Cognitive function as a prospective predictor of falls.

OBJECTIVE This study examined speed of processing, executive functioning, and psychomotor speed as independent prospective predictors of falls and recurrent falls across 3 years. METHOD The participants were 509 community-dwelling older adults. Measures of speed of processing, executive function, psychomotor speed, and known risk factors of falling were included in correlation and logistic regression analyses. RESULTS Poor executive function (Trail Making Test), slower speed of processing (Digit Symbol Substitution [DSS] Test), and slower psychomotor speed (Digit Symbol Copy Test) were significantly associated with falls. Poor executive function and speed of processing performance (Stroop Test, Trail Making Test, and DSS Test) as well as slower psychomotor speed were significantly related to recurrent falls. Logistic regression results indicated that only medication use, far visual acuity, and psychomotor speed were significant independent predictors of falls. Regarding recurrent falls, being white, medication use, and balance were significant predictors. DISCUSSION Although cognitive measures at baseline were significantly associated with falls and recurrent falls at follow-up, these measures did not predict falling after considering known risk factors of falls and psychomotor speed. Thus, it may be that simple measures of psychomotor speed are more salient predictors of falls than cognitive measures.