Skip main navigation

Impaired Maze Learning and Cerebral Glucose Utilization in Aged Hypertensive Rats

Originally published 1995;25:545–553


    Abstract To elucidate the effects of prolonged hypertension on brain function during aging, we examined learning of an eight-arm radial maze task and local cerebral glucose utilization in young-adult (3 to 4 months old) and aged (16 to 17 months old) spontaneously hypertensive rats (SHR) and Wistar-Kyoto rats (WKY). Young-adult SHR learned the task more slowly than young-adult WKY, but cerebral glucose utilization, measured by the [14C]2-deoxyglucose method in 24 brain structures, was not significantly different in the two groups. The aged SHR and WKY exhibited impaired learning ability. Cerebral glucose utilization was reduced (13% to 23%) in six regions in aged WKY and in 12 regions in aged SHR compared with values in the respective young-adult groups. Furthermore, the aged SHR showed a greater disturbance of learning acquisition and more profound reduction of cerebral glucose utilization in five regions than the aged WKY. In SHR, hypometabolism, indicated by a decrease in glucose utilization in 15 brain structures including the cerebral cortex, hippocampus, and visual system, was significantly correlated with impaired learning acquisition, indicated by an increase in total error choices. These findings show that (1) hypertension per se does not impair maze learning or cerebral glucose utilization in young-adult rats, and (2) brain function is impaired during aging and prolonged hypertension is an additional factor facilitating brain dysfunction associated with neuronal hypoactivities, resulting in behavioral deterioration including learning disability. Thus, early control of hypertension seems important for preventing or reducing brain dysfunction in senescence.

    Aged hypertensive patients have been reported to show cognitive impairment.123 It is controversial, however, whether long-standing hypertension itself modifies cognitive function without stroke or dementia.4 This is an important issue because hypertension can be controlled by medication.

    Spontaneously hypertensive rats (SHR) have been widely used as a model of human essential hypertension. The learning and memory abilities of SHR seem to differ depending on the type of task and age. Faster learning in avoidance tasks5678910 and better performance in a Hebb-William’s maze11 of SHR have been reported. However, poor performance has been observed in young-adult SHR in a two-way shuttle box task.10 Recently, Wyss et al12 reported mildly impaired learning ability in an eight-arm radial maze task in adult SHR compared with normotensive Sprague-Dawley rats. The eight-arm radial maze task is widely used for the evaluation of learning and memory ability of rats because it does not require any harmful stimulation or stress and can be used to evaluate two different aspects of memory. One is reference memory, or long-term memory, by which the rat can walk down to the end of the arms to obtain food. This memory is common across trials. The other is spatial working memory, or short-term memory, of a particular arm that the rat has already visited and obtained food. The memory should be renewed at the completion of each trial. Spatial memory, which is postulated to play an important role in the eight-arm radial maze task, has been related to functions of various brain structures. Olton et al13 found that destruction of the entorhinal cortex, septum, or fornix resulted in an impairment in radial maze performance, supporting the hypothesis that the hippocampus has an important role in producing information about spatial location. Lesions of other limbic structures, including the amygdala,14 olfactory bulb,15 and mediodorsal thalamic nucleus,16 may disrupt interconnections to septohippocampal systems or may fundamentally compromise memory systems. Lesions of the superior colliculus1718 and visual cortex18 caused spatial memory deficits, suggesting that visual information is essential for spatial memory. In contrast, memory function may not be directly correlated with the auditory, vestibular, or sensorimotor system.

    Wei et al19 found that local cerebral glucose utilization (LCGU) was decreased in adult SHR of 6 to 7 months old, suggesting impaired neuronal activities in SHR of this age. However, no detailed studies have been conducted on modifications of neuronal activities of SHR with age. Accordingly, we designed the present experiments to investigate the following points in SHR: (1) How does hypertension per se affect radial maze learning and LCGU in young-adult SHR after establishment of hypertension? (2) How does lifelong hypertension affect maze learning and LCGU in aged SHR? (3) Is there any correlation of impaired memory with hypometabolism in specific brain structures?


    Male SHR and age- and sex-matched normotensive Wistar-Kyoto rats (WKY) were used. The animals were kept in a pathogen-free facility of the Laboratory of Animal Experiments, Kyushu University, Fukuoka, Japan, at a constant room temperature of 24°C with a 12-hour light/dark schedule (lights on at 8 am) and were allowed free access to food and water.

    Studies were made in young-adult and aged rats of changes with age in physical parameters such as body weight, heart rate, and systolic pressure as well as survival rate, the ability of the rats to learn an eight-arm radial maze task, and LCGU, as an indicator of local neuronal activities in the brain. Finally, we correlated learning behavior with LCGU to determine the neuronal basis of changes in the learning behavior. All rats used were treated humanely on the basis of the Public Health Service policy of the National Institutes of Health (revised 1986).

    As shown in Table 1, we used 119 rats (74 SHR and 45 WKY): 77 (49 SHR and 28 WKY) rats for observation of survival rate (among them, 12 SHR and 12 WKY were used in studies on physiological changes) and 69 (44 SHR and 25 WKY) for maze learning (20 of these for LCGU determination after maze learning). Twenty-seven (19 SHR and 8 WKY) rats were also used for observation of both maze learning at 3 to 4 months of age and survival rate up to 22 months of age (Table 1).

    Changes in Physical Indicators

    We measured body weight, heart rate, and systolic pressure in 12 SHR and 12 WKY at 2, 6, 9, 12, 15, 18, and 22 months of age. Heart rate and systolic pressure were measured noninvasively by the standard tail-cuff method with a blood pressure monitor (model MK-1000, Muromachi Kikai). During measurement, the rats were kept in a case at 33°C warmed with an air bath.

