This study aimed to compare the ability of Timed Up-and-Go (TUG) and usual gait speed (UGS) to predict incident difficulty completing basic and instrumental activities of daily living (ADLs) in older adults free of disability at baseline, and to provide estimates for the probability of incident difficulty at various levels of baseline mobility performance. Data from the first two waves of The Irish Longitudinal Study on Ageing (TILDA), a study assessing health, economic and social aspects of ageing in adults who are 50 years and over. A nationally representative, population-based sample of community-dwelling adults was conducted. Participants aged ≥65 years, who completed mobility tests during a health assessment, had no reported difficulty in ADLs and a Mini-Mental State Examination (MMSE) score ≥24, were re-interviewed after 2 years (n=1,664). Participants completed the TUG and UGS at baseline and indicated difficulty in several basic and instrumental ADLs at follow-up. ROC analysis suggested that TUG and UGS are acceptable tools to predict disability in ADLs and IADLs (AUC=0.65-0.75) with no significant difference between them (p>0.05). Both were very good predictors of difficulty in higher level functioning tasks such as preparing hot meals, taking medications and managing money (AUC>0.80). Predictive probabilities were obtained across a range of performance levels.
(s): TUG and UGS have similar predictive ability in relation to incident disability in basic and instrumental ADLs. Predictive probabilities can be employed in the identification of those most at risk and in need of particular services. As improving physical function can prevent or delay dependence in ADLs, they can also provide performance goals and feedback during exercise interventions.