Advanced oxidation protein products and ischemia-modified albumin as prognostic biomarkers of long-term mortality in community-acquired pneumonia: a prospective observational study

Abstract

Oxidative stress is increasingly recognized as a contributor to the pathophysiology and outcomes of community-acquired pneumonia (CAP). This study examined whether plasma levels of advanced oxidation protein products (AOPPs) and ischemia-modified albumin (IMA) measured at hospital admission are associated with 100-day mortality. A cohort of 71 hospitalized CAP patients was analyzed. Plasma AOPPs and IMA were measured within 24 h of admission and evaluated in relation to clinical data, the Charlson Comorbidity Index (CCI), and routine laboratory parameters. Statistical analyses included receiver operating characteristic (ROC) curve evaluation, Kaplan–Meier survival estimates, and Spearman correlations. AOPPs correlated positively with D-dimer, while IMA showed positive associations with N-terminal pro–B-type natriuretic peptide and high-sensitivity cardiac troponin I, and inversely with serum albumin, indicating links between oxidative stress, inflammation, and cardiovascular dysfunction. Higher admission levels of AOPPs and IMA were independently associated with increased 100-day mortality. In ROC analysis, AOPPs demonstrated good discriminatory ability for 100-day mortality (area under the curve [AUC] = 0.75, p < 0.0001), and a combined multivariable model including AOPPs, IMA, and CCI further improved performance (AUC = 0.851, p < 0.0001). These findings suggest that oxidative stress biomarkers measured at admission may serve as accessible indicators of long-term mortality risk in CAP.

Introduction

Community-acquired pneumonia (CAP) remains the leading cause of infection-related hospital admissions and is defined as a pulmonary infection acquired outside the hospital setting or diagnosed within 48 h of admission in previously non-hospitalized patients. Its incidence ranges from 1 to 25 cases per 1,000 inhabitants annually¹. Among bacterial pathogens, Streptococcus pneumoniae accounts for the majority of pneumonia-related deaths, followed by Haemophilus influenzae. Viral agents, such as SARS-CoV-2, are also common contributors. Nevertheless, in nearly half of all cases, the etiologic agent remains unidentified, largely due to sampling challenges, prior antibiotic exposure, or limited access to molecular diagnostics². Despite advances in diagnostics, therapy, and prevention, CAP remains a major cause of mortality in developed countries³, especially among older and immunocompromised individuals⁴. Severe CAP carries up to 36% mortality despite appropriate antibiotics⁵, while mild to moderate cases resolve within two weeks; however, full recovery may take up to six months⁶. Among patients with comorbidities, long-term deterioration often reflects underlying chronic conditions rather than the acute infection⁷. These findings underscore the importance of prognostic biomarkers for accurate risk stratification and timely escalation of care.

Given the substantial clinical and public health burden of CAP, elucidating its pathophysiology remains essential. A key feature of CAP is a dysregulated inflammatory response, characterized by excessive production of reactive oxygen species (ROS) and the development of oxidative stress (OS). OS reflects an imbalance between ROS generation and endogenous antioxidant defenses. In CAP, ROS serve a dual role: at physiological levels, they contribute to antimicrobial defense, whereas excessive ROS generation leads to oxidative tissue injury, endothelial dysfunction, and cytokine overproduction⁸, ultimately contributing to pulmonary damage, acute respiratory distress syndrome, and multi-organ dysfunction. Although modulation of redox homeostasis and inflammatory signaling may mitigate tissue injury, the precise molecular mechanisms underlying these processes remain incompletely characterized^9,10.

