MacArthur SES & Health Network
MacArthur SES & Health Network

printable version

Memory Function and Hippocampal Formation Volume

Summary prepared by Sonia J. Lupien in collaboration with the Allostatic Load Working Group. Last revised January, 1997.

Chapter Contents

  1. Introduction
  2. Cognitive Measures of Hippocampal Function
  3. Structural Measures of Hippocampal Formation
  4. Magnetic Imaging Resonance Technic
  5. References


In the rat, sustained exposure to elevated glucocorticoid levels in later life is associated with an increased loss of hippocampal neurons, accompanied by severe memory impairments (Landfield et al., 1981). These results strongly support the idea that increased HPA activity accounts, in part, for individual differences in the occurrence of age-related hippocampal pathology and memory deficits.

The cognitive effects of elevated concentrations of corticosteroids in human populations have been studied in disorders affecting corticosteroid levels and using exogenous administration of the synthetic coumpound to healthy subjects. Mental disturbances mimicking mild dementia (such as decrements in simple and complex attentional tasks, verbal and visual memory, encoding, storage and retrieval) have been described in depressed patients with hypercortisolism (Weingartner et al., 1981; Cohen et al., 1982; Rubinow et al., 1984; Roy-Byrne et al., 1986; Wolkowitz et al., 1988, 1990), and in steroid psychosis following corticosteroids treatment (Hall et al., 1979; Ling et al., 1981; Varney et al., 1984; Wolkowitz et al., 1989). The role played by the hippocampus in HPA dysregulation in human populations is suggested by recent studies in patients with Cushing's syndrome which report significant positive correlations between hippocampal formation volume and scores on verbal memory tests and significant negative correlations between hippocampal formation volume and plasma cortisol levels (Starkman et al., 1992). Moreover, many investigators have reported inverse relationships between mean 24-hour cortisol levels and severity of cognitive decline in Alzheimer patients (De Leon et al., 1988; Oxenkrug et al., 1989; Martignoni et al., 1990).

The role of the hippocampal formation in human learning and memory is now well established (for a complete review, see Squire, 1992). More importantly, studies report that the hippocampus is essential for a specific kind of memory, notably declarative or explicit memory. In contrast, the hippocampus is not essential for nondeclarative or implicit memory. Explicit memory refers to conscious or voluntary recollection of previous information, whereas implicit memory refers to the fact that experience changes the facility for recollection of previous information without affording conscious access to it. Moreover, the hippocampus has been implicated in performance on several cognitive tasks other than declarative, particularly on those sensitive to the time-limited (Scolville & Milner, 1956), and spatial (O'Keefe & Nadel., 1978) aspects of memory. Patients with amnesia due to hippocampal dysfunction show normal retention at short delays and impaired retention at longer delays (Scolville & Milner, 1956), as well as spatial memory/spatial orientation deficits (O'Keefe & Nadel, 1978).

back to top

Cognitive Measures of Hippocampal Function

Declarative/Non-Declarative Memory

In order to measure declarative memory, a list of 12 word-pairs is presented to the subject (Lussier & Lupien, in preparation). The list of words is comprised of six moderately related word pairs (related-pairs) and six unrelated word pairs (unrelated-pairs). The subjects are presented with the list of word pairs which they have to read aloud. Then, the subjects make a cued recall where they have to recall a member of a pair when presented with the other. Non-declarative memory is thereafter tested by a word completion task (trigrams). Subjects are presented with the three first letter of words, and instructed to complete each presented syllable as fast as possible and with the first word that come to mind. Among these trigrams, 24 correspond to the first syllable of each word of the pairs learned previously, and recall of the words previously learned on the declarative task represents the non-declarative score.

Figure 1 Declarative Memory Performance

Figure 1 presents the declarative memory score of elderly subjects showing significant (PSE group) or moderate increases of cortisol levels with years (PSM group), or decrease (NS group) of cortisol levels with years (Lupien et al., 1995).

