Relationships between estimation errors and falls in healthy aged dwellers.

Background: Although fall predictions using motor ability have been well reported in elderly people, there are few reports on physical cognitive ability. Objective: To examine the relationship of the results of motor function tests that include physical cognitive ability on the ability to predict falls and to determine which test is the most appropriate. Methods: We studied 174 community-dwelling elderly adults (mean age 75.7 +/- 5.7, 41 males and 133 females), and measured grip strength, one-leg standing time (OLS), timed up and go test (TUG), functional reach test, sit and reach test, and maximal step length (MSL). The estimation error (EE), which was defined as the difference between the predicted and actual values, was calculated in all motor ability tests. Other assessments included the number of falls in the previous year, BMI, frequency of going out, Mini-Mental State Examination score, and Falls Efficacy Scale. In the baseline study, we divided the subjects into a fall group (n = 33) and a nonfall group (n = 141) and compared motor ability and EE for the two groups. During a 1-year follow-up, the nonfall group (baseline study) was assessed for the same measurements by using the same methods. Results: In the baseline study, the fall group had significantly lower values of OLS and MSL. Furthermore, the fall group significantly overestimated their OLS, TUG, and MSL. In logistic regression analysis, EE of TUG (OR = 1.27) and EE of MSL (OR = 1.08) were detected as risk factors for falls. During follow-up, 11 subjects (7.8%) experienced falls. In logistic regression analysis, TUG (OR = 1.89) and EE of MSL (OR = 1.06) were detected as significant risk factors for falls. Since EE of MSL had higher values of both the area under the receiver operating characteristic curve and the sum of sensitivity and specificity than EE of TUG, the nonfall group was divided into two groups with a cutoff value of 2 cm for EE of MSL. A significant distribution disparity in falls between the two groups was found during follow-up and showed a relative risk of 18.78 for EE of MSL. Conclusions: We suggest that EE of MSL is a potent predictor for falls among healthy elderly adults. (C) 2014 S. Karger AG, Basel