Comparisons of four cognitive-frailty measures in predicting dementia and disability

Background: Several cognitive-frailty (CF) measurements, such as traditional CF, the CF phenotype, physio-cognitive decline syndrome (PCDS), and motoric cognitive risk syndrome (MCRS) have been developed but their predictive abilities for incident dementia and incident disability are seldom compared. We conducted a 2-year prospective study to compare the associations of traditional CF, the CF phenotype, PCDS, and MCRS with incident dementia and incident disability. Methods: In total, 755 individuals aged 65 years or older, without preexisting dementia or disability, participated in the baseline assessment and were subsequently monitored over a 2-year period. Data on cognitive and frailty components of traditional CF, the CF phenotype, PCDS, and MCRS, were collected. The logistic regression model was used to investigate independent associations of each CF measure with incident dementia and incident disability. Results: In total, 505 participants completed the two annual follow-ups. After adjusting for other CF measures, age, and sex, incident dementia was significantly associated with PCDS (odds ratio [OR] = 2.54; 95% confidence interval [CI], 1.25 ~ 5.19) but was not significantly associated with traditional CF, the CF phenotype, or MCRS, and incident disability was significantly associated with the CF phenotype (OR = 2.90; 95% CI, 1.59 ~ 5.30) but was not significantly associated with traditional CF, PCDS, or MCRS. After adjusting for other CF measures, age, sex, educational level, and other variables, incident dementia was not independently associated with any CF measure, while the association of incident disability with the CF phenotype remained significant (OR = 2.72; 95% CI, 1.45 ~ 5.11). Conclusions: The CF phenotype, MCRS, and PCDS can possibly identify a higher number of CF cases than can the traditional CF measure. While the CF phenotype was a significant predictor of incident disability, all four CF measures lacked an independent association with incident dementia over a 2-year period. Future studies with a longer study period are needed to validate our results.