Development and validation of composite inflammaging metrics: Findings from two prospective cohort studies

BackgroundAging-related inflammation is associated with chronic diseases and mortality. This study aimed to: 1) develop composite inflammaging metrics (CIMs) in UK biobank (UKB), and validate them in UKB and National Health and Nutrition Examination Survey (NHANES); 2) estimate mortality and CVD risk predictions of CIMs; 3) compare CIMs with single inflammatory blood biomarkers and conventional inflammatory indexes; 4) examine associations between lifestyles and CIMs. MethodsWe utilized algorithms including multiple linear regression, principal component analysis (PCA), allostatic load (AL), and Klemera and Doubal method (KDM), to develop four CIMs from five inflammatory blood biomarkers, using data of 438,321 adults (40-70 years) from UK Biobank (UKB). We validated these CIMs in UKB and 10,667 adults (20-84 years) from NHANES IV. We performed a parametric proportional hazard model based on Gompertz distribution to estimate CVD and mortality risk predictions of CIMs. Areas under receiver operating characteristic curves (AUCs) were calculated to compare the predictive abilities of CIMs. Multiple linear regression models were used to access associations between lifestyles and CIMs. ResultsWith adjustment for age and sex, four CIMs were significantly associated with higher risks of all-cause mortality and incident CVD in UKB, among which CIMKDM outperformed the others (all-cause mortality: hazard ratio [HR] = 1.48, 95% confidence interval [CI] = 1.46, 1.50; incident CVD: HR = 1.34, 95% CI = 1.33, 1.36). CIMKDM had the best discriminative ability for predicting 10-year survival and incident CVD in UKB (all-cause mortality: AUC = 0.728; incident CVD: AUC = 0.712). CIMs were responsive to lifestyle variables. For example, in UKB, compared to never smokers, current smokers had a significant increment in CIMKDM (coefficient = 0.30 SD, P < 0.001). Similar results were well validated in NHANES IV. ConclusionsWe developed and validated four novel CIMs that were predictive of mortality and CVD risk. CIMKDM outperformed the others and had the potential to be used in aging related preventive and intervention programs. Intervention programs targeting lifestyles could slow inflammaging and further reduce disease burden.