The neural correlates of implicit sequence learning in schizophrenia

Twenty-seven schizophrenia spectrum patients and 25 healthy controls performed a probabilistic version of the serial reaction time task (SRT) that included sequence trials embedded within random trials. Patients showed diminished, yet measurable, sequence learning. Postexperimental analyses revealed that a group of patients performed above chance when generating short spans of the sequence. This high-generation group showed SRT learning that was similar in magnitude to that of controls. Their learning was evident from the very 1st block; however, unlike controls, learning did not develop further with continued testing. A subset of 12 patients and 11 controls performed the SRT in conjunction with positron emission tomography. High-generation performance, which corresponded to SRT learning in patients, correlated to activity in the premotor cortex and parahippocampus. These areas have been associated with stimulus-driven visuospatial processing. Taken together, these results suggest that a subset of patients who showed moderate success on the SRT used an explicit stimulus-driven strategy to process the sequential stimuli. This adaptive strategy facilitated sequence learning but may have interfered with conventional implicit learning of the overall stimulus pattern.     

Impaired implicit memory for gist information in amnesia

In 2 experiments using a converging associates paradigm, the authors evaluated implicit memory for gist information in amnesic patients. In Experiment 1, participants saw multiple sets of associates, each converging on a nonpresented theme word, and were then tested using an implicit word stem completion test and an explicit cued recall test. Amnesic patients showed intact implicit and impaired explicit memory for studied words, but memory for nonpresented lures was impaired, regardless of retrieval instructions. To evaluate whether impaired implicit memory for lures was due to accelerated forgetting of gist information, short study lists were used in Experiment 2, each consisting of a single set of associates. Amnesics' implicit memory for lures was again impaired. These results point to an inability to encode robust gist representations as the cause of impaired gist memory in amnesia.     

Intact implicit habit learning in Alzheimer's disease

Habit learning refers to the incremental implicit learning of associations. Patients with Alzheimer's disease (AD) exhibit deficits in explicit memory and in conceptual implicit memory tasks that rely on the cortical areas damaged in AD. The authors tested patients with AD and controls on a probabilistic classification task in which participants implicitly acquire cue-outcome associations. Both groups showed evidence of learning across 50 trials, and performance did not differ significantly between the groups. In contrast, patients with AD exhibited a profound impairment in explicit memory for the testing episode. These results are consistent with the idea that habit learning relies on subcortical structures, including the basal ganglia, and is independent of the medial temporal and cortical areas damaged in AD.     

Impaired updating of working memory in schizophrenia.

This study investigated working memory in schizophrenia, using an auditory target detection task specifically designed to separate out brain activity related to the updating of working memory with new information from activity related to target detection and response. Event related potentials (ERPs) were recorded during task performance, using 31 electrodes, from 25 subjects with schizophrenia and 25 matched controls. Subjects with schizophrenia had a reduction in parietal P3 and Late Slow Wave amplitude in the ERP waveforms recorded when the task required subjects to remember a new stimulus. This P3 amplitude attenuation correlated with symptom measures of preoccupation and poor volition. Previous findings of a reduction in P3 amplitude during target detection by subjects with schizophrenia were replicated. These results suggest that there is a specific impairment in the ability to update working memory in schizophrenia, and that this is associated with poverty of engagement with the environment.     

Impaired higher order implicit sequence learning on the verbal version of the serial reaction time task in patients with Parkinson's disease

Although neuroimaging studies have strongly implicated basal ganglia involvement in implicit sequence learning, serial reaction time (SRT) studies with Parkinson's disease (PD) patients have yielded mixed results. The present research sought to examine the ability of people with PD to implicitly learn sequences with different sequential structures and to objectively assess explicit knowledge. A version of the SRT task that reduces motor demands was used to compare 19 patients with PD but not dementia and 37 matched controls. PD patients showed less implicit sequence-specific learning for both sequences and reduced response time improvement over sequential trials for the more complex sequence. A closer examination revealed that the deficit involved higher order sequential associations as well as the learning of pairwise information.     

Developmental differences in implicit learning of spatial context

The authors examined whether a form of implicit memory that has been unambiguously dissociated from conscious awareness--learning of spatial context on the contextual cuing task introduced by M. M. Chun and Y. Jiang (1998)--is mature in childhood as predicted by an evolutionary view of cognition. School-aged children did not show reliable learning relative to adults who performed the same version of the task or another version that slowed responses to match those of children. Thus, unreliable learning in childhood was mediated by immature implicit representations of spatial context rather than by slower baseline response speed. The present finding is inconsistent with the prediction of the evolutionary view of cognition but consistent with incomplete maturation of medial temporal lobes known to mediate contextual learning.     

