fully loaded
fully loaded
Interference of working memory load with long-term memory formation
Nikolai Axmacher 1,2 , Sven Haupt 1 , Michael X Cohen 1,2 , Christian E. Elger 1,2 and Juergen Fell 1
1 Department of Epileptology, University of Bonn, Sigmund-Freud-Str. 25, 53105 Bonn, Germany
2 Life and Brain Center of Academic Research, Bonn, Germany
Correspondence to Dr N. Axmacher, 1Department of Epileptology, as above.
E-mail: nikolai.axmacher@ukb.uni-bonn.de
Copyright Journal compilation © 2009 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
KEYWORDS
fMRI • long-term memory • medial temporal lobe • parahippocampal cortex • working memory
ABSTRACT
Traditionally, it has been assumed that the medial temporal lobe (MTL) is indispensable for long-term memory (LTM) encoding, but only plays a minor role for working memory (WM) maintenance. Recently, however, an increasing number of studies questioned this seemingly clear distinction by showing that the MTL does participate in some WM processes, especially if multiple items are being maintained. This would predict that WM maintenance of multiple items interferes with simultaneous LTM encoding. Here, we tested this idea in a functional magnetic resonance imaging paradigm that required subjects to encode stimuli into LTM during simultaneous WM maintenance of either single or multiple items. Indeed, we found that maintenance of multiple items deteriorates simultaneous LTM encoding as compared with maintenance of single items. WM-related activation of the hippocampus was more pronounced in the condition with high WM load; in contrast, hippocampal activation related to LTM encoding was stronger in the low WM load condition. Successful LTM encoding was associated with a high level of activity in the adjacent parahippocampal cortex (PHC), leading to pronounced parahippocampal subsequent memory effects in the high load condition. This suggests that the PHC is a locus of WM–LTM interaction. Functional connectivity analysis with a seed in the PHC confirmed this result by revealing strong connectivity with the medial frontal cortex, which was only active in the high WM load condition. Taken together, these findings suggest that high WM demands interfere with LTM encoding and thus support the idea that WM and LTM processes interact in the MTL.
training -> (+)WMC -> (-)WM load @ specified task difficulty -> +LTP
training likely either improves extrahippocampal WM processing efficiency thereby sparing hippocampal resources, improves intrahippocampal WM processing efficiency thereby increasing hippocampal resources (WM training has been shown to reduce hippocampal neurogenesis in animals), or both
and bingo was his name-o...
Nikolai Axmacher 1,2 , Sven Haupt 1 , Michael X Cohen 1,2 , Christian E. Elger 1,2 and Juergen Fell 1
1 Department of Epileptology, University of Bonn, Sigmund-Freud-Str. 25, 53105 Bonn, Germany
2 Life and Brain Center of Academic Research, Bonn, Germany
Correspondence to Dr N. Axmacher, 1Department of Epileptology, as above.
E-mail: nikolai.axmacher@ukb.uni-bonn.de
Copyright Journal compilation © 2009 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
KEYWORDS
fMRI • long-term memory • medial temporal lobe • parahippocampal cortex • working memory
ABSTRACT
Traditionally, it has been assumed that the medial temporal lobe (MTL) is indispensable for long-term memory (LTM) encoding, but only plays a minor role for working memory (WM) maintenance. Recently, however, an increasing number of studies questioned this seemingly clear distinction by showing that the MTL does participate in some WM processes, especially if multiple items are being maintained. This would predict that WM maintenance of multiple items interferes with simultaneous LTM encoding. Here, we tested this idea in a functional magnetic resonance imaging paradigm that required subjects to encode stimuli into LTM during simultaneous WM maintenance of either single or multiple items. Indeed, we found that maintenance of multiple items deteriorates simultaneous LTM encoding as compared with maintenance of single items. WM-related activation of the hippocampus was more pronounced in the condition with high WM load; in contrast, hippocampal activation related to LTM encoding was stronger in the low WM load condition. Successful LTM encoding was associated with a high level of activity in the adjacent parahippocampal cortex (PHC), leading to pronounced parahippocampal subsequent memory effects in the high load condition. This suggests that the PHC is a locus of WM–LTM interaction. Functional connectivity analysis with a seed in the PHC confirmed this result by revealing strong connectivity with the medial frontal cortex, which was only active in the high WM load condition. Taken together, these findings suggest that high WM demands interfere with LTM encoding and thus support the idea that WM and LTM processes interact in the MTL.
training -> (+)WMC -> (-)WM load @ specified task difficulty -> +LTP
training likely either improves extrahippocampal WM processing efficiency thereby sparing hippocampal resources, improves intrahippocampal WM processing efficiency thereby increasing hippocampal resources (WM training has been shown to reduce hippocampal neurogenesis in animals), or both
and bingo was his name-o...
medicalstudent | 3 years ago
Reply
Makes sense, if the brain is essentially a memory organ. It is reducible to one element: memory. Long- and short-term being extrapolations of this general faculty.
? | 3 years ago
Reply
medicalstudent | 3 years ago
Reply
Since the evidence above is undeniable and supports that there is a strong connection between the two, I would assume that the two evolved together. Further specification with respect to how WM interacts with LM will likely grant a few clues about what this evolutionary process may have looked like and what it will become. More thought about this may develop possible scenarios.
? | 3 years ago
Reply
