The neuroanatomy of remote memory

Abstract

In humans and experimental animals, damage to the hippocampus or related medial temporal lobe structures severely impairs the formation of new memory but typically spares very remote memory. Questions remain about the importance of these structures for the storage and retrieval of remote autobiographical memory. We carried out a detailed volumetric analysis of structural brain images from eight memory-impaired patients. Five of the patients had damage limited mainly to the medial temporal lobe. These patients performed normally on tests of remote autobiographical memory. Three patients had medial temporal lobe damage plus significant additional damage to neocortex, and these patients were severely impaired. These findings account for previously reported differences in the recollective ability of memory-impaired patients and demonstrate that the ability to recollect remote autobiographical events depends not on the medial temporal lobe but on widely distributed neocortical areas, especially the frontal, lateral temporal, and occipital lobes.

Figure 1

Magnetic Resonance Images Showing the Extent of Brain Damage for Five of the Eight Patients (A), (B), and (C) are T2-weighted axial images (T1-weighted axial images for P.H.) arranged from ventral (A) to dorsal (C). Damaged tissue is indicated by areas of bright signal (but by dark signal for P.H.). (D) in each row is a coronal, T1-weighted image taken at the level of the hippocampus. For all images, the left side of the brain is on the right side of the image. See text for detailed description of the lesions.

Figure 2

Volumes of Major Brain Regions for the Five Amnesic Patients with Lesions Limited Primarily to the Medial Temporal Lobe Patients are represented by filled circles. Separate control groups (n = 4; unfilled circles) were matched to patient E.P, patient G.P., and patients R.S., G.W., and J.R.W.

Figure 3

Volumes of Major Brain Regions for Three Amnesic Patients with Large Medial Temporal Lobe Lesions and Additional Damage to Neocortex Patients are represented by filled circles. Separate control groups (n = 4; unfilled circles) were matched to patient H.C., patient P.H., and patient G.T.

Figure 4

Performance on the Test of Remote Autobiographical Memory Participants were given 24 cue words (e.g., river, bottle, nail) and asked to recollect a specific event that involved the word. Patients were asked to recall events from the first third of their life before the onset of amnesia, and controls were asked for events from the same portion of their lives. Tape-recorded narratives were scored (0 to 3) for how well they described an event (0 = no response or a generic response, 1 = vague reference to a memory without any reference to time or place, 2 = memory that had some specificity but was not specific to one time and place, 3 = memory that was specific to one time and place). The bars show the mean number of narratives given each score, and the brackets show SEM. MTL, five patients with medial temporal lobe lesions; MTL+, three patients with medial temporal lobe lesions and additional lesions to neocortex; CON, 26 controls.

Figure 5

Performance on the Childhood Portion of the Autobiographical Memory Interview (A) Scores on items that assessed memory for autobiographical events (maximum score = 9). (B) Scores on items that assessed personal semantic knowledge (maximum score = 21). MTL, five patients with medial temporal lobe lesions; MTL+, three patients with medial temporal lobe lesions and additional lesions to neocortex; CON, 13 controls.

Figure 6

Characteristics of the Remote Autobiographical Memories (A) The percentage of Remember (R) and Know (K) responses for each autobiographical memory. A Remember response was scored if participants reported that a recollection was associated with the subjective feeling of mentally traveling back in time to the place that the event occurred. A Know response was scored if participants reported that the event had occurred but could not directly reexperience it. (B) Vividness of the visual imagery during autobiographical recall. Vividness scores are shown separately for autobiographical memories classified as Remember (R) and Know (K), as described in (A). Note that two of the eight controls did not report any Know memories, so that only six controls are shown in this condition. Participants were asked to rate the vividness of their mental imagery using a 5 point scale, 1 (“No image at all”) to 5 (“Perfectly clear and vivid as normal vision”). (C) The perspective from which visual imagery was viewed during autobiographical recall. Participants were asked to judge whether their recollections were viewed from the first-person perspective (from the participant's own viewpoint) or from the third-person perspective (from a third-person viewpoint). MTL, five patients with medial temporal lobe lesions; CON, 8 controls.