a cohort of 136,067 patients with nondialysis-dependent CKD stages 3–5 and ≥2 years’ postindex data or death within 2 years. We grouped participants into those with and without metabolic acidosis (serum bicarbonate 12 to <22 mEq/L vs. 22 to <30 mEq/L) as Asian, Black, Hispanic, non-Hispanic White individuals, or unknown. Cox proportional hazards models examined factors associated with (1) incident metabolic acidosis; and (2) time to the composite outcome of death, dialysis, transplant, or ≥40% decline from baseline eGFR (DD40) within each race/ethnic group. <b><i>Results:</i></b> Metabolic acidosis incidence was higher for Asian, Black, and Hispanic versus non-Hispanic White individuals (<i>p</i> values for adjusted hazard ratios [HR] all <0.001), but this higher hazard was mitigated in all groups with increasing community education. During the median follow-up of 4.2 years, 47,032 of 136,607 (34.6%) experienced a DD40 event. There was an independent association of metabolic acidosis with DD40 within each race/ethnic group. Adjusted HRs (95% confidence interval) for DD40 were 1.806 (1.312, 2.486), 1.420 (1.313, 1.536), 1.409 (1.211, 1.641), and 1.561 (1.498, 1.626) in Asian, Black, Hispanic, and non-Hispanic White groups, respectively (all <i>p</i> < 0.0001), for metabolic acidosis versus normal serum bicarbonate. <b><i>Discussion/Conclusion:</i></b> The higher incidence of metabolic acidosis observed in Asian, Black, and Hispanic individuals was mitigated by residing in higher education zip codes. Once established, metabolic acidosis was independently associated with DD40 in patients with CKD in all racial/ethnic groups examined.