Professional Documents
Culture Documents
Memory Dysfunction
Memory Dysfunction
review article
current concepts
Memory Dysfunction
Andrew E. Budson, M.D., and Bruce H. Price, M.D.
episodic memory
Episodic memory refers to the explicit and declarative memory system used to recall per-
sonal experiences framed in our own context, such as a short story or what you had for
dinner last night. Episodic memory has largely been defined according to the inability
of people with amnesia due to lesions of the medial temporal lobe to remember expe-
riences that healthy people can remember. Thus, this memory system depends on the
medial temporal lobes (including the hippocampus and the entorhinal and perirhinal
cortexes). Other critical structures in the episodic memory system (some of which are
associated with a circuit described by Papez in the right medial temporal and right frontal lobes are
19378) include the basal forebrain with the medial most active when learning visual scenes.7
septum and diagonal band of Broca’s area, the ret- One reason that the frontal lobes are important
rosplenial cortex, the presubiculum, the fornix, for encoding is that they permit the person to focus
mammillary bodies, the mammillothalamic tract, on the information to be remembered and to en-
and the anterior nucleus of the thalamus.2 A lesion gage the medial temporal lobes. Dysfunction of the
in any one of these structures may cause the im- frontal lobes may cause distortions of episodic
pairment that is characteristic of dysfunction of the memory as well as false memories, such as infor-
episodic memory system (Fig. 1). mation that is associated with the wrong context14
Memory loss attributable to dysfunction of the or with incorrect specific details.15 Extreme exam-
episodic memory system follows a predictable pat- ples of memory distortions include confabulation,
tern known as Ribot’s law, which states that events which occurs when “memories” are created to be
just before an ictus are most vulnerable to dissolu- consistent with current information,14 such as “re-
tion, whereas remote memories are most resistant. membering” that someone broke into the house
Thus, in cases of dysfunction of the episodic mem- and rearranged household items.
ory system, the ability to learn new information is These differences between deficits in episodic
impaired (anterograde amnesia), recently learned memory that occur because of damage to the medi-
information cannot be retrieved (retrograde am- al temporal lobes (and the Papez circuit) and those
nesia), and remotely learned information is usually that occur because of damage to the frontal lobes
spared.9 can be conceptualized in an oversimplified but clin-
Studies have shown that the episodic memory ically useful analogy.16 The frontal lobes are analo-
system includes the frontal lobes.5,10 Rather than gous to the “file clerk” of the episodic memory sys-
being responsible for the retention of information, tem, the medial temporal lobes to the “recent
the frontal lobes are involved in the registration, ac- memory file cabinet,” and other cortical regions to
quisition, or encoding of information6; the retrieval the “remote memory file cabinet.” Thus, if the
of information without contextual and other cues11; frontal lobes are impaired, it is difficult — but not
the recollection of the source of information12; and impossible — to get information in and out of stor-
the assessment of the temporal sequence and re- age. However, the information may be distorted ow-
cency of events.13 Studies have also shown that the ing to “improper filing” that leads to an inaccurate
left medial temporal and left frontal lobes are most source, context, or sequence. If, however, the medi-
active when a person is learning words,6 whereas al temporal lobes are rendered completely dysfunc-
Anterior thalamic
nucleus Fornix
Thalamus
Mammillothalamic tract
Mammillary body
Parahippocampal
Amygdala gyrus
Hippocampal
formation
tional, it will be impossible for recent information mentia and multiple sclerosis, progress in a step-
to be retained. Older information that has been con- wise manner. Other disorders of memory, such as
solidated over a period of months or years is thought those due to medications, hypoglycemia, tumors,
to be stored in other cortical regions and will there- and Korsakoff ’s syndrome, can have a more com-
fore be available even when medial temporal lobes plicated and variable time course.
and the Papez circuit are damaged. For example, Once a disorder of episodic memory is suspect-
although patients with depression and those with ed on the basis of a reported inability to remember
Alzheimer’s disease may exhibit episodic memory recent information and experiences accurately, ad-
dysfunction, the former have a dysfunctional “file ditional evaluation is warranted. A detailed history
clerk” and the latter have a dysfunctional “recent should be taken, with particular emphasis on the
memory file cabinet.” time course of the memory disorder. Interviewing a
Disorders of episodic memory may be transient, caregiver or other informant is usually critical for
such as those attributable to a concussion, a seizure, accuracy, since the patient will invariably not re-
or transient global amnesia. Static disorders, such member important aspects of the history. A history
as traumatic brain injury, hypoxic or ischemic injury, of other cognitive deficits (e.g., attention, language,
single strokes, surgical lesions, and encephalitis, visuospatial, and executive) should be elicited. A
typically are maximal at onset (or for several days), medical and neurologic examination should be per-
improve (sometimes over periods of two years or formed, with a focus on searching for signs of sys-
more), and then are stable. Degenerative diseases, temic illness, focal neurologic injury, and neurode-
including Alzheimer’s disease,17 dementia with generative disorders.