Systemic OS can be assessed using circulating biomarkers that reflect protein oxidation, such as advanced oxidation protein products (AOPPs) and ischemia-modified albumin (IMA). AOPPs, generated via the myeloperoxidase–hydrogen peroxide–halide system, have been linked to various inflammatory and metabolic disorders^11–13 and proposed as prognostic biomarkers in malignancies, including breast cancer¹⁴, as well as in end-stage renal disease¹⁵; however, data regarding their role in CAP remain limited. IMA, a structurally modified form of albumin produced under ischemic and oxidative conditions, is influenced by both demographic and clinical factors¹⁶ and typically increases in states of heightened OS¹⁷. Initially characterized in acute coronary syndromes¹⁶, IMA has subsequently been associated with infectious and inflammatory conditions, including sepsis and pneumonia^18,19, where elevated levels may reflect tissue injury and inflammatory burden, thereby underscoring its potential prognostic relevance. In parallel, endogenous enzymatic antioxidants, particularly superoxide dismutase (SOD) and glutathione peroxidase (GPx), play a crucial role in maintaining redox homeostasis by limiting ROS-mediated damage during lower respiratory tract infections^20,21.

Although OS has been implicated in the pathogenesis and progression of CAP^8–10, few studies have examined oxidative biomarkers as predictors of outcome. Evidence on AOPPs and IMA in this context is particularly limited^11–19, and their long-term prognostic value remains unknown. Most prior research addressed short-term outcomes or focused on sepsis rather than isolated CAP. To date, no study has evaluated whether admission levels of AOPPs and IMA predict 100-day mortality in CAP. By addressing this gap, the present study provides the first evidence that redox imbalance assessed at admission may identify patients at risk of delayed mortality. This work therefore advances current understanding of oxidative mechanisms in CAP and highlights the potential prognostic significance of integrating redox biomarkers into clinical risk stratification and decision-making.

Materials and methods

Study design

This prospective observational pilot study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Nicolaus Copernicus University in Toruń, Collegium Medicum in Bydgoszcz, Poland (KB/343/2021). Written informed consent was obtained from all participants. The study included 71 adults hospitalized with community-acquired pneumonia (CAP) at the Regional Center of Pulmonology in Bydgoszcz, Poland, between March 2023 and February 2024. CAP was diagnosed based on high-resolution computed tomography or chest radiography. Patients requiring noninvasive or mechanical ventilation were excluded from the study. All patients received standard oxygen therapy and routine laboratory testing on admission.

Measurements of AOPPs

The serum AOPP level was measured using a modified method by Witko-Sarsat, which utilizes absorbance at 340 nm²². Detection was further refined according to Hanasand’s improvements, where citric acid replaced acetic acid to enhance sample stability over time²³. Briefly, the reactive mixture was prepared by combining 1.875 mL of 0.2 M citric acid (Avantor, cat. no. PA-06-538210118) with 25 µL of 1.16 M potassium iodide (Avantor, cat. no. PA-06-743160117). Then, 1.9 mL of this solution was mixed with 100 µL of plasma, and absorbance at 340 nm was measured immediately using a JASCO V-550 UV-Vis spectrophotometer (Japan). Results were expressed in molar units as chloramine T equivalents.

Measurements of IMA

Measurement of IMA was conducted using the colorimetric method developed by Bar-Or et al.²⁴. A total of 100 µL of plasma was mixed with 50 µL of 0.2% cobalt(II) chloride (Sigma-Aldrich, cat. no. 232696) and incubated for 10 min to allow for cobalt–albumin binding. Then, 50 µL of 1.5 mg/mL dithiothreitol (Roche, cat. no. 10197777001) was added to react with unbound cobalt. After 2 min, 2 mL of 0.9% NaCl (Avantor, cat. no. PA-06-794121116) was added to stop the reaction, and absorbance was measured at 470 nm using a JASCO V-550 UV-Vis spectrophotometer (Japan).

Measurements of SOD

SOD activity was assayed in plasma using a commercial Superoxidase Dismutase Assay Kit (CaymanChemical, USA). Briefly, samples were thawed, diluted, and analyzed in duplicate with standard curve calibration. Absorbance was read at 450 nm (Multiskan GO, Thermo Scientific). The SOD standard value was plotted as a function of final SOD activity (U/ml), and SOD activity (U/mL) was calculated from the regression equation derived from the standard curve.

Measurements of GPx

GPx activity was measured using a commercial Glutathione Peroxidase Assay Kit (Cayman Chemical, USA) following the manufacturer’s instructions. Plasma samples were thawed, diluted, and analyzed in triplicate with background and control wells included. Absorbance was measured at 340 nm (Multiskan GO, Thermo Scientific), and enzyme activity was calculated according to the kit protocol.