Immediate and Delayed Memory

Immediate and delayed memory are measured using presentaiton of 15 non-complex line-drawings of objects of everyday use. The subject is presented with the 15 line-drawings for 3 seconds each, and is asked to name the object. Sujects are then asked to verbally recall as many line-drawings as possible, immediately after the presentation or 24 hours later (Lussier, 1995).

Figure 2 Immediate and Delayed Memory Performance

Figure 2 presents immediate and delayed memory performance in aged human subjects showing significant increase (PHC group) or decrease (NLC) of cortisol levels with years (Lupien et al, 1998).

back to top

Spatial Memory

Spatial memory function is measured using a human maze (designed by Dr Romedi Passini from the Dept. of Architecture, Université de Montréal). The surface area of the maze is 1,500 square feet and the walls are 6 feet high, with no extraneous cues, either on the floor or on the ceiling. The passages corresponded to a small, domestic corridor of one meter in width. The subject is shown a path by following the experimenter through the maze. The sujects have to learn a simple and a complex path. The complexity of the path is determined by the number of decision points in the path. A decision point is an intersection in the maze to which the suject must take a decision (turn left, right or go straight ahead). The simple path is comprised of 3 points of decision, while the complex path is comprised of 5 points of decision. The time taken to find the correct path serves as the measure of spatial memory function.

Figure 3 Spatial Memory Performance

Figure 3 presents spatial memory performance in aged human subjects showing significant increase (PHC group) or decrease (NLC) of cortisol levels with years (Lupien et al, 1995).

Structural Measures of Hippocampal Formation

Recent pathological analysis (Coleman & Flood, 1987) and neuroimagining studies (Convit at al; 1995 Golomb et al, 1993, 1994) have shown that early in the course of Alzheimer's disease (AD), there are pronounced atrophic changes in the hippocampal formation. In non-demented elderly subjects, similar but less pronounced hippocampal atrophy is associated with isolated memory dysfunctions; individuals with these milder disturbances are at increased risk for later dementia. That is, magnetic resonance imaging (MRI) studies have shown that, after controlling for age, hippocampal atrophy is the specific anatomic correlate of delayed recall performance in normal elderly (Golomb et al., 1993, 1994) and elderly with very mild memory impairments (Convit et al., 1995). Moreover, it has further been shown that hippocampal atrophy and mild memory changes in non-demented elderly subjects are both sensitive and specific predictors over 4 years of future clinical decline to the status of dementia and a diagnosis of probable Alzheimer's disease (de Leon et al., 1993; Flicker et al., 1991).

back to top

Experimental studies have demonstrated that the hippocampus plays a major role in both memory (Squire, 1992) and in mediating the response to stress (Reul & deKloet, 1993). The hippocampus has the highest concentration of glucocorticoid receptors and is considered the major central nervous system (CNS) site controlling the negative feedback to the hypothalamic-pituitary-adrenal (HPA) axis (McEwen, 1988). Damage to the hippocampus has been repeatedly demonstrated to disrupt the HPA axis resulting in its overactivation. Conversely, overactivation of the axis under conditions of chronic stress has been shown to damage the hippocampus thus leading to further increases in circulating glucocorticoids that potentially further damage the hippocampus (Sapolsky et al., 1986).

Magnetic Imaging Resonance Technic

(Taken from DeLeon et al., personal manuscript)

In order to obtain anatomically accurate MR volumes of cerebral anatomy, several requirements must be satisfied. First, the scan sequence must provide sufficient spatial and contrast resolution to define the boundary between the structure of interest and its surrounding tissue. Second, an axis of planar sampling must be selected to maximize the number of slices taken through the structure while minimizing partial volume effects. Finally, coverage must include the entire head.

We have developed an MRI protocol that yields a high resolution image with excellent gray matter/white matter/CSF contrast. This protocol has been in use for over two years and over 500 studies have been completed. This MRI study permits accurate measurements of small brain structures. It was specifically designed to be used in the quantification of the human hippocampus. The 3-D gradient echo sagittal scan data of 128slices at 1.2 mm collected using a GE 1.5T Signa system will satisfy those requirements (TR=24, TE=5, FOV=24, NEX=2, with a 256 x 192 matrix).