Lewy bodies, and frontotemporal dementia, begin Cursory cognitive testing may be performed by
insidiously and progress gradually. Disorders af- asking the patient to remember a short story or sev-
fecting multiple brain regions, such as vascular de- eral words, or with the use of instruments such as
sions in the basal ganglia or cerebellum who show has traditionally been divided into components that
impairment in learning procedural skills.40 Because process phonologic information (e.g., keeping a
the disease process in early Alzheimer’s disease af- phone number “in your head”) or spatial informa-
fects cortical and limbic structures while sparing tion (e.g., mentally following a route) and an exec-
the basal ganglia and cerebellum, these patients utive system that allocates attentional resources.44
show deficits in episodic memory but normal ac- Numerous studies have shown that working
quisition and maintenance of procedural skills. memory uses a network of cortical and subcortical
Parkinson’s disease is the most common disor- areas, depending on the particular task.45 However,
der affecting procedural memory. Other neurode- virtually all tasks involving working memory require
generative diseases that disrupt procedural memory participation of the prefrontal cortex (Fig. 2).5 Typ-
include Huntington’s disease and olivopontocere- ically, the network of cortical and subcortical areas
bellar degeneration. Patients in the early stages of includes posterior brain regions (e.g., visual-asso-
these disorders perform nearly normally on episod- ciation areas) that are linked with prefrontal regions
ic memory tests but show an impaired ability to to form a circuit. Studies have shown that phono-
learn skills.38,41 Tumors, strokes, hemorrhages, logic working memory tends to involve more re-
and other causes of damage to the basal ganglia or gions on the left side of the brain, whereas spatial
cerebellum may also disrupt procedural memory. working memory tends to involve more regions on
Patients with major depression have also been the right side.5 Studies have also shown that more
shown to have impairment in procedural memory, difficult tasks involving working memory require
perhaps because depression may involve dysfunc- bilateral brain activation, regardless of the nature
tion of the basal ganglia (Table 2).42 of the material being manipulated.46 Furthermore,
Disruption of procedural memory should be sus- there is an increase in the number of activated brain
pected when patients show evidence of either the regions in the prefrontal cortex as the complexity of
loss of previously learned skills or substantial im- the task increases.47
pairment in learning new skills. For example, pa- Because working memory depends on a network
tients may lose the ability to perform automatic, of activity that includes subcortical structures as
skilled movements, such as writing, playing a mu- well as frontal and parietal cortical regions, many
sical instrument, or swinging a golf club. Although neurodegenerative diseases impair working-mem-
they may be able to relearn the fundamentals of ory tasks. Studies have shown that patients with
these skills, explicit thinking is often required for Alzheimer’s, Parkinson’s, or Huntington’s disease
their performance. As a result, patients with damage or dementia with Lewy bodies, as well as less com-
to the procedural memory system may never achieve mon disorders such as progressive supranuclear
the automatic effortlessness of simple motor tasks palsy, may show impaired working memory (Table
that healthy people take for granted. 2).48,49 In addition to neurodegenerative diseases,
Evaluation of disorders of procedural memory almost any disease process that disrupts the frontal
is similar to that of disorders of episodic memory; lobes or their connections to posterior cortical re-
treatment of the underlying cause depends on the gions and subcortical structures can interfere with
specific disease process. It is worth noting that pa- working memory. Such processes include strokes,
tients whose episodic memory has been devastated tumors, head injury, and multiple sclerosis, among
by encephalitis, for example, have had success in re- others.50,51 Because phonologic working memory
habilitation by using the procedural memory sys- involves the silent rehearsal of verbal information,
tem to learn new skills.43 almost any kind of aphasia can also impair it. Al-
