ABSTRACT
Can Alzheimer's disease (AD) patients efficiently learn to perform a complex motor skill when relying on procedural knowledge? To address this question, the authors compared the golf-putting performance of AD patients, older adults, and younger adults in 2 different learning situations: one that promotes high error rates (thus increasing the reliance on declarative knowledge) or one that promotes low error rates (thus increasing the reliance on procedural knowledge). Motor performance was poorer overall for AD patients and older adults relative to younger adults in the high-error condition but equivalent between similar groups in the low-error condition. Also, AD patients in the low-error condition had better performance at the final putting distance relative to those in the high-error condition. This performance facilitation for AD patients likely stems from intact procedural knowledge.
Notes
1. AD patients assigned to the low-error condition and those assigned to the high-error condition did not differ on demographic scores (age, education), motor scores (Timed Up and Go test), and on cognitive scores (Mini-Mental State Examination, MARIN apathy scale, Geriatric Depression Scale, Instrumental Activities of Daily Living, Stroop test, and Trail Making test).
2. Initially, our experimental design required that half of the participants perform a secondary tone-counting task while putting at the last distance of 125 cm. By doing so, our goal was to assess whether knowledge representations were attention demanding (thus suggesting the use of declarative knowledge) or attention free (thus suggesting the use of procedural knowledge). But we do not report these preliminary data because AD patients were unable to perform the tone-counting task (even by itself in isolation).
3. Note that the correlational analyses showed uneven degrees of freedom because some neuropsychological tests could not be completed by each AD patients (due to a lack of available time). It is also worth pointing out that no other correlation effects were found between Victoria Stroop test part 3 (inhibition processes) and learning performance (through the five blocks) for young and older adults in both learning conditions. For information purposes, in the Victoria Stroop test part 3, the following nonsignificant correlations were found: for young adults, r(12) = .363, p = .911; and for older adults in low-error conditions r(12) = .446, p = .440; for young adults, r(12) = .266, p = .403; and for older adults in high-error conditions, r(12) = .127, p = .695.