Perceptually based implicit learning in severe closed-head injury patients

This study suggests that perceptually based implicit learning may either be preserved following a severe closed-head injury (CHI) or recover within the 1st year. Nineteen severe CHI patients and 19 controls searched visual matrices and indicated the quadrant location of a target. Participants were exposed to the following covariation pattern: AAAABAAA. For Covariation A blocks, the matrices systematically co-occurred with a unique location of the target. This relationship was altered for the B block. Despite CHI participants' overall slower response times (RTs), both groups demonstrated the expected decline in RTs across the first 4 Covariation A blocks followed by an increase when the covariation changed. Both groups also exhibited retention of their learning after a 20-min delay. Explicit knowledge tests indicated that participants lacked awareness for the covariation.     

Intact implicit probabilistic sequence learning in obstructive sleep apnea

Obstructive sleep apnea (OSA) belongs to the sleep-related breathing disorders and is associated with cognitive impairments in learning and memory functions. The impairments in attention-demanding cognitive functions such as working memory and executive functions are well established in OSA; however, it remains unknown if less attention-demanding implicit sequence learning is affected. In the present study, we examined implicit sequence learning in OSA to probe the functional integrity of this fundamental learning mechanism. We used listening span to measure complex working memory capacity and the alternating serial reaction time (ASRT) task, which enables us to measure general skill learning and sequence-specific learning separately. Twenty OSA patients and 20 healthy controls participated in this study. Our data show dissociation between working memory and implicit sequence learning in OSA. Surprisingly, OSA patients showed preserved general skill and sequence-specific learning in spite of the possible hypoxia and sleep restriction. In contrast, working memory performance measured by listening span task was impaired in the OSA group. This finding suggests selective susceptibility of more attention-demanding cognitive functions in this patient population, while implicit learning remains intact. Our findings draw attention the fact that disordered sleep may have less impact on the integrity of structures connected to implicit sequence learning.     

Impaired prefrontal inhibition in schizophrenia: relevance for cognitive dysfunction

Male adult rats that received an intragastric load of 2 ml of 12% NaCl (n=13) ingested both water (4.0+/-0.2 ml/2 h) and 0.9% NaCl (3.7+/-1.0 ml/2 h) when compared with rats that received intragastric load of 2 ml of water (water: 0.1+/-0.1; 0.9% NaCl: 0.5+/-0.3 ml/2 h, n=12) in a two-bottle test. Intragastric sodium load increased plasma sodium concentration and osmolality by 5% and reduced plasma renin activity by half compared to rats that received intragastric load of water. Intravenous infusion of 1.5 ml/10 min of 10% NaCl (n=16) also induced ingestion of water (6.2+/-0.8 ml/2 h) and 0.9% NaCl (2.9+/-0.8 ml/2 h) compared with intravenous infusion of 1.5 ml/10 min of 0.9% NaCl (water: 0.9+/-0.4; 0.9% NaCl: 0.5+/-0.2 ml/2 h, n=14). Therefore, a sodium load that raises natremia and plasma osmolality, and therefore induces cell dehydration, results in both 0.9% NaCl and water ingestion when the rats have a two-bottle choice.     

Impaired recruitment of the hippocampus during conscious recollection in schizophrenia

Poor attention and impaired memory are enduring and core features of schizophrenia. These impairments have been attributed either to global cortical dysfunction or to perturbations of specific components associated with the dorsolateral prefrontal cortex (DLPFC), hippocampus and cerebellum. Here, we used positron emission tomography (PET) to dissociate activations in DLPFC and hippocampus during verbal episodic memory retrieval. We found reduced hippocampal activation during conscious recollection of studied words, but robust activation of the DLPFC during the effort to retrieve poorly encoded material in schizophrenic patients. This finding provides the first evidence of hippocampal dysfunction during episodic memory retrieval in schizophrenia.     

Frames of reference during implicit and explicit learning

There is a significant overlap between the processes and neural substrates of spatial cognition and those subserving memory and learning. However, for procedural learning, which often is spatial in nature, we do not know how different forms of spatial knowledge, such as egocentric and allocentric frames of reference, are utilized nor whether these frames are differentially engaged during implicit and explicit processes. To address this issue, we trained human subjects on a movement sequence presented on a bi-dimensional (2D) geometric frame. We then systematically manipulated the geometric frame (allocentric) or the sequence of movements (egocentric) or both, and retested the subjects on their ability to transfer the sequence knowledge they had acquired in training and also determined whether the subjects had learned the sequence implicitly or explicitly. None of the subjects (implicit or explicit) showed evidence of transfer when both frames of reference were changed which suggests that spatial information is essential. Both implicit and explicit subjects transferred when the egocentric frame was maintained indicating that this representation is common to both processes. Finally, explicit subjects were also able to benefit from the allocentric frame in transfer, which suggests that explicit procedural knowledge may have two tiers comprising egocentric and allocentric representations.     