    The survival rates of 49 SHR and 28 WKY from 2 and 22 months of age were observed at the same times as the above observations. Of the 49 SHR, 12 were used for observations of physical parameters and 19 for studies on radial maze learning at 3 to 4 months of age. Of the 28 WKY, 12 were used for the above observations and 8 for experiments on radial maze learning at 3 to 4 months of age.

    Learning An Eight-Arm Radial Maze Task

    Radial maze learning was studied at 3 to 4 months of age (young-adult group) and 16 to 17 months of age (aged group) in 44 SHR (24 young-adult and 20 aged) and 25 WKY (13 young-adult and 12 aged). The eight-arm radial maze used in the present study was similar to that described by Olton and Samuelson.20 Each of the eight radial arms was 70 cm long and 15 cm wide, with walls 6 cm high on either side, and was equipped at its distal end with a white food cup 2.5 cm in diameter and 1 cm in depth. The maze was approximately 50 cm above the floor. The central platform was 41 cm in diameter with walls 12 cm in height and guillotine doors, 20 cm in height, which were manipulated up and down by a string from outside the experimental room. From 1 week before the learning experiment, food was withheld to reduce body weights to 80% to 85%. The reduced body weights were maintained throughout the experiment. The experimental room was quiet and dimly lit by contrast lights (approximately 10 lux) with numerous visual cues scattered around the room, including a picture, chairs, a light stand, and a desk. These cues were kept constant throughout the experiment. The experimenter watched the behavior of the rats from outside the test room through a window.

    Before learning sessions, rats received habituation trials once a day for 3 consecutive days, in which three or four rats were simultaneously placed in the radial maze for 15 minutes and allowed to explore freely and take 45 mg of food pellets scattered in the arms and in the food cups. At the start of each trial, one 45-mg food pellet was placed in the food cup at the end of each arm.

    One minute after the rat was placed on the central platform, the guillotine doors were raised at the beginning of one trial for each rat, and the rat was allowed to choose one of the eight arms to obtain food. Each trial lasted until the rat obtained all eight pellets or for 30 minutes. The behavior of each rat was monitored with a video camera and analyzed later. When the rat obtained the food, a correct response was registered. The number of correct responses within the first eight choices was designated as the number of correct choices (CCs). When the rat chose an arm previously visited, an error response was registered. The total number of error responses in a trial was designated as the total error choices (TECs). Each rat was given one trial per day between 4 pm and 8 pm on 6 of 7 days. The full acquisition of learning was defined as eight correct responses among the first nine choices in 3 consecutive days. Three consecutive trials made one block; the maze learning was continued for 36 trials (12 blocks). The CCs and TECs in each block were averaged. Three young-adult SHR that successively chose an arm next to the one currently explored and succeeded in getting all the pellets without failure were excluded from the present study.

    The animals subjected to the learning study were subsequently divided into two groups, one for LCGU determination at the age of 4 or 17 months, and the other for observation of the survival rate.

    LCGU Determination

    Twenty rats were used for LCGU determination, 5 of each strain of each age group. After maze learning, 5 rats in each group were selected randomly and in a blinded manner. LCGU was determined as described before.21 Briefly, rats were anesthetized with halothane, and polyethylene catheters (PE-50, Clay Adams) were inserted into the right femoral artery and vein. The hindquarters of the rats were then restrained with a loose-fitting plaster cast and anchored with tape. After 3 hours or more to allow complete recovery from anesthesia, mean arterial blood pressure, heart rate, and rectal temperature were measured. Then, approximately 80 μL of arterial blood was withdrawn for measurement of hematocrit, pH, Paco2, and Pao2 (model 1304 Blood Gas Analyzer, Instrumentation Laboratories). Basal arterial plasma glucose concentration was also determined (Glucose Analyzer 2, Beckman Instruments). A bolus of 4.63 MBq/kg of 2-deoxy-d-[1-14C]glucose (2-DG; specific activity, 2.04 GBq/mmol; American Radiolabeled Chemicals) was injected intravenously into fully conscious, resting rats, and 11 timed arterial blood samples were taken during the next 45 minutes for determination of the plasma concentrations of glucose and 14C radioactivity (Liquid Scintillation Counter LSC 1000, Aloka). Then the rats were decapitated, and the brain was promptly removed and frozen in isopentane cooled to −30°C with dry ice. The brain was cut serially with a cryostat into coronal sections 30 μm thick (Cryocut 2, American Optical) maintained at −22°C. The sections were mounted on glass slides and dried on a hot plate at 60°C. Autoradiographs were prepared in x-ray cassettes by exposing the dried sections to x-ray films (Kodak SB-5) for 7 to 10 days, together with [14C]methylmethacrylate standard disks including a blank and six serial calibrated 14C levels. Optical densities of regions of interest of cerebral structures were measured on the autoradiograph with a drum-scan densitometer (PDI-10 model 2605, Konica). These optical densities were transformed to 14C concentrations of given tissues from the standard curve expressing the relation between 14C concentration and optical density. From the time courses of changes in concentrations of [14C]2-DG and glucose in plasma, together with the local tissue concentrations of 14C, LCGU was calculated by an operational equation,22 using the rate constants and lumped constant for normal adult rats reported by Sokoloff et al.22 According to the age-dependent difference in the lumped constants observed in one rat strain (Fischer-344 rats: 0.50 at 3 months, 0.46 at 12 months, and 0.42 at 24 months of age23 ), we also used age-corrected lumped constant values. The LCGU for a given cerebral structure was measured bilaterally in two slices, and the averaged value was recorded. Values are quantified in 24 gray matter structures.

