Background Maintenance hemodialysis patients experience higher morbidity and mortality from COVID-19, partly due to comorbidities like diabetes and cardiovascular disease. However, kidney disease-related metabolic processes may also contribute. Methods In this prospective, multi-center, observational study, we analyzed 201 routine serum samples from 30 hemodialysis patients (average age 59.2 ± 13.3 years, 57% male) with confirmed COVID-19, collected from 60 days before and 60 days after diagnosis. Untargeted liquid chromatography/mass spectrometry was used to profile metabolites. Linear and semi-parametric mixed-effects models were applied to assess changes across four phases: baseline (−60 to −15 days), putative incubation period (PIP; −14–0 days), acute (1–14 days), and post-COVID (15–60 days). Because infection and symptoms may vary across individuals, −14–0 days were used as an approximate pre-diagnosis window rather than a precise incubation interval. Results Among 417 metabolomic features, 10 showed significant changes between baseline and PIP. Two metabolites, α-guanidinoglutaric acid and N-acetylneuraminic acid, were identified through library matching, while the remainder were characterized by mass and retention time. Temporal analysis revealed both transient metabolic shifts, which returned to baseline, and persistent changes, which remained altered post-COVID. Conclusions These findings suggest that early metabolic changes before COVID-19 diagnosis may be detected in routine serum samples, offering opportunities to develop predictive models for early detection. Identifying these unique metabolomics fingerprints could improve personalized surveillance strategies and enhance understanding of COVID-19’s impact on hemodialysis patients.
Dias et al. (Fri,) studied this question.