The Free Triiodothyronine/Free Thyroxine Ratio Is Associated with Frailty in Older Adults: A Longitudinal Multisetting Study.

Background: Various models have been proposed to predict frailty, including those based on clinical criteria and phenotypes. However, a simple biomarker associated with frailty has been not yet identified. The aim of this study is to evaluate the relationship between free triiodothyronine (fT3)/free thyroxine (fT4) ratio value and the degree of frailty among three different cohorts of older individuals: (1) acutely ill hospitalized patients, (2) nursing-home (NH) residents, and (3) home-dwelling centenarians. Methods: We performed a secondary analysis of de-identified patient-level data from two prospective observational studies on acutely hospitalized older patients (Geriatric Acute Unit [GAU]), and home-dwelling centenarians (CENT), and a retrospective-prospective observational study on older NH residents. Demographic characteristics, along with a 30-items Frailty Index (FI) and serum thyrotropin, fT3 and fT4 measurements were obtained. Results: Six hundred fifteen individuals (aged 86.4 ± 8.9 years; 55.1% females) were included in the study, including 298 (48.5%) GAU, 250 (40.6%) NH, and 67 (10.9%) CENT. A significant inverse relationship between fT3/fT4 ratio and FI values was observed ( ρ s  = -0.17 [confidence interval; CI: -0.092 to 0.252], p  < 0.001), and this was confirmed by logistic multivariate analysis ( β  = -0.44, odds ratio [OR]: 0.64 [CI: 0.47-0.87], p  < 0.001) (after adjustment for age, sex, and cohorts). Moreover, a progressively decreased mortality risk was associated with rising fT3/fT4 ratio (OR 0.60 [CI: 0.44-0.80] β  = -0.51, p  < 0.001]. Conclusions: The fT3/fT4 ratio value was inversely correlated with frailty degree and mortality risk in a large cohort of older individuals, including centenarians, regardless of their sex and clinical condition. fT3/fT4 ratio value could represent an easily measured independent biochemical marker of frailty degree in older people.