    Selected brain sections corresponding to the autoradiographs were stained with cresyl violet for histological verification of structures as well as detection of hemorrhagic lesions and infarctions. Before and during sectioning, macroscopic observations were also made for detection of lesions. Descriptions of brain anatomy are according to the atlas of the rat brain of Paxinos and Watson.24

    Statistical Analysis

    Values are mean±SEM. Multiple comparisons of four groups were performed by one-way ANOVA and Scheffé’s test. Nonparametric tests (Kruskal-Wallis ANOVA and Spearman’s rank correlation) were used for analyses of behavioral scores because these scores were not distributed normally in the entire group. Computations were carried out using the statistical package BMDP 7D25 and an IBM system 4381 computer. Correlations between behavioral scores (mean number of TECs in the last 10 trials) and LCGU of 24 brain structures of each strain of young-adult and aged rats were evaluated by Spearman’s rank correlation.


    Physical Data During Aging

    Fig 1 shows survival rate, body weight, systolic pressure, and heart rate of SHR and WKY 2 to 22 months of age. The survival rate of SHR was significantly lower from 15 months of age (Fig 1a). The body weight of WKY increased until 12 months of age, whereas that of SHR was almost constant after 6 months, resulting in significantly lower body weights of SHR than WKY from 9 months of age (Fig 1b). The systolic pressure and heart rate of SHR were significantly higher than those of WKY throughout the 22-month experimental period (Fig 1c and 1d). The systolic pressure of SHR rose between 2 and 6 months and then remained constant (Fig 1c). The heart rates of WKY and SHR decreased significantly (both P<.05) between 2 and 6 months of age (Fig 1d).

    Radial Maze Task

    Correct Choice (CC)

    The CC score of young-adult WKY was 6.0±0.2 (mean±SEM) in the first block of three trials, increased markedly in the second block, and then increased gradually until the eighth block, when it reached a plateau (Fig 2a). The maximal CC was 7.9±0.1 in the eighth block. The CC scores of young-adult SHR were significantly lower than those of young-adult WKY from the second to fifth block, although their CC score in the first block (6.4±0.1) was significantly higher than that of young-adult WKY (Fig 2a). The maximal CC score of the young-adult SHR was 7.6±0.1 in the last block.

    The CC score of aged WKY was lower than that of young-adult WKY in blocks 2 and 4 through 12 (Fig 2a). The CC score of aged WKY was 5.7±0.2 in the first block and reached a maximum of 7.3±0.1 in the last block. The CC score of aged SHR was significantly lower than that of both young-adult groups in all blocks (P<.005, respectively) except young-adult WKY in the first block and young-adult SHR in the second and third blocks (Fig 2a). The final CC score of aged SHR was lower than that of aged WKY (P<.05). The initial CC score of aged SHR was 5.9±0.1, and the maximal scores were 6.7±0.1 and 6.7±0.2 in the 11th and 12th blocks, respectively.

    Total Error Choice (TEC)

    The TEC score of young-adult WKY was 6.0±0.8 (mean±SEM) in the first block, decreased markedly in the second block, and decreased gradually thereafter (Fig 2b), with a minimum of 0.1±0.1 in the eighth block. The TEC of young-adult SHR tended to be higher than that of young-adult WKY, but the difference was not significant (Fig 2b). The initial TEC of young-adult SHR was 6.8±0.8, and the minimal value was 0.6±0.1 in the last block.

    The TECs of aged WKY were significantly higher than those of young-adult WKY in blocks 2 and 7 through 9. The initial TEC of aged WKY was 8.0±1.2, and the minimal score was 2.0±0.4 in the last block. The aged SHR showed the highest TECs of the four groups; these TEC values were significantly greater in blocks 2 through 12 than those of young-adult WKY and SHR in blocks 2 and 4 through 12 as well as in blocks 4 through 6 and 8 through 12 than those of aged WKY.

    Full Acquisition of Task Learning

    All young-adult WKY attained the full acquisition level in an average of 12 trials (Table 2), with a range from 6 to 20 trials, and 22 of 24 (92%) young-adult SHR attained the full acquisition level in an average of 23 (range, 9 to 35) trials. Although the rates of attaining full acquisition were not significantly different in young-adult SHR and young-adult WKY, the number of trials needed for attaining the full acquisition level was significantly greater in young-adult SHR (P<.01). Seven of 12 (58%) aged WKY reached the full acquisition level at an average of 25 (14 to 35) trials. The incidence of full acquisition in aged WKY was significantly less (P<.05), and the number of trials for full acquisition was significantly greater (P<.01) than that of young-adult WKY. Four of 20 (20%) aged SHR reached the full acquisition level in an average of 23 (18 to 32) trials, the incidence of acquisition being significantly less than in young-adult SHR (P<.01), aged WKY (P<.05), and young-adult WKY (P<.01). The number of trials necessary for attaining the full acquisition level, however, was not significantly different from those of young-adult SHR and aged WKY.

    Local Cerebral Glucose Utilization

    There were no differences in physiological variables at the start of LCGU measurement between WKY and SHR in either age group, except for mean arterial blood pressure, which was significantly higher in SHR of both age groups. Body weight was significantly greater in the aged WKY and SHR (Table 3). Macroscopic and histological examinations of the brain after the completion of the experiment revealed no hemorrhagic lesions or infarctions in any brain region of any rats.

    The mean LCGU values in the 24 cerebral structures of the four subgroups are listed in Table 4. Although LCGU values were generally lower in the young-adult SHR than in the young-adult WKY, there were no statistically significant differences between the values in any structures in the two young-adult groups.

    LCGU tended to be lower in most structures of aged WKY than in those of young-adult WKY. Significant decreases were seen in six structures: the frontal, auditory, and visual cortices; striatum; olfactory tubercle; and cochlear nucleus (Table 4). The decreases were 13% to 23%, with the frontal cortex, olfactory tubercle, and cochlear nucleus showing the greatest decreases of more than 20%.

