Abstract
Background/Objectives: The CTI-FI integrates inflammatory–metabolic dysfunction with accumulated physiological vulnerability. Previous evidence linking CTI-FI to incident CMM has been limited to the Chinese CHARLS cohort, without external validation in other populations. We assessed baseline and cumulative average CTI-FI in relation to incident CMM and evaluated incremental and survival-based prediction across CHARLS and ELSA. Methods: Adults aged ≥45 years were analyzed in the China Health and Retirement Longitudinal Study (CHARLS; 2011–2020) and the English Longitudinal Study of Ageing (ELSA; 2004–2018). Baseline CTI-FI and cumulative average CTI-FI (caCTI-FI; the mean of two CTI-FI measurements) were examined using Cox regression, restricted cubic splines, subgroup and sensitivity analyses, and survival discrimination. Six survival algorithms were compared in CHARLS; the selected Elastic Net Cox models were evaluated in ELSA without updating. Results: Baseline cohorts included 7108 participants (627 CMM events). Fully adjusted hazard ratios per unit of baseline CTI-FI were 1.46 (95% CI 1.35–1.58) in CHARLS and 1.31 (1.17–1.46) in ELSA; corresponding estimates for caCTI-FI were 1.61 (1.41–1.84) and 1.35 (1.11–1.64). Adding caCTI-FI increased apparent Harrell’s C from 0.721 to 0.759 in CHARLS and from 0.756 to 0.776 in ELSA. In within-cohort Cox comparisons, gains over FI were modest in CHARLS and uncertain in ELSA. The selected models achieved external Harrell’s C values of 0.715 (0.680–0.752) and 0.719 (0.654–0.777), respectively, but substantially overestimated 4-year absolute risk in ELSA. Conclusions: Baseline and cumulative average CTI-FI were associated with incident CMM in both cohorts, supporting their potential as epidemiological risk markers. The survival models retained discrimination in ELSA, although their absolute-risk estimates require recalibration.