though the pathophysiology is not well understood,
working memory disorders that diminish attentional resources, such
as attention deficit–hyperactivity disorder, obses-
Working memory is a combination of the tradition- sive–compulsive disorder, schizophrenia, and de-
al fields of attention, concentration, and short-term pression, can also impair working memory.52-54
memory. It refers to the ability to temporarily main- A disorder of working memory can present in
tain and manipulate information that one needs to several ways. Most commonly, the patient will show
keep in mind. Because it requires active and con- an inability to concentrate or pay attention. Diffi-
scious participation, working memory is an explicit culty performing a new task involving multistep in-
and declarative memory system. Working memory structions may be seen. A disorder of working mem-
ory may also present as a problem with episodic duced converging and complementary lines of evi-
memory. In such cases, the evaluation will show a dence, suggesting that memory is composed of sep-
primary failure of encoding, because in order to arate and distinct systems. A single disease process
transfer information into episodic memory, the in- (such as Alzheimer’s disease) may impair more than
formation must first be “kept in mind” by working one memory system. Improved understanding of
memory.5 the types of memory will aid clinicians in the diag-
Evaluation of working memory is similar to that nosis and treatment of their patients’ memory dis-
of disorders of episodic memory. Treatment de- orders. This knowledge will become increasingly
pends on the specific cause; for instance, stimulants important as more specific strategies emerge for
have been approved by the FDA to treat attention the treatment of memory dysfunction.
deficit–hyperactivity disorder.55,56 Dr. Budson reports having received consultation or lecture fees
from Eisai, Forest Pharmaceuticals, Janssen, and Pfizer.
We are indebted to Daniel Schacter, David Wolk, Daniel Press, Jef-
conclusion frey Joseph, Dorene Rentz, Paul Solomon, and Hyemi Chong for
their helpful comments on the manuscript, figures, and supplemen-
tary appendix.
Traditionally, memory has been viewed as a simple
concept. In fact, the use of various methods has pro-
refer enc es
1. Newman MF, Kirschner JL, Phillips- source monitoring. Neuroreport 1997;8: 23. Tariot PN, Farlow MR, Grossberg GT,
Bute B, et al. Longitudinal assessment of 1317-20. Graham SM, McDonald S, Gergel I. Me-
neurocognitive function after coronary- 13. Kopelman MD, Stanhope N, Kingsley mantine treatment in patients with moder-
artery bypass surgery. N Engl J Med 2001; D. Temporal and spatial context memory in ate to severe Alzheimer disease already re-
344:395-402. [Erratum, N Engl J Med patients with focal frontal, temporal lobe, ceiving donepezil: a randomized controlled
2001;344:1876.] and diencephalic lesions. Neuropsycholo- trial. JAMA 2004;291:317-24.
2. Mesulam M-M. Principles of behavioral gia 1997;35:1533-45. [Erratum, Neuropsy- 24. Moretti R, Torre P, Antonello RM, Caz-
and cognitive neurology. 2nd ed. New York: chologia 1998;36:796.] zato G, Bava A. Use of galantamine to treat
Oxford University Press, 2000. 14. Johnson MK, O’Connor M, Cantor J. vascular dementia. Lancet 2002;360:1512-3.
3. Schacter DL, Tulving E. Memory systems Confabulation, memory deficits, and frontal 25. McKeith I, Del Ser T, Spano P, et al. Effi-
1994. Cambridge, Mass.: MIT Press, 1994. dysfunction. Brain Cogn 1997;34:189-206. cacy of rivastigmine in dementia with Lewy
4. Squire LR. Memory and the hippocam- 15. Budson AE, Sullivan AL, Mayer E, Daf- bodies: a randomised, double-blind, place-
pus: a synthesis from findings with rats, fner KR, Black PM, Schacter DL. Suppres- bo-controlled international study. Lancet
monkeys, and humans. Psychol Rev 1992; sion of false recognition in Alzheimer’s dis- 2000;356:2031-6.
99:195-231. [Erratum, Psychol Rev 1992; ease and in patients with frontal lobe lesions. 26. Cummings JL. Alzheimer’s disease.
99:582.] Brain 2002;125:2750-65. N Engl J Med 2004;351:56-67.
5. Fletcher PC, Henson RNA. Frontal lobes 16. Budson AE, Price BH. Memory: clinical 27. Press DZ. Parkinson’s disease dementia
and human memory: insights from function- disorders. In: Encyclopedia of life sciences. — a first step? N Engl J Med 2004;351:
al neuroimaging. Brain 2001;124:849-81. Vol. 11. London: Nature Publishing Group, 2547-9.
