The effects of aging in delay and trace human eyeblink conditioning

Classical eyeblink conditioning is a model system for characterizing age-related changes in human learning and memory. In this fundamental form of learning a neutral conditioned stimulus (CS) is paired with a biologically salient, unconditioned stimulus (US). After repeated pairings, the CS alone reliably elicits a conditioned response (CR), indicating that a CS-US association has been formed. In the delay conditioning procedure the CS and US coterminate while in trace conditioning, a silent “trace interval” elapses between the CS offset and US onset. Age-related effects on behavioral performance and neural activity during eyeblink conditioning have been investigated (Solomon, Beal, & Pendlebury, 1988; Woodruff-Pak, 1988). Numerous human studies have consistently demonstrated that older adults showed fewer conditioned responses than younger adults (Bellebaum & Daum, 2004; Finkbiner & Woodruff-Pak, 1991; Knuttinen, Power, Preston, & Disterhoft, 2001; Solomon, Pomerleau, Bennett, James, & Morse, 1989; Woodruff-Pak & Thompson, 1988). These age-related learning deficits are evident in a number of different conditioning paradigms. Single cue delay (Solomon et al., 1989; Woodruff-Pak & Thompson, 1988), single cue trace (Finkbiner & Woodruff-Pak, 1991), differential trace (Knuttinen et al., 2001), and conditional discrimination (Bellebaum & Daum, 2004) studies reported that younger adults produced higher conditioning levels than older adults. In addition to its correlation with conditioning performance, normal aging is also associated with functional and morphological changes within two brain regions that are often targeted during investigations of memory and cognition, including eyeblink conditioning: the hippocampus and cerebellum (Rosenzweig & Barnes, 2003; Solomon et al., 1988; Wilson, Gallagher, Eichenbaum, & Tanila, 2006; Woodruff-Pak, 1988). In the hippocampus, fewer synaptic connections, weakened synaptic plasticity, and a volumetric reduction have been found (Barnes, Rao, & Houston, 2000; Geinisman, de Toledo-Morrell, Morrell, Persina, & Rossi, 1992; Golomb et al., 1993; Raz et al., 2004). Similarly in the cerebellum, Purkinje cell loss, fewer cerebellar cortical synapses, and a reduction in parallel fiber conduction velocity have been found in older animals (Andersen, Gundersen, Pakkenberg, 2003; Glick & Bondareff, 1979; Rogers, Zornetzer, & Bloom, 1981). These age-related changes in the brain are often linked to impairments in cognitive functioning. For example, magnetic resonance imaging (MRI) data showed that cerebellar volume strongly correlated with the production of conditioned eyeblink responses (Woodruff-Pak, Goldenberg, Downey-Lamb, Boyko, & Lemieux, 2000) and the degree of hippocampal atrophy predicted performance on a variety of memory tests (Golomb et al., 1996; Petersen et al., 2000). Further, the hippocampus has been shown to be critical for trace learning (Moyer, Deyo, & Disterhoft, 1990; Solomon, Vander Schaaf, Thompson, & Weisz, 1986) and significant hippocampal activity has also been reported during delay learning (Berger, Alger, & Thompson, 1976). Functional MRI (fMRI) results from a recent eyeblink conditioning study (Cheng, Disterhoft, Power, Ellis, & Desmond, 2008) showed trace trials elicited significantly greater human hippocampal responding than delay trials, suggesting that this region is critically relied upon during the associative processes mediating trace conditioning. One possible reason for the conditioning deficits typically seen in older adults may be that both the cerebellum and hippocampus experience significant changes across the lifespan. The present study is a behavioral investigation of whether older adults can demonstrate both delay and trace conditioning within the same acquisition period. Younger adults have shown significant learning during a conditioning design in which both delay and trace trials are presented within the same acquisition session (Cheng et al., 2008; Knight, Cheng, Smith, Stein, & Helmstetter, 2004) while older adults have only been tested in their capacity to learn either delay or trace conditioning separately (Finkbiner & Woodruff-Pak, 1991; Solomon et al., 1989; Knuttinen et al.. 2001; Woodruff-Pak & Thompson, 1988). Presenting two different conditioning trial types during acquisition introduces a modest level of complexity not present in single-cue experiments. Behavioral CRs from these previous studies (Cheng et al., 2008; Knight et al., 2004) suggest that delay and trace conditioning was possible using this within-subjects design in younger adults, but given the increased design complexity and the effects of aging on the hippocampus and cerebellum, older adults may have difficulty demonstrating significant learning under such conditions.