Longitudinal investigation of eyeblink classical conditioning in elderly human subjects.

Eyeblink classical conditioning (EBCC) is an important tool in the study of learning, memory, and aging, but few longitudinal data have been collected on EBCC in humans of any age. Our aim was to determine if EBCC would remain stable across time. Fifteen subjects with an initial mean age of 83.2 years ("old-old") were tested three times in just over 2 years, and a subset of 8 of these subjects were tested a fourth time after the third year. Fifteen additional subjects with a mean age of 69.1 years ("young-old") were tested two times in just over one year. Subjects were tested with identical procedures in the 400 msec delay EBCC paradigm. Cognitively normal young-old and old-old adults showed stability in EBCC. In the old-old group, poor EBCC (< 25% conditioned responses; CRs) predicted significantly poorer Blessed Information Memory Concentration scores. Old-old subjects with poor initial EBCC were more likely to become demented or die than were old-old subjects producing over 25% CRs.