Quality control

All assays were performed in duplicate, and intra-assay coefficients of variation were below 5%, indicating high analytical precision and reproducibility.

Routine laboratory measurements

Standard biochemical and hematological parameters, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), C-reactive protein (CRP), D-dimers, high-sensitivity cardiac troponin I (hs-cTnI), N-terminal pro–B-type natriuretic peptide (NT-proBNP), lactate dehydrogenase (LDH), and albumin, were obtained from hospital records at admission. All analyses were performed using validated, automated methods in the accredited reference laboratory of the Regional Center of Pulmonology in Bydgoszcz, following standard operating procedures and internal quality control protocols. Hematological indices such as white blood cell count (WBC), red blood cell count (RBC), hemoglobin (HGB), hematocrit (HCT), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), platelet count (PLT), Red Cell Distribution Width-Standard Deviation (RDW-SD), and Red Cell Distribution Width-Coefficient of Variation (RDW-CV) were analyzed using the same standardized procedures.

Statistical analysis

Statistical analyses and visualizations were performed using Statistica software, version 13.3 (StatSoft^®, Kraków, Poland). Data distribution was assessed with the Shapiro–Wilk test. As most variables deviated from normality, intergroup comparisons were conducted using the nonparametric Mann–Whitney U test. Associations between OS biomarkers and clinical or laboratory parameters were evaluated using Spearman’s rank correlation coefficients. Mortality risk was analyzed by receiver operating characteristic (ROC) curve analysis, with predictive accuracy expressed as the area under the curve (AUC). Optimal cut-off values for AOPPs and IMA were determined using the Youden index, maximizing the sum of sensitivity and specificity for predicting 100-day mortality. These thresholds (AOPPs > 7 µM and IMA > 0.92) were subsequently used to stratify patients in Kaplan–Meier survival analyses followed by the log-rank test. The prognostic value of biomarkers was further examined using multivariable logistic regression analysis and ROC curves constructed for the combined models. To assess the internal validity and potential overfitting of the prognostic models, bootstrap resampling with 1000 iterations was performed. In each bootstrap sample, the logistic regression models including AOPPs and IMA, and AOPPs, IMA, and CCI were refitted, and the area under the ROC curve was recalculated. The optimism-corrected AUC was obtained as the apparent AUC minus the bootstrap-estimated optimism. These analyses were considered exploratory internal validation rather than proof of external generalizability. Internal validation analyses were performed in R version 4.5.2 using the RStudio Desktop environment version 2025.09.2–418 (Posit, Boston, MA, USA). A two-sided p-value < 0.05 was considered statistically significant.

Results

Baseline characteristics of the study population

The study population comprised 71 patients hospitalized with CAP. The median age was 70 years (interquartile range [IQR], 62–76 years), and all patients were followed for a period of 100 days. Comorbidities were evaluated using the Charlson Comorbidity Index (CCI)²⁵, which assigns weighted scores (1–6) to 19 predefined conditions. Data on cardiovascular, renal, hepatic, metabolic, and malignant diseases were extracted from medical records, with CCI scoring based on the most severe manifestation within each category (Table 1).

Table 1.

Comorbidities in patients hospitalized for community-acquired pneumonia (n = 71).

Common comorbidities	Number of patients (%)
Cardiovascular disease	35 (49.30%)
Neurological disease	10 (14.08%)
Respiratory disease	16 (22.54%)
Rheumatic disease	5 (7.04%)
Gastrointestinal disease	4 (5.63%)
Metabolic disease	12 (16.90%)
Renal disease	2 (2.82%)
Cancer	22 (30.99%)