An ERP analysis of implicit structured sequence learning

When task exposure facilitates performance without producing corresponding changes in verbalizable knowledge, learning is said to be implicit. In Experiment I, event-related potentials (ERPs) were recorded as individuals practiced an implicit structured sequence learning (ISSL) task wherein only some target events required a response. With practice, the ERPs to targets that obeyed the underlying grammar diverged from those that did not at around 200 ms; grammatical targets appeared to be more positive between 200 and 500 ms because a similar positivity for the ungrammatical targets was delayed. In Experiment 2, the grammar was simplified allowing a direct comparison to be made between an implicit learning group and an explicit group, who were taught the grammar prior to recording. The results of the comparison revealed a remarkable similarity but did implicate at least partially nonidentical neural mechanisms in implicit and explicit strured sequence learning.     

Impaired heat discrimination learning after capsaicin treatment

A heat discrimination experiment has been designed with the intention of deciding whether rats treated with capsaicin are insensitive to heat or their thermolytic reactions are impaired. The animals were trained in a T-maze on an 11°C discrimination. The capsaicin-treated rats were slow in learning the task, and their running time was significantly longer than that of the controls even when they had mastered the problem. No such differences were found in the visual discrimination experiment in the same T-maze. It is suggested that the deficiency of the thermolytic reactions in the capsaicin treated rats may be due to an impaired heat sensitivity.     

Neural correlates of encoding and expression in implicit sequence learning

It has been shown that high-frequency repetitive transcranial magnetic stimulation (rTMS) to the human primary motor hand area (M1-HAND) can induce a lasting increase in corticospinal excitability. Here we recorded motor evoked potentials (MEPs) from the right first dorsal interosseus muscle to investigate how sub-threshold high-frequency rTMS to the M1-HAND modulates cortical and spinal excitability. In a first experiment, we gave 1500 stimuli of 5 Hz rTMS. At an intensity of 90% of active motor threshold, rTMS produced no effect on MEP amplitude at rest. Increasing the intensity to 90% of resting motor threshold (RMT), rTMS produced an increase in MEP amplitude. This facilitatory effect gradually built up during the course of rTMS, reaching significance after the administration of 900 stimuli. In a second experiment, MEPs were elicited during tonic contraction using weak anodal electrical or magnetic test stimuli. 1500 (but not 600) conditioning stimuli at 90% of RMT induced a facilitation of MEPs in the contracting FDI muscle. In a third experiment, 600 conditioning stimuli were given at 90% of RMT to the M1-HAND. Using two well-established conditioning-test paradigms, we found a decrease in short-latency intracortical inhibition (SICI), and a facilitation of the first peak of facilitatory I-waves interaction (SICF). There was no correlation between the relative changes in SICI and SICF. These results demonstrate that subthreshold 5 Hz rTMS can induce lasting changes in specific neuronal subpopulations in the human corticospinal motor system, depending on the intensity and duration of rTMS. Short 5 Hz rTMS (600 stimuli) at 90% of RMT can selectively shape the excitability of distinct intracortical circuits, whereas prolonged 5 Hz rTMS (900 stimuli) provokes an overall increase in excitability of the corticospinal output system, including spinal motoneurones.     

Implicit learning of visuospatial sequences in schizophrenia

The authors examined whether patients with schizophrenia learned sequential patterns in a probabilistic serial response time task in which pattern trials alternated with random ones. Patients showed faster and more accurate responses to pattern trials than to random trials, but controls showed greater sensitivity to patterns. The highest level of regularity learned in both groups was information about runs of 3 events. Pattern learning occurred largely outside of awareness, as participants could not describe patterns. Controls with higher memory spans learned the sequential pattern better than those with lower memory spans, suggesting that working memory influences implicit pattern learning. Pathology in motor sequencing systems and poor working memory may lead to deficits in learning sequence structure in schizophrenia.     

Impaired conditioned taste aversion learning in APP transgenic mice

Cognition in transgenic mouse models of Alzheimer's disease (AD) has been predominantly characterized in explicit spatial orientation tasks. However, dementia in AD encompasses also implicit memory systems. In the present study a line of transgenic mice (TgCRND8) encoding a double mutated allele of the human amyloid precursor protein (APP) genes was evaluated in an implicit associative learning task of conditioned taste aversion (CTA). CTA is a form of Pavlovian classical conditioning, in which a mouse learns to avoid a novel taste of saccharine (conditioned stimulus) paired with an experimentally induced (systemic injection of lithium chloride) nausea (unconditioned stimulus). In contrast to conditioned non-Tg mice, TgCRND8 APP mice developed weaker aversion against saccharine and quickly increased its consumption in repeated tests. These results indicate that TgCRND8 mice show a significant impairment not only in explicit spatial memory, as has been previously shown [Nature 408 (2000) 979], but also in implicit memory. Control experiments confirmed that TgCRND8 and non-Tg mice had comparable taste sensitivities in response to appetitive as well as aversive tastes. The study suggests that the CTA paradigm can be a sensitive tool to evaluate deficits in implicit associative learning in APP transgenic mouse models of AD.