Lower limb reaction time discriminates between multiple and single fallers

Abstract

Despite research supporting its validity in predicting falls, simple reaction time (SRT) is not typically included in a clinical falls assessment battery because there is no standardised, clinically feasible testing protocol with published cutoff scores. This study aimed to determine whether SRT scores obtained using a clinically viable protocol could discriminate between multiple and single fallers. SRT scores were obtained from 147 participants (88 women and 59 men) over the age of 65 years (mean 80.1 years; SD 8.4 years) who had reported one or more falls in the previous 12 months. Forty-eight single and 99 multiple fallers were recruited from the subacute public hospital sector. SRT scores significantly discriminated between faller groups (Wilk's λ = 0.89, p = 0.05), with SRT score being the dominating predictor between groups. A receiver operating characteristic (ROC) curve showed good diagnostic accuracy with the area under curve = 0.78. A cutoff score of 394.5 ms provided the best balance between sensitivity and specificity. Once data were dichotomized as being above or below this cutoff value, a binary regression analysis revealed an odds ratio of 7.18 (95% CI 3.3–15.6), with a positive predictive value of 84.5%. The clinically feasible lower limb SRT testing protocol described provided good discrimination between single and multiple fallers. This test may be useful in clinical practice to help identify older people at greater risk of future falls.