Associations between oxidative stress biomarkers and clinical parameters

Significant correlations were observed between oxidative stress biomarkers and selected laboratory parameters (Table 2). Although ALT and AST activities remained within reference ranges, both showed significant positive correlations with GPx activity (ALT: r = 0.3786, p < 0.01; AST: r = 0.4006, p < 0.001). D-dimer concentrations showed a significant positive correlation with AOPPs (r = 0.2676, p < 0.05). In addition, N-terminal pro–B-type natriuretic peptide (NT-proBNP) levels correlated significantly with IMA (r = 0.3353, p < 0.01). Hematological parameters showed significant associations with oxidative stress biomarkers. RBC, HGB, and hematocrit were inversely correlated with IMA (all p < 0.05). In addition, RDW-SD and RDW-CV correlated positively with IMA and inversely with SOD and GPx activity. Serum albumin concentration was reduced (median: 3.14 g/dL) and showed a significant inverse correlation with IMA (r = − 0.3941, p < 0.001). In addition, LDH activity correlated positively with GPx activity (r = 0.4052, p < 0.001).

Table 2.

Demographic, clinical, and laboratory characteristics, as well as their correlations with oxidative stress biomarkers (AOPPs, IMA, SOD, and GPx), in patients with community-acquired pneumonia.

Parameters (Units)	Median	Interquartile range	Reference Values	Correlation with
				AOPPs	IMA	SOD	GPx
Age (years)	70.00	62.00–76.00		-0.0220	0.2287	-0.0560	-0.0545
ALT (U/L)	23.95	17.40–50.10	< 40.00	0.0880	0.0849	0.0835	0.3786**
AST (U/L)	27.50	21.20–40.20	< 37.00	0.0087	-0.0208	-0.0244	0.4006***
CRP (mg/L)	55.93	20.46 − 133.04	< 5.00	-0.1084	0.2245	0.1323	-0.0784
D-Dimers (ng/mL)	1342.46	922.72–2970.00	< 500.00	0.2676*	0.1242	-0.1675	-0.0178
hsTNI (ng/L)	4.70	0.08–16.30	< 19.00	0.1955	0.2814*	-0.1211	0.0174
NT-proBNP (pg/mL)	767.00	224.00–2118.00	< 125.00	0.2098	0.3353**	-0.1402	-0.1303
LDH (U/L)	396.00	297.00-535.00	225.00-450.00	0.3024*	-0.0067	-0.0457	0.4052***
Albumin (g/dL)	3.14	2.80–3.53	3.90–5.10	-0.1119	-0.3941***	-0.1567	0.0099
WBC (10^3/µL)	9.72	7.42–12.30	4.00–10.00	0.2010	0.1136	0.0036	-0.1959
RBC (10^6/µL)	3.91	3.43–4.36	4.50–5.50	-0.0441	-0.4095***	0.0024	0.2651*
HGB (g/dL)	11.80	9.90–13.20	14.00–18.00	-0.0395	-0.3242**	0.0989	0.2423*
HCT (%)	35.70	30.50–39.90	40.00–54.00	-0.0484	-0.3830**	0.0402	0.2360*
MCV (fL)	90.50	87.30–93.50	84.00–94.00	0.0006	0.0599	0.0519	-0.1690
MCH (pg)	30.20	28.90–30.90	27.00–34.00	0.0140	0.1022	0.1983	0.0369
MCHC (g/dL)	32.90	32.20–34.10	31.00–37.00	0.0161	0.1434	0.2187	0.2502*
PLT (10^3/µL)	295.00	210.00–393.00	130.00-350.00	-0.0028	0.0806	0.0126	-0.1342
RDW-SD (fL)	45.90	42.80–50.30	35.10–43.90	0.1875	0.3338**	-0.3147**	-0.2407*
RDW-CV (%)	13.60	12.80–15.60	11.60–14.40	0.1882	0.2862*	-0.3259**	-0.1900
CCI (points)	5.00	2.00–6.00		0.1297	0.2965*	-0.1986	-0.0219
AOPPs (µM)	6.59	5.42–8.70		–	0.2864*	0.0627	0.0927
IMA	0.99	0.74–1.18		0.2864*	–	0.0622	-0.0734
SOD (U/mL)	0.89	0.28–1.45		0.0627	0.0622	–	0.1141
GPx (U/mL)	1.62	1.35–1.95		0.0927	-0.0734	0.1141	–