6. Wagner AD, Schacter DL, Rotte M, et al. 2002:529-36. 28. Schacter DL, Wagner AD, Buckner RL.
Building memories: remembering and for- 17. Solomon PR, Budson AE. Alzheimer’s Memory systems of 1999. In: Tulving E,
getting of verbal experiences as predicted by disease. Clin Symp 2003;54:1-40. Craik FIM, eds. The Oxford handbook of
brain activity. Science 1998;281:1188-91. 18. Folstein MF, Folstein SE, McHugh PR. memory. New York: Oxford University Press,
7. Brewer JB, Zhao Z, Desmond JE, Glover “Mini-mental state”: a practical method for 2000:627-43.
GH, Gabrieli JD. Making memories: brain grading the cognitive state of patients for 29. Corkin S. Lasting consequences of bilat-
activity that predicts how well visual experi- the clinician. J Psychiatr Res 1975;12:189-98. eral medial temporal lobectomy: clinical
ence will be remembered. Science 1998; 19. Blessed G, Tomlinson BE, Roth M. The course and experimental findings in H.M.
281:1185-7. association between quantitative measures Semin Neurol 1984;4:249-59.
8. Papez JW. A proposed mechanism of of dementia and of senile change in the cere- 30. Vaidya CJ, Zhao M, Desmond JE, Gabri-
emotion. Arch Neurol Psychiatry 1937;38: bral grey matter of elderly subjects. Br J Psy- eli JD. Evidence for cortical encoding speci-
725-43. chiatry 1968;114:797-811. ficity in episodic memory: memory-induced
9. Ribot T. Diseases of memory: an essay 20. Welsh KA, Butters N, Mohs RC, et al. re-activation of picture processing areas.
in the positive psychology. Smith WH, trans. The Consortium to Establish a Registry for Neuropsychologia 2002;40:2136-43.
Vol. 41 of The international scientific series. Alzheimer’s Disease (CERAD). V. A norma- 31. Damasio H, Grabowski TJ, Tranel D,
New York: Appleton, 1882. tive study of the neuropsychological battery. Hichwa RD, Damasio AR. A neural basis for
10. Simons JS, Spiers HJ. Prefrontal and Neurology 1994;44:609-14. lexical retrieval. Nature 1996;380:499-505.
medial temporal lobe interactions in long- 21. Solomon PR, Hirschoff A, Kelly B, et al. [Erratum, Nature 1996;381:810.]
term memory. Nat Rev Neurosci 2003;4: A 7 minute neurocognitive screening battery 32. Perani D, Cappa SF, Schnur T, et al. The
637-48. highly sensitive to Alzheimer’s disease. Arch neural correlates of verb and noun process-
11. Petrides M. The mid-ventrolateral pre- Neurol 1998;55:349-55. ing: a PET study. Brain 1999;122:2337-44.
frontal cortex and active mnemonic retriev- 22. Winblad B, Engedal K, Soininen H, et al. 33. Price JL, Morris JC. Tangles and plaques
al. Neurobiol Learn Mem 2002;78:528-38. A 1-year, randomized, placebo-controlled in nondemented aging and “preclinical”
12. Johnson MK, Kounios J, Nolde SF. Elec- study of donepezil in patients with mild to Alzheimer’s disease. Ann Neurol 1999;45:
trophysiological brain activity and memory moderate AD. Neurology 2001;57:489-95. 358-68.
34. Lidstrom AM, Bogdanovic N, Hesse C, ory in neurodegenerative disease: biologi- mer’s disease. J Neurol Neurosurg Psychia-
Volkman I, Davidsson P, Blennow K. Clus- cal, cognitive, and clinical perspectives. try 2001;70:157-64.
terin (apolipoprotein J) protein levels are in- Cambridge, England: Cambridge University 49. Gotham AM, Brown RG, Marsden CD.
creased in hippocampus and in frontal cor- Press, 1998:210-25. ‘Frontal’ cognitive functions in patients with
tex in Alzheimer’s disease. Exp Neurol 42. Sabe L, Jason L, Juejati M, Leiguarda R, Parkinson’s disease ‘on’ and ‘off’ levodopa.
