30 June 2010. Theories about the neurobiology underlying cognitive deficits in schizophrenia lead to inaccurate predictions about working memory impairments in the disorder, says a study in the June Archives of General Psychiatry. James Gold of the University of Maryland School of Medicine in Baltimore and colleagues questioned whether, as some theories suggest, working memory would be unstable and imprecise in schizophrenia. Using novel methods that enabled them to tease apart different aspects of working memory, they found no evidence that people with schizophrenia have more fleeting or off-the-mark memories than healthy subjects. Rather, they are unable to simultaneously hold as many items at once in working memory. In light of these findings, Gold et al. recommend a trip back to the drawing board to revise current theories of the neurobiology of working memory deficits in schizophrenia.
Schizophrenia-related cognitive deficits hamper patients’ lives (see Green, 1996), and a recent meta-analysis found consistent evidence of large working memory deficits in subjects with schizophrenia as compared to healthy subjects (Forbes et al., 2009). Working memory, the ability to briefly store and manipulate information, guides goal-directed behavior through other cognitive processes.
Several researchers have tried to explain the biological basis of cognitive symptoms in schizophrenia. For instance, John Lisman and colleagues (Lisman et al., 2008) suggest that reduced N-methyl-D-aspartate (NMDA) channel function impairs memory in schizophrenia by disinhibiting pyramidal cells in the hippocampus, thereby reducing gamma waves (see SRF related news story). Edmund Rolls and associates (Rolls et al., 2008) suggest that reduced dopamine in schizophrenia decreases NMDA currents, causing neural networks to randomly fire, adding noise that drowns out information-carrying signals, rendering the networks unstable. Daniel Durstewitz and Jeremy Seamans (2008) propose that imbalanced activation of dopamine D1 and D2 receptors in the prefrontal cortex may result in excess noise and overly frail representations in memory. These theories could lead one to expect imprecise or unstable working memory in schizophrenia.
Yet, when Gold and colleagues looked at previous findings regarding working memory in schizophrenia, they found reason to doubt that schizophrenia causes faster-decaying or less accurate memories. Members of the research team, Wei Zhang and Steven Luck, both of the University of California at Davis, had recently devised a way to separately examine the number of representations in memory, the precision of those memories, and their stability over time. This provided an opportunity for the research group to test current theories of deficits in the working memory of individuals with schizophrenia.
The researchers’ approach involved a case-control design in which the researchers tested the working memory of 31 clinically stable patients who met criteria for schizophrenia or schizoaffective disorder. They compared them with 26 mentally healthy subjects who had no history of psychosis. Controls mirrored the patients in age, sex, ethnicity, and parental education. All participants were presented with three or four different colors on a computer screen. After a pause when the screen went blank, subjects were to indicate the color shown in a particular spot by selecting and clicking on it on a color wheel. Subjects who stored the color in memory and recalled it when tested should select colors similar to those actually shown. Those who did not would have to guess.
By examining the distribution of errors, Gold and colleagues determined the probability that subjects held an item in memory at test time and the precision of that representation. To check the stability of working memory representations, they tried both a one-second and a four-second delay. They reasoned that if patients with schizophrenia have unstable memories, they should perform worse than control subjects after a short delay.
A mixed bag
The results suggest that schizophrenia reduces working memory capacity, causing subjects with schizophrenia to store fewer items. Even so, patients recalled items that had been stored in memory with the same precision as healthy subjects. Furthermore, the length of delay made no difference in either the number of items recalled or the precision of recall for either group, contrary to expectations of less stable memories in schizophrenia. “In our view, the recent biological accounts discussed above are at odds with much of the behavioral literature, and clearly at odds with the data presented here,” write Gold and associates.
While Gold et al. acknowledge that a longer delay might bring out unseen differences between the two groups of subjects, they think that a four-second delay should be sufficient for detecting the disruptive working memory deficits expected in schizophrenia. They cannot explain why the working memory of subjects with schizophrenia would hold fewer items, although they note that neuroimaging studies point to the parietal cortex, perhaps in league with the prefrontal cortex, in setting capacity for visual working memory. They write, “Unfortunately, there is very little understanding of the origins of capacity limits in the basic cognitive neuroscience literature.”
All of the patients in the study were undergoing treatment with antipsychotic medication, including clozapine in 18 cases, suggesting that other treatments had failed them (see SRF related news story). This indicates that clozapine may boost the signal-to-noise ratio (see SRF related news story). Gold and colleagues warn that untreated patients in the early stages of schizophrenia might show a different pattern of deficits. They also caution that the results might not extend to other working memory tasks that activate different neural pathways and/or do not involve color judgments.
Despite these findings, or maybe because of them, Gold and colleagues see “a great need” for models that explain the actual working memory deficits seen in schizophrenia. Even so, they write, “these models must accurately capture the behavioral endpoint, which is characterized primarily by reductions in storage capacity and not by an instability of the working memory representations.”—Victoria L. Wilcox.
Gold JM, Hahn B, Zhang WW, Robinson BM, Kappenman ES, Beck VM, Luck SJ. Reduced capacity but spared precision and maintenance of working memory representations in schizophrenia. Arch Gen Psychiatry. 2010 Jun;67(6):570-7. Abstract