Abbreviations: ALT: Alanine Aminotransferase; AST: Aspartate Aminotransferase; CRP: C-Reactive Protein; hsTNI: High-sensitivity troponin I; NT-proBNP: N-terminal pro b-type Natriuretic Peptide; LDH: Lactate Dehydrogenase; WBC: White Blood Cell Count; RBC: Red Blood Cell Count; HGB: Hemoglobin; HCT: Hematocrit; MCV: Mean Corpuscular Volume; MCH: Mean Corpuscular Hemoglobin; MCHC: Mean Corpuscular Hemoglobin Concentration; PLT: Platelet Count; RDW-SD: Red Cell Distribution Width-Standard Deviation; RDW-CV: Red Cell Distribution Width-Coefficient of Variation; CCI: Charlson Comorbidity Index; AOPPs: Advanced Oxidation Protein Products; IMA: ischemia-modified albumin; SOD: superoxide dismutase; GPx: glutathione peroxidase. Correlation coefficients and significance between parameters were calculated according to Spearman’s method: *p < 0.05, **p < 0.01, ***p < 0.001.

Association of oxidative stress biomarkers with 100-day mortality

Figure 1  shows the differences in plasma levels of AOPPs (Fig. 1A) and IMA (Fig.  1B) between survivors and non-survivors with CAP. Both biomarkers were significantly higher in patients who died within 100 days of hospital admission (p < 0.05).

Fig. 1.

Box plots showing plasma levels of AOPPs (A) and IMA (B) in survivors and non-survivors with CAP. The boxes represent the interquartile range with the median line shown, and whiskers indicate non-outlier ranges. Data points exceeding 1.5-fold the interquartile range are classified as outlier values. Statistically significant differences between groups (p < 0.05) are indicated.

Discriminatory performance of oxidative stress biomarkers for 100-day mortality

The discriminatory performance of oxidative stress biomarkers for 100-day mortality in CAP was evaluated using ROC analysis (Fig. 2). Among individual markers, AOPPs and IMA demonstrated the highest discriminatory ability (AUC = 0.750 and 0.740, respectively; Table 3). The combined model incorporating both biomarkers showed improved performance (AUC = 0.803). Further inclusion of the Charlson Comorbidity Index (CCI) resulted in additional improvement in model performance (AUC = 0.851; sensitivity, 91.7%; specificity, 71.1%; p < 0.0001). In bootstrap internal validation with 1000 resamples, the apparent and optimism-corrected AUC values were similar. For the model including AOPPs and IMA, the apparent AUC was 0.803 with an estimated optimism of 0.014, yielding an optimism-corrected AUC of 0.789. For the model including AOPPs, IMA, and CCI, the apparent AUC was 0.851 with an estimated optimism of 0.020 and an optimism-corrected AUC of 0.831. Optimism-corrected calibration slopes were approximately 1 for the single-predictor models; the slope was 0.914 for AOPPs + IMA and 0.884 for AOPPs + IMA + CCI. These findings suggest only modest overfitting; however, the estimates should still be interpreted with caution, given the limited number of events and the absence of external validation.

Fig. 2.

Graphs showing Receiver Operating Characteristic (ROC) curves for the levels of AOPPs, IMA, GPx, SOD, and CCI (A) and for logistic regression models AOPPs + IMA and AOPPs + IMA + CCI (B) for predicting long-term mortality.

Table 3.

Results of the prognostic accuracy of the studied biomarkers.

Parameter/Model (Units)	Cut-off	AUC	CI	Sensitivity %	Specificity %	p-value
AOPPs (µM)	7.00	0.750	0.633–0.867	79.20	71.70	< 0.0001
IMA	0.92	0.740	0.623–0.857	87.50	51.10	0.0001
SOD (U/mL)	0.218	0.565	0.419–0.712	36.00	82.20	0.3820
GPx (U/mL)	1.737	0.516	0.371–0.662	48.00	71.10	0.8248
CCI (points)	3.00	0.725	0.609–0.842	100.00	39.10	0.0001
AOPPs + IMA	0.2668	0.803	0.700-0.905	83.30	66.70	< 0.0001
AOPPs + IMA + CCI	0.2594	0.851	0.762–0.940	91.70	71.10	< 0.0001

Abbreviations: AOPPs: Advanced Oxidation Protein Products; IMA: ischemia-modified albumin; SOD: superoxide dismutase; GPx: glutathione peroxidase; CCI: Charlson Comorbidity Index; AUC: Area under the curve; CI: Confidence Interval. Statistically significant differences between groups are indicated by bold p-values.