1998;154:511-21. Starkstein SE. Dissociation between declar- Brain 1988;111:299-321.
35. Balota DA, Watson JM, Duchek JM, Fer- ative and procedural learning in dementia 50. Kubat-Silman AK, Dagenbach D, Ab-
raro FR. Cross-modal semantic and homo- and depression. J Clin Exp Neuropsychol sher JR. Patterns of impaired verbal, spatial,
graphic priming in healthy young, healthy 1995;17:841-8. and object working memory after thalamic
old, and in Alzheimer’s disease individuals. 43. Glisky EL, Schacter DL. Extending the lesions. Brain Cogn 2002;50:178-93.
J Int Neuropsychol Soc 1999;5:626-40. limits of complex learning in organic amne- 51. Sfagos C, Papageorgiou CC, Kosma KK,
36. Green JD, Hodges JR. Identification of sia: computer training in a vocational do- et al. Working memory deficits in multiple
famous faces and famous names in early main. Neuropsychologia 1989;27:107-20. sclerosis: a controlled study with auditory
Alzheimer’s disease: relationship to antero- 44. Baddeley AD. Recent developments in P600 correlates. J Neurol Neurosurg Psychi-
grade episodic and general semantic memo- working memory. Curr Opin Neurobiol atry 2003;74:1231-5.
ry. Brain 1996;119:111-28. 1998;8:234-8. 52. Egeland J, Sundet K, Rund BR, et al. Sen-
37. Hodges JR. Frontotemporal dementia 45. Rowe JB, Toni I, Josephs O, Frackowiak sitivity and specificity of memory dysfunc-
(Pick’s disease): clinical features and assess- RS, Passingham RE. The prefrontal cortex: tion in schizophrenia: a comparison with
ment. Neurology 2001;56:S6-S10. response selection or maintenance within major depression. J Clin Exp Neuropsychol
38. Heindel WC, Salmon DP, Shults CW, working memory? Science 2000;288:1656- 2003;25:79-93.
Walicke PA, Butters N. Neuropsychological 60. 53. Klingberg T, Forssberg H, Westerberg
evidence for multiple implicit memory sys- 46. Newman SD, Carpenter PA, Varma S, H. Training of working memory in children
tems: a comparison of Alzheimer’s, Hun- Just MA. Frontal and parietal participation with ADHD. J Clin Exp Neuropsychol 2002;
tington’s, and Parkinson’s disease patients. in problem solving in the Tower of London: 24:781-91.
J Neurosci 1989;9:582-7. fMRI and computational modeling of plan- 54. Purcell R, Maruff P, Kyrios M, Pantelis C.
39. Daselaar SM, Rombouts SA, Veltman ning and high-level perception. Neuropsy- Cognitive deficits in obsessive-compulsive
DJ, Raaijmakers JG, Jonker C. Similar net- chologia 2003;41:1668-82. disorder on tests of frontal-striatal function.
work activated by young and old adults dur- 47. Jaeggi SM, Seewer R, Nirkko AC, et al. Biol Psychiatry 1998;43:348-57.
ing the acquisition of a motor sequence. Does excessive memory load attenuate acti- 55. Elia J, Ambrosini PJ, Rapoport JL. Treat-
Neurobiol Aging 2003;24:1013-9. vation in the prefrontal cortex? Load-depen- ment of attention-deficit–hyperactivity dis-
40. Exner C, Koschack J, Irle E. The differ- dent processing in single and dual tasks: order. N Engl J Med 1999;340:780-8.
ential role of premotor frontal cortex and functional magnetic resonance imaging 56. Mehta MA, Goodyer IM, Sahakian BJ.
basal ganglia in motor sequence learning: study. Neuroimage 2003;19:210-25. Methylphenidate improves working memo-
evidence from focal basal ganglia lesions. 48. Calderon J, Perry RJ, Erzinclioglu SW, ry and set-shifting in AD/HD: relationships
Learn Mem 2002;9:376-86. Berrios GE, Dening TR, Hodges JR. Percep- to baseline memory capacity. J Child Psychol
41. Salmon DP, Lineweaver TT, Heindel tion, attention, and working memory are Psychiatry 2004;45:293-305.
WC. Nondeclarative memory in neurode- disproportionately impaired in dementia Copyright © 2005 Massachusetts Medical Society.
generative disease. In: Troster AI, ed. Mem- with Lewy bodies compared with Alzhei-