    Aged SHR showed lower LCGU than young-adult SHR in all gray structures. Significant decreases in LCGU were found in 12 structures of aged SHR: the frontal, parietal, auditory, and visual cortices; striatum; entopeduncular nucleus; anteroventral thalamic nucleus; hippocampus; olfactory tubercle; piriform cortex; superior colliculus; and cochlear nucleus (Table 4), with reduction rates varying between 13% and 26%. In particular, the frontal and visual cortices, anteroventral thalamic nucleus, olfactory tubercle, and cochlear nucleus showed reduction rates of more than 20%.

    Comparison of LCGU in the two aged groups revealed lower values in aged SHR in all gray structures. Significant reductions were found in five structures: the parietal and visual cortices, anteroventral thalamic nucleus, hippocampus, and ventral tegmental area (Table 4), with reduction rates varying between 13% and 23%.

    Correlation Between Radial Maze Task Learning and LCGU

    The correlation between the averaged TEC in the last 10 trials and the LCGU of each structure revealed no significant relationship in young-adult or aged WKY. Young-adult and aged SHR, however, showed significant correlations between TECs and LCGU values in 15 of 24 structures (Table 5). These 15 structures were scattered throughout the brain: the frontal and parietal cortices, hippocampus, hippocampal formation (the subiculum) and hippocampal connections (the medial septal nucleus), other limbic structures (the amygdala, olfactory tubercle, piriform cortex, hypothalamus, ventral tegmental area, and basal nucleus of Meynert), and the visual system (the visual cortex and superior colliculus).


    The main findings in the present study were (1) hypertension per se did not impair maze learning and LCGU in young-adult SHR, (2) long-standing hypertension, associated with aging, impaired maze learning more severely and reduced LCGU more profoundly and more widely in aged SHR than aged WKY, and (3) impaired learning ability may be correlated with hypometabolism in 15 brain structures: the cerebral association cortex and visual and limbic systems (including the hippocampus, medial septal nucleus, amygdala, and basal nucleus of Meynert).

    Radial Maze Learning

    Most (92%) young-adult SHR reached the full acquisition level but needed more trials (23 trials) than young-adult WKY (12 trials). In contrast with this finding, faster learning of young-adult SHR has been observed in a radial maze task with six of eight arms baited.12 The reason for this discrepancy is unknown, but the slower rate of learning of young-adult SHR might be due to behavioral and emotional alterations rather than impaired learning ability: Decreased attention or poor motivation associated with increased exploration (locomotor hyperactivity),26 aggressiveness,2728 irritability,26 or decreased thigmotaxis11 have been reported in SHR.

    Aged normotensive WKY showed significant impairment of learning ability compared with young-adult WKY. This finding is consistent with previous reports on Wistar2930 and Fischer-34431 rats, indicating “primary” neuronal hypoactivity during aging. The aged SHR showed a more marked decrease of learning ability than the aged WKY. The more marked learning deficit in aged SHR may be correlated with more pronounced neuronal hypoactivities as reflected by the more profound reduction of LCGU than in aged WKY: The TECs and LCGU values showed significant correlations in SHR. These findings indicate that during senescence, additional factors, particularly the effect of long-standing hypertension on brain function, have important effects on neuronal hypoactivity, which consequently lead to learning deficits. Wyss et al12 reported that 12-month-old adult SHR showed higher error scores and slower learning acquisition in a radial maze task than age-matched Sprague-Dawley rats but that all the adult SHR ultimately reached the full acquisition level. Precise comparison between their study and ours is not possible because of some differences in experimental design and criteria for learning acquisition, but it seems that impairment of learning progresses between 12 months and 16 to 17 months of age.

    Local Cerebral Glucose Utilization

    Young-adult SHR of 4 months old showed no significant LCGU decreases in any of the 24 cerebral structures compared with age-matched WKY, although hypertension had already been established by that age. This finding is in general agreement with results on 11- to 12-week-old SHR by Kadekaro et al32 and 20-week-old SHR by Hayashi and Nakamura.33 Although enlargement of ventricles3435 and brain atrophy343637 were observed in young-adult SHR, their neuronal density did not differ from that of WKY.37 The present study indicates that the neuronal activities of SHR are well maintained in 4-month-old young adults. In normotensive rats, hindlimb immobilization does not alter LCGU.38 The effect of immobilization stress on LCGU of SHR is not fully known, but it may be little, if any, because LCGU and plasma glucose concentration of young-adult SHR were not different from those of young-adult WKY (Tables 3 and 4) or another normotensive strain, the Sprague-Dawley rat.21

    The reports in different strains of normotensive rats, such as Fischer-34439 and Sprague-Dawley40 rats, and our findings in aged WKY indicate that neuronal activities deteriorate in scattered brain regions during aging in normotensive rats. Similar LCGU reduction was seen in aged SHR. The number of structures showing LCGU decrease, compared with those in the respective young-adult groups, was greater in aged SHR (12 structures) than in aged WKY (6 structures). In 5 structures, the decreases were greater in aged SHR than in aged WKY. These findings indicate that the neuronal activities decreased more extensively and in more cerebral structures in aged SHR than in aged WKY and suggest that there are additional factors, probably such as long-standing hypertension, causing more extensive and more widespread neuronal hypoactivity in aged SHR. LCGU was similar in 2- to 5-month-old young-adult SHR and WKY.3233 A study of adult SHR19 suggested that LCGU decreases in adult SHR from 6 to 7 months of age, and our data on aged SHR indicated that LCGU reduction persisted and progressed later on, resulting in more extensive LCGU reduction in SHR than in WKY at 16 to 17 months of age. In aged rats, the cerebral structures in which LCGU decreased were not limited to a particular functional system (Tables 4), indicating that the decreases are not necessarily due to transneuronal effects.

