Supplementary Material for: Combination of the Clock Drawing Test with the Physical Phenotype of Frailty for the Prediction of Mortality and Other Adverse Outcomes in Older Community Dwellers without Dementia

Background: Cognitive assessment is thought to increase the ability of the physical phenotype of frailty to identify older persons at a higher risk for adverse outcomes. Objective: Data from a cohort of dementia-free community dwellers were used to investigate whether the clock drawing test (CDT), a quick and easy cognitive screening test, is associated with adverse health outcomes independently of the physical phenotype of frailty. Methods: This was a prospective population-based cohort study of 766 dementia-free Italian community dwellers aged 65 years or older. Baseline assessment included the physical phenotype of frailty, 3 different CDT protocols [Sunderland, Shulman, and the clock drawing interpretation scale (CDIS)], and several health confounders. Hazard ratios (HR) and odds ratio (OR) along with their corresponding 95% confidence intervals (CI) from models adjusted for frailty and sociodemographic and health confounders were used to estimate the independent association of the CDT with the 7-year risk of all-cause mortality and the 3-year risk of new and worsening disability, hospitalization, and fractures. Results: After adjustment for confounders, the Sunderland CDT was significantly associated with all-cause mortality independently of the physical phenotype of frailty (HR = 1.44, 95% CI 1.03-2.01, p = 0.031). However, compared to all nonfrail participants with a normal Sunderland CDT, the HR was 1.57 (95% CI 1.09-2.26, p = 0.016) for those with impairment on the Sunderland CDT only, 2.48 (95% CI 1.46-4.20, p = 0.001) for those with frailty only, and 2.52 (95% CI 1.34-4.77, p = 0.004) for those with both frailty and impairment on the Sunderland CDT. Mortality was unrelated to the CDIS CDT (p = 0.359) and the Shulman CDT (p = 0.281). No statistically significant relationship was found between nonlethal outcomes and any CDT protocol, although trends were found for an association of both the Sunderland CDT (p = 0.118) and the CDIS CDT with worsening disability (p = 0.154). Conclusions: In older persons, depending on the scoring system, the CDT may predict the mortality risk independently of the physical phenotype of frailty. However, combining the two measurements does not improve their individual prognostic abilities.