Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

The Brain Circuitry of Relational Memory Impairment in Schizophrenia

25 Oct 2015

October 26, 2015. People with schizophrenia tend to perform poorly on tests of long-term memory that require establishing relationships between items. New research published in the September issue of JAMA Psychiatry, led by J. Daniel Ragland of the University of California, Davis, reports that impaired relational memory in schizophrenia is accompanied by reduced activation of the dorsolateral prefrontal cortex (DLPFC) and hippocampus.

"[A] picture is emerging with findings supporting that the [DLPFC] is pertinent for relational encoding and that reduced activity in specific regions of the PFC and [medial temporal lobe] may be associated with the selective relational memory impairment observed in schizophrenia," wrote Martin Lepage of McGill University, Montreal, Canada; Colin Hawco of the Centre for Addiction and Mental Health, Toronto, Canada; and Michael Bodnar of McGill in a commentary accompanying the article.

There are several different ways to remember things, Ragland told SRF. For example, you can remember an item by focusing on its specific features, such as color and shape. Alternatively, you can recall by creating relationships between items, such as remembering both eggs and butter as ingredients for cookies. People with schizophrenia have difficulty remembering things from their past, and studies show that this is due to a specific impairment in relational memory (Achim et al., 2003).

The Cognitive Neuroscience Test Reliability and Clinical Applications for Schizophrenia (CNTRACS) consortium previously developed the Relational and Item-Specific Encoding (RISE) paradigm in order to disentangle the contributions of item-specific and relational memory during the encoding and retrieval components of long-term memory in schizophrenia, finding that patients showed the greatest memory deficits following relational encoding, but relatively intact memory following item-specific encoding (Ragland et al., 2012).

"Item-specific and relational memory involve different systems in the brain," Ragland said. The two types of memory are mediated by distinct regions of the PFC and the MTL. Because relational memory is subserved by the DLPFC and hippocampus, the researchers suspected that these brain areas would be preferentially affected during the relational memory portions of the RISE task in patients with schizophrenia. To test their hypothesis, they collected fMRI data from 52 volunteers with schizophrenia and 57 healthy controls at five different sites in the U.S.

The RISE task includes separate encoding and retrieval components, during which

patients alternate between item-specific and relational blocks. For the encoding phase, patients were shown pairs of images of objects and instructed to answer item-specific (Is either object living?) or relational (Can one object fit inside the other?) questions. Similarly, during the recognition phase, patients were asked to make either item-specific (Was this item presented earlier?) or associative (Were the two images paired together earlier?) responses.

Both the schizophrenia and control groups showed better retrieval of items following relational encoding than following item retrieval encoding. Consistent with earlier studies, schizophrenia patients performed worse during relational encoding than healthy controls. Schizophrenia patients also showed impaired associative recognition compared to controls. This was accompanied by reduced activation of the DLPFC in cases relative to controls.

During memory retrieval, hippocampal activity was associated with retrieval success in healthy control patients. Those with schizophrenia correctly recognized fewer images that were relationally encoded than controls. They also displayed reduced hippocampal activation during correctly identified relationally encoded but not item-specific encoded images, compared to controls.

"These fMRI data also show that the relational impairment seems to be specifically due to impaired activity in the DLPFC during encoding and to impaired hippocampal activity during retrieval," Ragland said. These regions have previously been implicated in relational memory, and so the current findings establish the neural construct validity of the RISE paradigm.

Studies measuring the underlying neural architecture of specific memory deficits in schizophrenia are important for several reasons, wrote Lepage, Hawco, and Bodnar: "Considering the significant association between memory impairment and poor outcome in schizophrenia, the behavioral and neurobiological results from these studies provide novel targets for future pharmacological, psychosocial, or neurostimulation interventions."

For example, in thinking about helping patients improve their memory, the current work suggests two strategies, said Ragland. We can encourage patients to use the more intact system, focusing on item features when they're trying to learn new information, he said. Moving forward, we can also concentrate on identifying treatments that improve these relational memory systems, the DLPFC and hippocampus, he added.

One specific next step in this line of research is to try to understand what's happening at the neuron level using electrophysiology and magnetic resonance spectroscopy. The researchers plan to test the hypothesis that alterations in inhibitory neurotransmission lead to neural oscillation defects that underlie these memory deficits. "If we can show this at the neural level, then we have even more motivation to develop treatments for the impairment," said Ragland, suggesting drugs that improve GABA signaling or transcranial alternating current brain stimulation, which could deliver local oscillatory stimulation to specific brain areas.—Allison Marin


Lepage M, Hawco C, Bodnar M. Relational Memory as a Possible Neurocognitive Marker of Schizophrenia. JAMA Psychiatry . 2015 Sep 1 ; 72(9):946-7. Abstract

Ragland JD, Ranganath C, Harms MP, Barch DM, Gold JM, Layher E, Lesh TA, Macdonald AW, Niendam TA, Phillips J, Silverstein SM, Yonelinas AP, Carter CS. Functional and Neuroanatomic Specificity of Episodic Memory Dysfunction in Schizophrenia: A Functional Magnetic Resonance Imaging Study of the Relational and Item-Specific Encoding Task. JAMA Psychiatry . 2015 Sep 1 ; 72(9):909-16. Abstract