Dysregulated Metabolic Homeostasis as a Unifying Death Mechanism Underlying the Diverse Clinical Manifestations of Covid-19: Insights from a Retrospective Analysis of Sequential Blood Variables

Abstract

Background. COVID-19 presents diverse clinical manifestations associated with increased mortality, yet a unifying death mechanism remains elusive; here, we suggest such a mechanism that implies a simple way to lower deaths. This work differs from previous studies that use machine learning to identify mortality predictors. Methods. Viewing clinical deterioration to a severe stage as a distinct "junction" in disease progression, we collected 173 medical records of COVID-19 patients who deteriorated and divided them into two groups: those who died (nonsurvivors) and those who recovered after deterioration (survivors). We aligned patients' medical records by clinical deterioration time and statistically compared the two groups using standard blood variables. Results. Key differences between the groups emerged only in the first week after clinical deterioration: nonsurvivors showed a rapid, simultaneous rise in lactate dehydrogenase (p≤0.0001) and D-dimer (p≤0.0001), followed by a decrease in platelet counts in the second week (p≤0.0001). Other variables remained consistent throughout hospitalization. Older patients showed similar but less significant response patterns. Based on these clinical results, we hypothesized that the mechanism of death in COVID-19 involves an abrupt glycolytic surge during deterioration, driven by concurrent hypoxemia and virus-induced mitochondriopathy, resulting in significant disruption of metabolic homeostasis, which leads to imbalanced hemostasis and death. Conclusions. Our findings highlight the importance of timing in COVID-19 treatment. Using an available machine learning algorithm to predict imminent deterioration enables prompt, short-term intervention with prophylactic mechanical ventilation and optimal antiglycolytic therapy. Implementing this approach requires further experimental and clinical validation. Identifying metabolism-related genetic or epigenetic anomalies in nonsurvivors will support our hypothesis and aid in classifying the high-risk patients.