    It is not yet known how long-standing systemic hypertension causes brain hypometabolism, but three possibilities may be considered: SHR show morphological and functional changes of cerebral blood vessels, including a decrease in the inner diameter of pial arterioles,41 impaired endothelium-mediated dilatation,424344 and increased contractility.4344 These changes may result in disturbances of the cerebral microcirculation because of microcirculatory stasis, and this may lead to “secondary” neuronal deterioration. Moreover, aged SHR have a reduced cerebral circulatory reserve to a transient drop of blood pressure because of a higher shift of autoregulation for cerebral blood flow45 and are more vulnerable to a transient ischemic episode.46 The CA1 regions of the hippocampus, striatum, and frontal and parietal cortices, where we found a reduced LCGU in aged SHR, are known to be vulnerable to ischemia. Second, long-lasting hypertension may increase the permeability of the blood-brain barrier,47 resulting in cerebral edema, and this may cause neuronal dysfunction. An increased incidence of cerebral edema has been reported in 8- to 28-month-old male SHR.48 Third, dysfunctions in the peripheral sensory organs of aged SHR, including the auditory49 and visual50 systems, probably due to microcirculatory impairments of these sensory organs, would cause diminished inputs to the central nervous system, and this may result in decreased neuronal activities. We therefore assume that the deterioration of brain function of aged SHR is due to a combination of “primary” neuronal dysfunction and “secondary” causes mainly resulting from cerebral microcirculatory disturbance.

    Behavioral-Metabolic Correlation

    The significant correlation between LCGU values and TECs in SHR (Table 5) does not necessarily mean a specific involvement of learning ability, because the structures with reduced LCGU were not limited to those structures primarily or secondarily relevant to the learning process. Rather, they were scattered in multifunctional systems, and the learning disability may have been one manifestation of the multifunctional dysfunctions. However, it is important to point out that learning ability is also involved more severely in the course of aging in the rats with long-standing hypertension.

    Some of the structures play a role in the learning process in the radial maze task: The hippocampus (CA1, CA3, or dentate gyrus)203051 is considered to be a key structure associated with memory. The medial septal nucleus projects cholinergic fibers to the hippocampus and composes hippocampal connections (septohippocampal systems).1352 The amygdala14 is considered to be another important structure of memory. The anteroventral thalamic nucleus composes the traditional Papez (1937) circuit (hippocampus–fornix–mamillary body–anterior thalamic nucleus–cingulate cortex–entorhinal cortex–hippocampus). The ventral tegmental area projects dopaminergic fibers to the hippocampus, amygdala, and frontal association cortex (mesolimbic system).5354 The parietal cortex plays an important role in processing information about space that is external to the body.55 Visual information is essential for spatial memory,1718 because a rat should recognize the directions of arms it has or has not visited. The basal nucleus of Meynert56 is involved in the cholinergic system, suggesting the cholinergic hypothesis for memory function.57 The striatum may be associated with spatial orientation58 and is involved in the general learning system.59 Damages of these brain structures may be associated with impaired learning ability in aged SHR.

    The present study clearly demonstrated a subclinical decrease of learning acquisition as one manifestation of psychosomatic deterioration during aging. This decrease was more marked in aged SHR than in aged WKY, probably because of the additional factor of long-standing hypertension, which leads to neuronal hypofunction as reflected by a decrease of the LCGU. Long-standing hypertension, therefore, seems to accelerate the aging process, manifesting an earlier decline in brain function and metabolism and shorter life span. Although the precise mechanism of the decrease in learning ability secondary to long-standing hypertension is not clear, our study suggests the importance of earlier control of hypertension for preventing or reducing brain dysfunction, including learning and memory disturbances. Further study is necessary to elucidate whether earlier control of hypertension has a favorable effect on secondary neuronal hypofunction in aged SHR.

    Reprint requests to Motohiro Kato, MD, Department of Clinical Neurophysiology, Neurological Institute, Faculty of Medicine, Kyushu University 60, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812, Japan.

          Figure 1.

    Figure 1. Line graphs show survival rates (a), body weights (b), systolic pressures (c), and heart rates (d) of Wistar-Kyoto rats (WKY) and spontaneously hypertensive rats (SHR) from 2 to 22 months of age. Values are mean±SEM. *P<.01 vs WKY.

          Figure 2.

    Figure 2. Line graphs show time course of changes in mean correct choice (a) and total error choice (b) in the eight-arm radial maze task in young-adult and aged Wistar-Kyoto rats (WKY) and spontaneously hypertensive rats (SHR). Correct and total error choices were observed in 13 young-adult WKY, 24 young-adult SHR, 12 aged WKY, and 20 aged SHR. Values are mean±SEM. *P<.05 vs young-adult WKY, +P<.01 vs young-adult SHR, #P<.05 vs aged WKY.

    Table 1. Number of Rats Used in Each Part of the Experiment

    (1) Survival301 (+192 =49)201 (+82 =28)50 (77)
    (2) Maze24 >(=192 +5)2013 >(=82 +5)1269
    Survival after maze219827
    (3) LCGU3555520
    Total: (1)+(2)54203312119

    SHR indicates spontaneously hypertensive rats; WKY, Wistar-Kyoto rats; and LCGU, local cerebral glucose utilization. Young-adult rats were 3 to 4 months old; aged, 16 to 17 months old.

    1Physiological indicators were measured in 12 of the 30 SHR and 12 of the 20 WKY until 22 months of age.

    2Survival rate was also observed in 19 young-adult SHR and 8 young-adult WKY after maze learning, and in totals of 49 SHR and in 28 WKY.

    3LCGU was measured after maze learning.

    Table 2. Learning Acquisition of Eight-Arm Radial Maze Task of Young-Adult and Aged WKY and SHR

    Measured VariableWKYSHR
    No. of rats13122420
    No. of rats with full acquisition (%)13 (100)7 (58)122 (92)4 (20)123
    Trials required for full acquisition11.5±1.425.4±3.3122.5±1.6123.0±3.3

    WKY indicates Wistar-Kyoto rats; SHR, spontaneously hypertensive rats. Young-adult rats were 3 to 4 months old; aged, 16 to 17 months old. Values are mean±SEM.

