Body composition, physical capacity, and immuno-metabolic profile in community-acquired pneumonia caused by COVID-19, influenza, and bacteria: a prospective cohort study

Community-acquired pneumonia (CAP) is a leading cause of hospitalization and death from infectious diseases and is caused by various bacteria and viruses. The emergence of SARS-CoV-2 underscored the importance of viral causes of CAP. Risk factors such as age, male sex, obesity, diabetes, hypertension, and cardiovascular disease are associated with severe outcomes in COVID-19, and similar factors are implicated in severe bacterial and influenza-related CAP. Obesity is linked to physical inactivity, dyslipidemia, insulin resistance, chronic low-grade inflammation, and lower adiponectin, which may impair immune responses and increase susceptibility to severe infections. Dysregulated or exaggerated cytokine responses can drive immunopathology, as documented in COVID-19. The study’s purpose was to compare body composition, physical capacity, glucometabolic status, and inflammatory profiles among patients hospitalized with CAP due to SARS-CoV-2, influenza virus, or bacteria.

Prior studies identified obesity and metabolic disorders as risk factors for severe outcomes in COVID-19, influenza, and bacterial CAP. Obesity’s chronic low-grade inflammation and reduced adiponectin can impair immune responses, increasing infection severity. COVID-19 has been associated with abnormal cytokine profiles and more severe outcomes than influenza or bacterial CAP in some reports. Hyperglycemia at admission and dyslipidemia (notably low HDL-cholesterol) are common in CAP and linked to worse outcomes. However, before this study, no direct comparisons existed across CAP etiologies for body composition, physical capacity, glucose and lipid metabolism, and systemic inflammatory responses.

Design and setting: Prospective cohort study at Nordsjællands Hospital, Denmark. Recruitment from January 8, 2019 to May 26, 2020. Population: Adults ≥18 years with CAP defined by a new infiltrate on chest X-ray/CT plus at least one symptom (fever ≥38.0 °C, cough, pleuritic chest pain, dyspnea, or focal auscultatory findings). Exclusion: No pathogen detected in blood, airways, or urine (Legionella pneumophila antigen or Streptococcus pneumoniae antigen). Outcomes included ICU admission, hospital length of stay, in-hospital, 30-day, and 180-day mortality. Data collection: Demographics, comorbidities, symptoms, medical history. Severity assessed using CURB-65 (mild 0–1, moderate 2, severe 3–5). Comorbidity burden by Charlson Comorbidity Index and categorized as none, one, or more than one condition. Microbiological testing recorded. Blood sampling and assays: Within 48 h of admission, collected venous blood for complete blood count, CRP, coagulation, procalcitonin, kidney/liver function, diabetes biomarkers, and lipid status. Cytokines measured from plasma using V-PLEX Viral Panel 3 (IL-5, IFN-γ, IL-1β, IL-4, IL-6, IL-8, IL-10, IL-12p70, TNF-α) and U-PLEX IL-18 assay; adiponectin measured by Human Adiponectin Kit. Cytokines run in duplicates with averages reported. Definitions: Known diabetes from records and/or antidiabetic medication. Undiagnosed diabetes: HbA1c ≥48 mmol/mol (≥6.5%). Pre-diabetes: HbA1c 39–47 mmol/mol (5.7–6.4%). Admission glycemia categorized: normoglycemia <5.99 mmol/L, mild hyperglycemia 6.0–10.99 mmol/L, severe hyperglycemia ≥11.0 mmol/L. Questionnaires within 48 h: IPAQ Short Form for physical activity (categorized as low or moderate-to-high), EQ-5D-5L for health-related quality of life, Barthel Index-100 for activities of daily living, FRAIL scale for frailty, Nutritional Risk Screening-2002 for nutritional status. Anthropometry and body composition within 48 h: Weight by electronic scale; height self-reported; BMI calculated (kg/m²). Overweight BMI ≥25 kg/m²; obesity BMI ≥30 kg/m². Waist circumference measured; body composition (FM and FFM) via bioelectrical impedance analysis (BioScan touch i8). Handgrip strength: Measured with electronic dynamometer (Saehan) seated, elbow 90°, dominant hand; best of three attempts recorded. Statistical analysis: Normality by histograms, Q–Q plots, Shapiro–Wilk; variance by Levene’s test. Missingness assumed MCAR. Parametric data as mean (SD) with one-way ANOVA and Tukey post hoc; non-parametric as median (IQR) with Kruskal–Wallis and Dunn’s post hoc. Categorical data by chi-squared with Bonferroni correction as appropriate. Cytokines log-transformed prior to analysis; ANOVA with Tukey post hoc; back-transformed and reported as geometric means with 95% CI. Risk ratios (RR) with 95% CI for ICU admission and mortality. Two-sided p<0.05 considered significant. Software: IBM SPSS 25.0 and GraphPad Prism 8.0.2. Ethics: Approved by Capital Region of Denmark Ethics Committee (H-18024256); registered at ClinicalTrials.gov (NCT03795662); informed consent obtained.