In contrast, enzymatic antioxidants SOD and GPx demonstrated limited discriminatory ability for 100-day mortality, with AUC values of 0.565 and 0.516, respectively, and non-significant p-values (Table 3).

Kaplan-Meier survival analysis according to AOPPs and IMA cut-off values

Figure 3 presents Kaplan-Meier survival curves illustrating 100-day survival stratified by ROC-derived cut-off values for AOPPs and IMA. Differences between strata were assessed using the log-rank test. Patients with AOPPs levels > 7 µM or IMA absorbance > 0.92 showed significantly shorter survival compared with those below these thresholds (p = 0.0001 and p = 0.0026, respectively).

Fig. 3.

Kaplan–Meier survival curves for the levels of AOPPs (A) and IMA (B) in the study group, based on cut-off points determined by ROC analysis. Significant differences between groups are indicated by bold p-values (Log-rank test).

Discussion

In this prospective observational study, we investigated the association between oxidative stress biomarkers and long-term mortality in patients hospitalized with community-acquired pneumonia, as well as their relationships with routinely assessed clinical and laboratory parameters. The main finding is that higher admission levels of AOPPs and IMA were associated with increased 100-day mortality. In addition, these biomarkers showed significant associations with markers of hepatic injury, coagulation, cardiac stress, and hematological alterations, supporting their relevance to the severity of systemic disease. AOPPs and IMA also demonstrated moderate discriminatory performance for mortality, which improved when combined with comorbidity burden, indicating their potential relevance for risk stratification. Importantly, these findings should be interpreted as associations rather than evidence of causality, while still providing clinically relevant and hypothesis-generating insight into the role of oxidative stress in the long-term outcomes of CAP^9,10,18.

Associations of oxidative stress biomarkers with routine clinical and laboratory parameters

In our study, IMA showed significant positive correlations with NT-proBNP and high-sensitivity troponin I (hsTnI). While natriuretic peptides are well-established markers of disease severity and short-term prognosis in CAP²⁶, previous studies evaluating IMA in CAP did not include parallel assessment of cardiac biomarkers¹⁸. Associations between IMA and natriuretic peptides have been reported primarily in cardiovascular populations¹⁶. Our findings, therefore, extend existing observations and suggest that systemic oxidative stress during pneumonia may be linked to secondary myocardial strain. However, these relationships should be interpreted as associations rather than indicators of direct cardiac injury mechanisms.

Oxidative stress biomarkers were also associated with markers of coagulation and endothelial perturbation. In particular, AOPPs correlated positively with D-dimer concentrations, consistent with prior reports linking oxidative stress to hemostatic activation in severe infection. These findings support the concept that protein oxidation markers may reflect the broader systemic response to inflammation and vascular dysfunction in CAP²⁴.

In contrast, enzymatic antioxidants demonstrated a different pattern of associations. SOD and GPx showed limited discriminatory ability for mortality and appeared to be more closely related to routine laboratory parameters. GPx activity correlated positively with liver enzymes (ALT, AST) and LDH, as well as with red blood cell indices, suggesting a potential adaptive antioxidant response in the context of systemic inflammation and cellular injury. Similar associations between GPx activity and hepatic enzymes have been reported in other inflammatory conditions²⁰. SOD activity, however, showed fewer systemic associations, which may reflect its intracellular localization and context-dependent regulation^21,27.

Albumin-related parameters further highlight the systemic nature of oxidative stress in CAP. The inverse relationship between serum albumin and IMA underscores albumin’s role as both a major extracellular antioxidant and a target of oxidative modification¹⁸. Associations between IMA and hematological indices, including red blood cell count, hemoglobin, and RDW, suggest that oxidative albumin modification may integrate redox imbalance with inflammation-related hematological alterations. These findings are consistent with previous observations linking oxidative stress to anemia and erythrocyte heterogeneity in inflammatory states.