    1P<.05 vs young-adult WKY.

    2P<.05 vs young-adult SHR.

    3P<.05 vs aged WKY.

    Table 3. Physiological Parameters in WKY and SHR

    Young-Adult (n=5)Aged (n=5)Young-Adult (n=5)Aged (n=5)
    Body weight, g271±4387±101262±15362±92
    Mean arterial blood pressure, mm Hg115±4120±2176±51183±43
    Heart rate, bpm420±9384±8376±23368±18
    Pao2, mm Hg109±5106±4117±9103±3
    Paco2, mm Hg27±127±130±128±1
    Plasma glucose, mmol/L7.6±0.26.9±0.58.1±0.67.4±0.5

    WKY indicates Wistar-Kyoto rats; SHR, spontaneously hypertensive rats. Young-adult rats were 4 months old; aged, 16 to 17 months old. Values are mean±SEM.

    1P<.01 vs young-adult WKY.

    2P<.01 vs young-adult SHR.

    3P<.01 vs aged WKY.

    Table 4. Local Cerebral Glucose Utilization in WKY and SHR

    Cerebral cortex
    Basal ganglia
    Entopeduncular nucleus54.7±2.250.0±1.250.0±1.443.7±1.124
    Substantia nigra65.2±2.861.4±3.260.4±3.555.3±1.9
    Limbic structure
    Dentate gyrus69.7±3.061.7±3.460.9±1.558.7±3.1
    Olfactory tubercle127.4±5.199.9±4.91117.7±4.994.7±3.22
    Piriform cortex104.1±2.392.6±4.599.5±2.781.7±5.72
    Medial septal nucleus72.3±2.966.3±2.965.6±1.857.0±3.3
    Basal nucleus of Meynert98.3±3.485.5±4.989.5±0.982.0±4.2
    Mamillary nucleus, posterior130.1±4.9116.9±4.9117.7±4.9106.9±9.3
    Brain stem
    Superior colliculus103.9±5.792.0±1.794.8±1.878.9±3.72
    Ventral tegmental area71.6±1.167.5±2.163.7±2.656.2±2.434
    Cochlear nucleus138.0±5.7106.6±3.11124.6±7.298.8±4.52

    Definitions and ages are as in Table 2. Values are mean±SEM (micromoles per 100 g per minute) for five rats in each group.

    1P<.05 vs young-adult WKY.

    2P<.05 vs young-adult SHR.

    3P<.05 vs aged WKY.

    4P<.05 vs aged WKY using age-adjusted values of local cerebral glucose utilization.

    Table 5. Correlation Between Behavioral Scores (Total Errors) and Local Cerebral Glucose Utilization in 10 Young-Adult and Aged SHR

    Visual system
    Visual cortex−.845.01121
    Superior colliculus−.820.01391
    Limbic system
    Medial septal nucleus−.821.0138
    Olfactory tubercle−.821.0138
    Piriform cortex−.772.0206
    Anteroventral thalamic nucleus−.815.0145
    Ventral tegmental area−.723.0302
    Basal nucleus of Meynert−.771.0207
    Association system
    Frontal cortex−.705.0344
    Parietal cortex−.863.0096

    SHR indicates spontaneously hypertensive rats. Young-adult rats were 3 to 4 months old.

    1Significant using age-adjusted values of local cerebral glucose utilization.

    This study was supported in part by a grant from the Japan Health Sciences Foundation. We thank Kohei Akazawa, PhD, Medical Informatics, Kyushu University, for the statistical analysis.