Sample: 164 CAP patients with detected pathogens: COVID-19 n=40 (24.4%), influenza A n=25 (15.2%), bacterial CAP n=99 (60.4%). Baseline: Groups comparable in age, sex, CCI, nursing home residence. COVID-19 patients had fewer comorbidities overall. Severity: CURB-65 distribution similar across groups (mild 57%, moderate 34%). ICU and mortality: ICU admission higher in COVID-19 (15%) vs influenza (4%) and bacterial (4%); RR vs influenza 3.75 (95% CI 0.48–29.34), RR vs bacterial 3.71 (95% CI 1.11–12.46). In-hospital mortality: COVID-19 13% vs influenza 4% vs bacterial 3%. Thirty-day mortality: 13% vs 4% vs 3%. 180-day mortality: 15% vs 8% vs 8%. Length of stay median 5 days across groups. Metabolic profile: Admission glucose, HbA1c, adiponectin, and lipid levels (LDL, HDL, total cholesterol, triglycerides) did not differ by etiology. Overall prevalence: known diabetes 15.2%, undiagnosed diabetes 6.1%, pre-diabetes 51.3%; admission hyperglycemia 81% (random glucose ≥6.0 mmol/L); HDL-cholesterol <1.0 mmol/L in 60% of patients. Body composition: BMI, total FM, FFM similar across groups. Overweight 33%; obesity 22%. Abdominal obesity prevalent (73% had waist circumference above recommended thresholds); mean waist circumference lower in COVID-19 vs influenza (96.6 ± 10.1 cm vs 110.4 ± 20.2 cm; p<0.05). Physical capacity and status: Handgrip strength low in >80% of patients; median handgrip ~23 kg across groups. Low physical activity levels in ~85% across groups. Similar Barthel Index and frailty/nutritional risk across groups. COVID-19 patients reported higher pre-hospital quality of life than influenza patients. Inflammatory profile: Compared to influenza, COVID-19 associated with higher IFN-γ (MD 4.14; 95% CI 1.36–12.58; p=0.008), IL-4 (MD 1.82; 95% CI 1.12–2.97; p=0.012), IL-5 (MD 2.22; 95% CI 1.09–4.52; p=0.024), and IL-6 (MD 2.41; 95% CI 1.02–5.68; p=0.044). Compared to bacterial CAP, COVID-19 had higher IFN-γ (MD 6.10; 95% CI 2.53–14.71; p<0.001) and IL-10 (MD 2.68; 95% CI 1.53–4.69; p<0.001). No significant differences among groups for IL-1β, TNF-α, IL-8, IL-18, IL-12p70, CRP, or adiponectin. COVID-19 patients had lower leukocyte counts (and lower neutrophils/monocytes vs bacterial CAP).

Despite similar pre-admission metabolic phenotype, body composition, and physical capacity across CAP etiologies, patients with COVID-19 demonstrated a more pronounced systemic inflammatory response, with elevated IFN-γ and selected cytokines (IL-4, IL-5, IL-6, IL-10) relative to influenza and/or bacterial CAP. The high prevalence of abdominal obesity, pre-diabetes/diabetes, admission hyperglycemia, and low HDL-cholesterol across all groups underscores an unhealthy metabolic profile common among hospitalized CAP patients. However, the augmented inflammatory response in COVID-19, alongside higher ICU admission and mortality, suggests that COVID-19 pathophysiology involves immune activation beyond what can be attributed solely to obesity or metabolic dysregulation. The cytokine pattern is consistent with enhanced adaptive immune activation in COVID-19. Given that COVID-19 patients had a lower mean waist circumference than influenza patients, the greater inflammatory activation appears not to be driven by abdominal adiposity alone. These findings support the hypothesis that SARS-CoV-2 elicits distinct host immune responses that may contribute to more severe outcomes compared with other CAP etiologies.

In hospitalized CAP patients, metabolic profile, body composition, and physical capacity were broadly similar across SARS-CoV-2, influenza, and bacterial etiologies. Nonetheless, COVID-19 was associated with heightened inflammatory responses and higher risks of ICU admission and mortality, indicating that dysregulated metabolism or abdominal obesity alone cannot explain the greater severity observed in COVID-19. Future research should investigate mechanistic links between SARS-CoV-2–specific immune responses and outcomes, and assess targeted interventions that modulate inflammation independent of metabolic status.

Single-center study with potential consent/selection bias as informed consent was required within 24 hours of admission. Self-reported measures (e.g., physical activity, quality of life) may introduce information bias and misclassification despite standardized staff training. Recruitment occurred at different time periods for each etiology (COVID-19 during March–May 2020; bacterial CAP across January 2019–May 2020; influenza during 2018/2019 and 2019/2020 seasons), which may introduce temporal effects. Group sizes were imbalanced, with a relatively small influenza cohort. There were substantial missing data for some measures (e.g., handgrip strength, IPAQ, body composition, lipids, cytokines), though missingness was assumed MCAR. Generalizability may be limited to similar hospital settings and populations.