The temporal lobe analyses is based on reformatted coronal slices with a 2 mm thickness. Therefore, depending on the extent of the hippocampus and temporal lobe chosen, we study between 20 and 50 MRI slices per case. Using multiplanar reformatting of the final data sets, for each hemisphere, a beginning and an ending slice that is fixed with respect to discrete anatomic reference points is identified. This technique ensures that the target anatomy is always represented in a geometrically uniform and standardized format for volume sampling. Image analysis is done on a graphic workstation (Sun Microsystems Sparc) on Unix operating system using our locally developed "Midas" software. Anatomical regions of interest (ROIs) are drawn on the coronal images so as to outline individual medial and lateral temporal lobe structures. The volume of the structure is then calculated by summing the areas of the ROI across slices and then compensating for slice thickness and field of view.

In our volumetric measurements, the medial temporal lobe is made up of the hippocampus and the parahippocampal gyrus. A separate estimation of the amygdala volume is performed using a different reformatted set of images. The lateral temporal lobe is composed of the fusiform gyrus, and the superior, medial, and inferior temporal lobe gyri. To correct for head size variations across individuals, we obtain an intracranial supratentorial volume. Every third sagittal image (mid-points every 3.6mm) is used to trace the outline of the supratentorial compartment by following the dural and tentorial margins. Anatomic description of the precise boundaries for individual structures are published and available on request. Table 1 presents hippocampal and temporal lobe structure measures in aged human subjects showing significant increase (PHC group) or decrease (NLC group) of cortisol levels with years (Lupien et al, submitted).

back to top

Table 1. Per slide brain regional volumes in cubic centimeters.

Left Hippocampus 0.17 ± 0.06 0.20 ± 0.05*
Right Hippocampus 0.18 ± 0.08 0.21 ± 0.05*
R + L Hippocampus 0.35 ± 0.09 0.41 ± 0.07*
Parahippocampus 0.24 ± 0.02 0.24 ± 0.03
Fusiform 0.36 ± 0.04 0.37 ± 0.03
Superior Termporal 0.91 ± 0.04 0.86 ± 0.06
Mid-Inferior Termporal 1.13 ± 0.06 1.26 ± 0.05
Temporal Lobe 2.77 ± 0.13 2.89 ± 0.12
Lateral Temporal Lobe 4.68 ± 0.23 4.86 ± 0.18
*: Significant group difference on T-test (T=1,9); p<.05 in all cases.
Figure 4 Cortisol Levels and Hippocampal Volume

Figure 4 presents the correlation between changes (increase or decrease) in cortisol levels with years and hippocampal volume in 11 elderly human subjects (Lupien et al, 1998).

back to top


Coleman PD, Flood DG (1987) Neuron numbers and dendritic extent in normal aging and Alzheimer's disease. Neurobiology of Aging 8:521-545.

Convit A, de Leon MJ, Tarshish C, de Santi S, Kluger A, Rusinek H, George A (1995) Hippocampal volume losses in minimally impaired elderly. The Lancet 345:266.

DeLeon M, McRae T, Tsai J, George A, Marcus D, Freedman M, Wolf A, Mc Ewen B (1988) Abnormal cortisol response in Alzheimer's disease linked to hippocampal atrophy. Lancet 2:391-392.

de Leon MJ, Golomb J, George AE, Convit A, Tarshish CY, McRae T, de Santi S, Smith G, Ferris SH, Noz M, Rusinek H (1993) The radiologic prediction of Alzheimer's disease: The atrophic hippocampal formation. Amer J Neuroradiology 14:897-906.

Flicker C, Ferris SH, Reisberg B (1991) Mild cognitive impairment in the elderly: Predictors of dementia. Neurology 41:1006-1009.

Golomb J, de Leon MJ, Kluger A, George AE, Tarshish C, Ferris SH (1993) Hippocampal atrophy in normal aging: An association with recent memory impairment. Arch Neurology 50:967-976.