    • 1 Waldstein SR, Manuck SB, Ryan CM, Muldoon MF. Neuropsychological correlates of hypertension: review and methodologic considerations. Psychol Bull.1991; 110:451-468. CrossrefMedlineGoogle Scholar
    • 2 Wilkie F, Eisdorfer C. Intelligence and blood pressure in the aged. Science.1971; 172:959-962. CrossrefMedlineGoogle Scholar
    • 3 Hertzog C, Schaie KW, Gribbin K. Cardiovascular disease and changes in intellectual functioning from middle to old age. J Gerontol.1978; 33:872-883. CrossrefMedlineGoogle Scholar
    • 4 Farmer ME, White LR, Abbott RD, Kittner SJ, Kaplan E, Wolz MM, Brody JA, Wolf PA. Blood pressure and cognitive performance: The Framingham Study. Am J Epidemiol.1987; 126:1103-1114. CrossrefMedlineGoogle Scholar
    • 5 Campbell RJ, Di Cara LV. Running-wheel avoidance behavior in the Wistar/Kyoto spontaneously hypertensive rat. Physiol Behav.1977; 19:473-480. CrossrefMedlineGoogle Scholar
    • 6 Danysz W, Pláznik A, Pucilowski O, Plewako M, Obersztyn M, Kostowski W. Behavioral studies in spontaneously hypertensive rats. Behav Neural Biol.1983; 39:22-29. CrossrefMedlineGoogle Scholar
    • 7 Knardahl S, Karlsen K. Passive-avoidance behavior of spontaneously hypertensive rats. Behav Neural Biol.1984; 42:9-22. CrossrefMedlineGoogle Scholar
    • 8 Knardahl S, Sagvolden T. Two-way active avoidance behavior of spontaneously hypertensive rats: effect of intensity of discontinuous shock. Behav Neural Biol.1982; 35:105-120. CrossrefMedlineGoogle Scholar
    • 9 Randich A, Maixner W. Acquisition of conditioned suppression and responsivity to thermal stimulation in spontaneously hypertensive, renal hypertensive and normotensive rats. Physiol Behav.1981; 27:585-590. CrossrefMedlineGoogle Scholar
    • 10 Sutterer JR, Perry J, DeVito W. Two-way shuttle box and lever-press avoidance in the spontaneously hypertensive and normotensive rat. J Comp Physiol Psychol.1980; 94:155-163. CrossrefMedlineGoogle Scholar
    • 11 Rogers LJ, Sink HS, Hambley JW. Exploration, fear and maze learning in spontaneously hypertensive and normotensive rats. Behav Neural Biol.1988; 49:222-233. CrossrefMedlineGoogle Scholar
    • 12 Wyss JM, Fisk G, van Groen T. Impaired learning and memory in mature spontaneously hypertensive rats. Brain Res.1992; 592:135-140. CrossrefMedlineGoogle Scholar
    • 13 Olton DS, Walker JA, Gage FH. Hippocampal connections and spatial discrimination. Brain Res.1978; 139:295-308. CrossrefMedlineGoogle Scholar
    • 14 Ermakova IV, Loseva EV, Valouskova V, Bures J. The effect of embryonal amygdala grafts on the impairment of spatial working memory elicited in rats by kainate-induced amygdaloid damage. Physiol Behav.1989; 45:235-241. CrossrefMedlineGoogle Scholar
    • 15 Hall RD, Macrides F. Olfactory bulbectomy impairs the rat’s radial-maze behavior. Physiol Behav.1983; 30:797-803. CrossrefMedlineGoogle Scholar
    • 16 Stokes KA, Best PJ. Mediodorsal thalamic lesions impair radial maze performance in the rat. Behav Neurosci.1988; 102:294-300. CrossrefMedlineGoogle Scholar
    • 17 Dean P, Key C. Spatial deficits on radial maze after large tectal lesions in rats: possible role of impaired scanning. Behav Neural Biol.1981; 32:170-190. CrossrefMedlineGoogle Scholar
    • 18 Foreman N, Stevens R. Visual lesions and radial maze performance in rats. Behav Neural Biol.1982; 36:126-136.CrossrefMedlineGoogle Scholar
    • 19 Wei L, Lin SZ, Tajima A, Nakata H, Acuff V, Patlak C, Pettigrew K, Fenstermacher J. Cerebral glucose utilization and blood flow in adult spontaneously hypertensive rats. Hypertension.1992; 20:501-510. LinkGoogle Scholar
    • 20 Olton DS, Samuelson RJ. Remembrance of places passed: spatial memory in rats. J Exp Psychol Anim Behav Proc.1976; 2:97-116. CrossrefGoogle Scholar
    • 21 Hosokawa S, Kato M, Shima F, Tobimatsu S, Kuroiwa Y. Local cerebral glucose utilization altered in rats with unilateral electrolytic striatal lesions and modification by apomorphine. Brain Res.1984; 324:59-68. CrossrefMedlineGoogle Scholar
    • 22 Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak CS, Pettigrew KD, Sakurada O, Shinohara M. The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem.1977; 28:897-916. CrossrefMedlineGoogle Scholar
    • 23 Takei H, Fredericks WR, Rapoport SI. The lumped constant in the deoxyglucose procedure declines with age in Fischer-344 rats. J Neurochem.1986; 46:931-938. CrossrefMedlineGoogle Scholar
    • 24 Paxinos G, Watson C. The Rat Brain in Stereotaxic Coordinates. North Ryde, Australia: Academic Press; 1986. Google Scholar
    • 25 Dixon WJ. Description of groups (Strata) with histograms and analysis of variance. In: Dixon WJ, ed. BMDP Statistical Software. Berkeley, Calif: University of California Press; 1983:105-115. Google Scholar
    • 26 Sasagawa S, Yamori Y. Quantitative analysis on the behavior of spontaneously hypertensive rats (SHR) and stroke-prone SHR. Jpn Heart J.1975; 16:313-315. CrossrefMedlineGoogle Scholar
    • 27 Eichelman B, Dejong W, Williams RB. Aggressive behavior in hypertensive and normotensive rat strains. Physiol Behav.1973; 10:301-304. CrossrefMedlineGoogle Scholar
    • 28 Knardahl S, Chindaduangratn C. Residential-maze behavior of spontaneously hypertensive rats. Behav Neural Biol.1984; 41:84-89. CrossrefMedlineGoogle Scholar
    • 29 Ingram DK, London ED, Goodrick CL. Age and neurochemical correlates of radial maze performance in rats. Neurobiol Aging.1981; 2:41-47. CrossrefMedlineGoogle Scholar
    • 30 Kadar T, Silbermann M, Brandeis R, Levy A. Age-related structural changes in the rat hippocampus: correlation with working memory deficiency. Brain Res.1990; 512:113-120. CrossrefMedlineGoogle Scholar
    • 31 Kobayashi S, Kametani H, Ugawa Y, Osanai M. Age difference of response strategy in radial maze performance of Fischer-344 rats. Physiol Behav.1988; 42:277-280. CrossrefMedlineGoogle Scholar
