Abstract Background Chronological age (CA) is a predictor of adverse coronavirus disease 2019 (COVID-19) outcomes; however, CA alone does not capture individual responses to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. Here, we evaluated the influence of aging metrics PhenoAge and PhenoAgeAccel to predict adverse COVID-19 outcomes. Furthermore, we sought to model adaptive metabolic and inflammatory responses to severe SARS-CoV-2 infection using individual PhenoAge components. Method In this retrospective cohort study, we assessed cases admitted to a COVID-19 reference center in Mexico City. PhenoAge and PhenoAgeAccel were estimated using laboratory values at admission. Cox proportional hazards models were fitted to estimate risk for COVID-19 lethality and adverse outcomes (intensive care unit admission, intubation, or death). To explore reproducible patterns which model adaptive responses to SARS-CoV-2 infection, we used k-means clustering using PhenoAge components. Results We included 1068 subjects of whom 222 presented critical illness and 218 died. PhenoAge was a better predictor of adverse outcomes and lethality compared to CA and SpO2 and its predictive capacity was sustained for all age groups. Patients with responses associated to PhenoAgeAccel >0 had higher risk of death and critical illness compared to those with lower values (log-rank p < .001). Using unsupervised clustering, we identified 4 adaptive responses to SARS-CoV-2 infection: (i) inflammaging associated with CA, (ii) metabolic dysfunction associated with cardiometabolic comorbidities, (iii) unfavorable hematological response, and (iv) response associated with favorable outcomes. Conclusions Adaptive responses related to accelerated aging metrics are linked to adverse COVID-19 outcomes and have unique and distinguishable features. PhenoAge is a better predictor of adverse outcomes compared to CA.