Prognostic relevance of AOPPs and IMA

AOPPs and IMA have been increasingly recognized as markers reflecting oxidative stress in respiratory infections, including community-acquired pneumonia (CAP)^9,28; however, evidence supporting their clinical relevance remains limited. In our cohort, patients who died within 100 days of hospital admission had significantly higher plasma levels of AOPPs and IMA compared with survivors. Previous studies have reported elevated AOPPs in CAP patients relative to healthy controls^29,30, while Bolatkale et al. demonstrated the diagnostic utility of IMA in CAP in the emergency department, without assessing its relationship with patient survival¹⁸. Our findings, therefore, extend existing observations by linking admission levels of these biomarkers with long-term outcomes following CAP.

In addition, survival analyses demonstrated that patients stratified by ROC-derived cut-off values of AOPPs and IMA exhibited significantly different 100-day survival probabilities, further supporting an association between higher oxidative stress burden at admission and adverse long-term outcomes. Importantly, AOPPs and IMA demonstrated moderate discriminatory performance for 100-day mortality, which improved when combined with comorbidity burden. These results suggest that oxidative stress biomarkers may provide complementary information to routine clinical assessment rather than serving as standalone prognostic indicators. Given the observational nature of the study and the limited number of events, the present results should be viewed as exploratory and require confirmation in larger cohorts. Taken together, these associations support the relevance of systemic oxidative stress in the post-acute course of CAP. While the underlying mechanisms cannot be inferred from the present study, our results highlight the value of further investigating AOPPs and IMA in larger, multicenter cohorts with longitudinal follow-up and direct comparison to established severity scores.

Limitations of the study

Despite the promising findings, several limitations should be acknowledged. First, the relatively small sample size, limited number of events, and single-center design may restrict the generalizability of the results. Second, although selected clinical variables were included, several potentially important confounding factors, such as comorbidities, nutritional status, renal function, inflammatory burden, and medication use, were not fully adjusted for and may independently influence oxidative stress biomarkers and clinical outcomes. Third, biomarker measurements were obtained at a single time point on admission; thus, longitudinal changes in oxidative stress during the disease course and recovery could not be assessed. Fourth, the prognostic performance of AOPPs and IMA was not compared with established CAP severity scores, such as CURB-65 or the Pneumonia Severity Index, which limits conclusions regarding their incremental value in clinical risk stratification. Fifth, the clinical applicability of these biomarkers remains constrained by the limited standardization of assays and their absence from routine diagnostic workflows. Finally, as this was an observational study, causal inferences cannot be drawn. Future studies should validate these findings in larger, multicenter cohorts, incorporate longitudinal biomarker assessment, and directly compare oxidative stress markers with established severity scores to better define their potential role in prognostic evaluation of CAP.

Conclusion

This study indicates that oxidative stress-related biomarkers, including AOPPs and IMA, are associated with systemic redox imbalance in patients hospitalized with CAP. Higher admission levels of these biomarkers were associated with increased 100-day mortality, suggesting their relevance to longer-term outcomes beyond the acute phase of infection. The combined assessment of AOPPs, IMA, and comorbidity burden showed improved discriminatory performance, indicating that oxidative stress markers may provide complementary information for risk stratification. Given the observational design and methodological limitations, these findings should be considered hypothesis-generating and warrant validation in larger, multicenter studies with longitudinal follow-up before clinical implementation.

Author contributions

M.N.-M. and T.W.: Conceptualization; T.W., M.N.-M., P.W., J.S., W.M.: methodology, M.N.-M.: data analysis; M.N.-M. and T.W.: investigation; M.P.-C. and J.P.: resources; M.N.-M., T.W., and B.C.: writing—original draft; M.N.-M., B.C., and T.W.: writing—review and editing; M.N.-M.: visualization; B.C. and S.K.: supervision; M.N.-M. and T.W.: project administration. All authors have read and agreed to the published version of the manuscript.

Data availability

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical statement

This study was conducted in accordance with the guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of Nicolaus Copernicus University in Toruń, Collegium Medicum in Bydgoszcz, Poland (KB 343/2021).