Golomb J, Kluger A, de Leon MJ, Ferris S, Convit A, Mittelman M, Cohen J, Rusinek H, de Santi S, George A (1994) Hippocampal formation size in normal human aging: A correlate of delayed secondary memory performance. Learning and Memory 1:45-54.

Hall RC, Popkin MK, Stickney SK, Gardner ER (1979) Presentation of the steroid psychoses. J Nerv Ment Dis 167:229-236.

Landfield PW, Baskin RW, Pitter TA (1981) Brain-aging correlates: Retardation by hormonal pharmacological treatments. Science 214:581-584.

Ling M, Perry P, Tsuang M (1981) Side effects of cortico steroid therapy. Arch Gen Psychiat 38:471-477.

Lupien SJ, DeLeon M, DeSanti S, Convit A, Tarshish C., Nair NPV, McEwen BS, Hauger RL & Meaney M (1998). Longitudinal increase in cortisol during human aging predicts hippocampal atrophy and memory deficits. Nature Neuroscience, 1:69-73.

Lupien S, Ngô T, Rainville C, Nair NPV, Hauger RL, Meaney MJ (1995) Spatial memory as measured by a human maze in aged subjects showing various patterns of cortisol secretion and memory function. Society for Neuroscience 21:1709.

Martignoni E, Petraglia F, Costa A, Bono G, Genazzani AR, Nappi G (1990) Dementia of the Alzheimer type and hypothalamus-pituitary-adrenocortical axis : Changes in cerebrospinal fluid, corticotropin releasing factor and plasma cortisol levels. Acta Neurolo Scand 81:452-456.

McEwen BS, Weiss JM, Schwartz LS (1968): Selective retention of corticosterone by limbic structures in rat brain. Nature 220:911-912.

back to top

O'Keefe, J, Nadel, L. The Hippocampus as a Cognitive Map, Oxford: Clarendon Press, 1978.

Oxenkrug GF, Gurevich D, Siegel B, Dumiao MS, Gershon S (1989) Correlation between brain-adrenal axis activation and cognitive impairment in Alzheimer's disease : Is there a gender effect? Psychiatry Res 29:169-175.

Reul JMHM, De Kloet ER (1985) Two receptor systems for corticosterone in rat brain: Microdistribution and differential occupation. Endocrinology 117:2505-2511.

Rubinow D, Post R, Savard R, Gold P (1984) Cortisol hypersecretion and cognitive impairment in depression. Arch Gen Psychiat 41:279-283.

Sapolsky RM, Krey LC, McEwen BS. (1986) The neuroendocrinology of stress and aging: The glucocorticoid cascade hypothesis. Endocr Rev 7:284-301.

Scolville WB, Milner B (1957) Loss of recent memory after bilateral hippocampal lesions. J Neurol, Neurosurg Psychiatry 20:11-21.

Squire LR (1992) Memory and the hippocampus: A synthesis from findings with rats, monkeys, and humans. Psychol Rev 99:195-231.

Starkman MN, Gebarski SS, Berent S, Schteingart DE (1992) Hippocampal formation volume, memory dysfunction, and cortisol levels in patients with Cushing's syndrome. Biol Psychiatry 32:756-765.

Varney NR, Alexander B, MacIndoe JH (1984) Reversible steroid dementia in patients without steroid psychosis. Am J Psychiatry 141:369-372.

Weingartner H, Cohen RM, Martello J (1981) Cognitive processes in depression. Arch Gen Psychiatry 38:42-47.

Wolkowitz OM, Weingartner H (1988) Defining cognitive changes in depression and anxiety: A psychobiological analysis. Psychiatr Psychobiol 3:1315-1385.

Wolkowitz OM, Reus VI, Weingartner H, Thompson K, Breier A, Doran A, Rubinow D, Pickar D (1990) Cognitive effects of corticosteroids. Am J Psychiatry 147:1297-1303.

back to top


UCSF Home About UCSF Search UCSF UCSF Medical Center