Does the timed up and go test predict future falls among British community-dwelling older people? Prospective cohort study nested within a randomised controlled trial

Gotaro Kojima, Tahir Masud, Denise Kendrick, Richard Morris, Sheena Gawler, Jonathan Treml, Steve Iliffe

Falling is common among older people. The Timed-Up-and-Go Test (TUG) is recommended as a screening tool for falls but its predictive value has been challenged. The objectives of this study were to examine the ability of TUG to predict future falls and to estimate the optimal cut-off point to identify those with higher risk for future falls. This is a prospective cohort study nested within a randomised controlled trial including 259 British community-dwelling older people ≥65 years undergoing usual care. TUG was measured at baseline. Prospective diaries captured falls over 24 weeks. A Receiver Operating Characteristic curve analysis determined the optimal cut-off point to classify future falls risk with sensitivity, specificity, and predictive values of TUG times. Logistic regression models examined future falls risk by TUG time. Sixty participants (23%) fell during the 24 weeks. The area under the curve was 0.58 (95% confidence interval (95% CI) = 0.49-0.67, p = 0.06), suggesting limited predictive value. The optimal cut-off point was 12.6 seconds and the corresponding sensitivity, specificity, and positive and negative predictive values were 30.5%, 89.5%, 46.2%, and 81.4%. Logistic regression models showed each second increase in TUG time (adjusted for age, gender, comorbidities, medications and past history of two falls) was significantly associated with future falls (adjusted odds ratio (OR) = 1.09, 95% CI = 1.00-1.19, p = 0.05). A TUG time ≥12.6 seconds (adjusted OR = 3.94, 95% CI = 1.69-9.21, p = 0.002) was significantly associated with future falls, after the same adjustments. TUG times were significantly and independently associated with future falls. The ability of TUG to predict future falls was limited but with high specificity and negative predictive value. TUG may be most useful in ruling in those with a high risk of falling rather than as a primary measure in the ascertainment of risk.