Immune–inflammatory vulnerability index for risk stratification in older adults with acute exacerbations of COPD: a prospective cohort study

BACKGROUND: Acute exacerbations of chronic obstructive pulmonary disease (AECOPD) are a major cause of morbidity, hospitalization, and short-term mortality in older adults. Growing evidence suggests that aging-related immune remodeling and immune-inflammatory vulnerability (IMV) substantially contribute to biological heterogeneity, susceptibility, and adverse outcomes in this population. However, clinically applicable tools that quantitatively capture this multidimensional biological vulnerability remain limited. This study aimed to develop and externally validate an Immune-Inflammatory Vulnerability Index, hereafter referred to as IAS, conceptualized as a biologically informed index of IMV, to support immune phenotype stratification and short-term risk assessment in older adults hospitalized with AECOPD. METHODS: In this prospective observational cohort study, 219 patients aged ≥ 60 years hospitalized for AECOPD were enrolled in the derivation cohort, and 136 independent patients from an external center were included for validation. Immune phenotyping was performed within 48 h of admission, including PD-1⁺CD4⁺ T cells, CD28⁻CD8⁺ T cells, and serum interleukin-6 (IL-6). IAS was constructed using biologically prespecified immune biomarkers with LASSO-based coefficient weighting, using infection-related versus non-infection-related AECOPD as a supervised biological contrast to enhance immune phenotype differentiation rather than as a direct predictive endpoint. Primary objectives were to evaluate IAS for immune phenotype stratification and to assess its associations with predefined short-term clinical outcomes, specifically 30-day readmission and 90-day all-cause mortality. RESULTS: IAS demonstrated strong immune phenotype stratification performance for distinguishing infection-related from non-infection-related AECOPD, with an area under the curve (AUC) of 0.842 (95% CI: 0.771-0.901) in the derivation cohort and 0.826 (95% CI: 0.754-0.885) in the external validation cohort. For prediction of 90-day all-cause mortality, IAS achieved an AUC of 0.824 (95% CI: 0.750-0.890). Higher IAS was independently associated with increased 30-day readmission, prolonged hospitalization, and elevated 90-day mortality risk after adjustment for age, disease severity, and established clinical risk scores (adjusted HR = 3.42, 95% CI: 2.01-5.82). Sensitivity analyses showed that IAS outperformed IL-6 alone and retained discriminatory capacity even after exclusion of IL-6, supporting its broader immune-inflammatory relevance beyond acute inflammatory burden. Longitudinal analyses demonstrated partial remission-associated declines in IAS, consistent with state-responsive biological vulnerability rather than a fixed immune-aging trait. CONCLUSIONS: IAS is a biologically informed, geriatric-oriented index of IMV that integrates aging-related immune remodeling with dynamic inflammatory responses in older adults with AECOPD. Rather than serving as a definitive measure of fixed immune-aging burden or as a dedicated infection classification tool, IAS provides a quantitative framework for biologically relevant immune-inflammatory vulnerability stratification and short-term prognostic assessment. Further prospective multicenter studies are warranted to refine clinical applicability, evaluate implementation feasibility, and improve translational potential before routine clinical adoption can be recommended.