    • 32 Kadekaro M, Savaki HE, Kutyna FA, Davidsen L, Sokoloff L. Metabolic mapping in the sympathetic ganglia and brain of the spontaneously hypertensive rat. J Cereb Blood Flow Metab.1983; 3:460-467. CrossrefMedlineGoogle Scholar
    • 33 Hayashi T, Nakamura K. Cerebral neuronal activity in spontaneously hypertensive rats as demonstrated by the 14C-deoxyglucose method. Naunyn Schmiedebergs Arch Pharmacol.1981; 316:331-339. CrossrefMedlineGoogle Scholar
    • 34 Ritter S, Dinh TT. Progressive postnatal dilation of brain ventricles in spontaneously hypertensive rats. Brain Res.1986; 370:327-332. CrossrefMedlineGoogle Scholar
    • 35 Bendel P, Eilam R. Quantitation of ventricular size in normal and spontaneously hypertensive rats by magnetic resonance imaging. Brain Res.1992; 574:224-228. CrossrefMedlineGoogle Scholar
    • 36 Nelson DO, Boulant JA. Altered CNS neuroanatomical organization of spontaneously hypertensive (SHR) rats. Brain Res.1981; 226:119-130. CrossrefMedlineGoogle Scholar
    • 37 Tajima A, Hans FJ, Livingstone D, Wei L, Finnegan W, DeMaro J, Fenstermacher J. Smaller local brain volumes and cerebral atrophy in spontaneously hypertensive rats. Hypertension.1993; 21:105-111. LinkGoogle Scholar
    • 38 Soncrant TT, Holloway HW, Stipetic M, Rapoport SI. Cerebral glucose utilization in rats is not altered by hindlimb restraint or by femoral artery and vein cannulation. J Cereb Blood Flow Metab.1988; 8:720-726. CrossrefMedlineGoogle Scholar
    • 39 London ED, Nespor SM, Ohata M, Rapoport SI. Local cerebral glucose utilization during development and aging of the Fischer-344 rat. J Neurochem.1981; 37:217-221. CrossrefMedlineGoogle Scholar
    • 40 Smith CB, Goochee C, Rapoport SI, Sokoloff L. Effects of ageing on local rates of cerebral glucose utilization in the rat. Brain.1980; 103:351-365. CrossrefMedlineGoogle Scholar
    • 41 Nordborg C, Fredriksson K, Johansson BB. The morphometry of consecutive segments in cerebral arteries of normotensive and spontaneously hypertensive rats. Stroke.1985; 16:313-320. CrossrefMedlineGoogle Scholar
    • 42 Baumbach GL, Heistad DD. Cerebral circulation in chronic arterial hypertension. Hypertension.1988; 12:89-95. LinkGoogle Scholar
    • 43 Lüscher TF, Diederich D, Weber E, Vanhoutte PM, Bühler FR. Endothelium-dependent responses in carotid and renal arteries of normotensive and hypertensive rats. Hypertension.1988; 11:573-578. LinkGoogle Scholar
    • 44 Mayhan WG, Faraci FM. Cerebral vasoconstrictor responses to serotonin during chronic hypertension. Hypertension.1990; 15:872-876. LinkGoogle Scholar
    • 45 Fujishima M, Sadoshima S, Ogata J, Yoshida F, Shiokawa O, Ibayashi S, Omae T. Autoregulation of cerebral blood flow in young and aged spontaneously hypertensive rats (SHR). Gerontology.1984; 30:30-36. CrossrefMedlineGoogle Scholar
    • 46 Yao H, Sadoshima S, Ooboshi H, Sato Y, Uchimura H, Fujishima M. Age-related vulnerability to cerebral ischemia in spontaneously hypertensive rats. Stroke.1991; 22:1414-1418. CrossrefMedlineGoogle Scholar
    • 47 Knox CA, Yates RD, Chen I, Klara PM. Effects of aging on the structural and permeability characteristics of cerebrovasculature in normotensive and hypertensive strains of rats. Acta Neuropathol (Berl).1980; 51:1-13. CrossrefMedlineGoogle Scholar
    • 48 Wexler BC, McMurtry JP, Iams SG. Histopathologic changes in aging male vs female spontaneously hypertensive rats. J Gerontol.1981; 36:514-519. CrossrefMedlineGoogle Scholar
    • 49 Tachibana M, Yamamichi I, Nakae S, Hirasugi Y, Machino M, Mizukoshi O. The site of involvement of hypertension within the cochlea: a coparative study of normotensive and spontaneously hypertensive rats. Acta Otolaryngol (Stockh).1984; 97:257-265. CrossrefMedlineGoogle Scholar
    • 50 Rogers LJ, Bolden SW, Patrech AS, Ehrlich D. Visual dysfunction in the spontaneously hypertensive rat. Physiol Behav.1993; 54:903-907. CrossrefMedlineGoogle Scholar
    • 51 Geinisman Y, de Toledo Morrell L, Morrell F. Aged rats need a preserved complement of perforated axospinous synapses per hippocampal neuron to maintain good spatial memory. Brain Res.1986; 398:266-275. CrossrefMedlineGoogle Scholar
    • 52 Yamada S. The effects of lesions in the medial septal area to spatial learning in rats [in Japanese]. Fukuoka Igaku Zasshi.1993; 84:480-487. MedlineGoogle Scholar
    • 53 Lee JM, Ross ER, Gower A, Paris JM, Martensson R, Lorens SA. Spatial learning deficits in the aged rat: neuroanatomical and neurochemical correlates. Brain Res Bull.1994; 33:489-500. CrossrefMedlineGoogle Scholar
    • 54 Brozoski TJ, Brown RM, Rosvold HE, Goldman PS. Cognitive deficit caused by regional depletion of dopamine in prefrontal cortex of Rhesus monkey. Science.1979; 205:929-932. CrossrefMedlineGoogle Scholar
    • 55 DiMattia BD, Kesner RP. Spatial cognitive maps: differential role of parietal cortex and hippocampal formation. Behav Neurosci.1988; 102:471-480. CrossrefMedlineGoogle Scholar
    • 56 Hepler DJ, Wenk GL, Cribbs BL, Olton DS, Coyle JT. Memory impairments following basal forebrain lesions. Brain Res.1985; 346:8-14. CrossrefMedlineGoogle Scholar
    • 57 Bartus RT, Dean RL III, Beer B, Lippa AS. The cholinergic hypothesis of geriatric memory dysfunction. Science.1982; 217:408-417. CrossrefMedlineGoogle Scholar
    • 58 Potegal M. Role of the caudate nucleus in spatial orientation of rats. J Comp Physiol Psychol.1969; 69:756-764. CrossrefMedlineGoogle Scholar
    • 59 Thompson R. Centrencephalic theory, the general learning system, and subcortical dementia. Ann N Y Acad Sci.1993; 702:197-223.CrossrefMedlineGoogle Scholar


    eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

    Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.