Nutrition-Related Parameters and Imaging Severity of COVID-19 in Hospitalized Adults: A Secondary Analysis of a Prospective Multicenter Cohort.

Background/Objectives: Nutrition-related parameters may be associated with COVID-19 severity, but their relationships with complementary pulmonary imaging measures remain incompletely characterized. This study aimed to examine associations between selected nutrition-related parameters, pulmonary imaging severity, and clinical outcomes in hospitalized adults with COVID-19. Methods: This secondary exploratory analysis included 555 adults from a prospective multicenter cohort. Dietary intake assessed during the first week of hospitalization and derived from medical history, serum albumin and total protein concentrations, and selected bioimpedance parameters were examined in relation to computed tomography severity score (CTSS), lung ultrasonography (LUS) parameters, and selected clinical outcomes. Results: First-week hospital dietary intake was weakly inversely associated with CTSS (ρ = -0.110; p = 0.038), LUSSB (ρ = -0.274; p = 0.002), and LUSSBmax (ρ = -0.329; p < 0.001). History-derived dietary intake was associated with LUSSB (ρ = -0.253; p = 0.007) and LUSSBmax (ρ = -0.239; p = 0.011), but not with CTSS (ρ = -0.091; p = 0.120). Lower serum albumin was weakly associated with higher CTSS (ρ = -0.252; p = 0.008), whereas total protein was not (ρ = -0.199; p = 0.300). In unadjusted analyses, first-week hospital dietary intake, albumin, and total protein were associated with selected escalation-of-care outcomes, but not with in-hospital death. Nominal BIA-LUSSC associations were observed, but none remained statistically significant after correction for multiple testing. Conclusions: Selected nutrition-related parameters showed weak and non-uniform associations with pulmonary imaging severity in hospitalized adults with COVID-19. The findings do not establish independent prognostic relationships, and the exploratory bioimpedance results require confirmation in larger, appropriately designed studies.