12.2AGING

Go to:
Age-associated cognitive impairment has been described in a variety of species, including rats, macaque monkeys and humans.172–175 In this second part of the chapter, I will review the main memory alterations that characterize cognitive decline associated with aging in humans and experimental animals (notably rodents). In each case, the neurobiological mechanisms linked to such declines will follow the phenomenological descriptions.

12.2.1Memory Deficits in Aging Human Population

As stated above, there are considerable individual differences in the course of aging, with particularly large variation occurring in humans.148 Establishing what represents normal cognitive decay is complicated by the difficulties of distinguishing stable mild impairments and deficits related to early symptoms of neurodegenerative diseases such as Alzheimer’s disease that show progressive deteriorations of brain function and behavior.174 In fact, most aged humans experience some form of age-related neural pathology such as Alzheimer’s disease (AD), Parkinson’s disease, diabetes, hypertension, and arteriosclerosis. Other difficulties for determining the cognitive alterations due to aging are the limitations intrinsic to the types of studies that can be done with human subjects. Instead of providing proper experimental evidence, studies on aged human subjects normally provide only correlational evidence and therefore cannot be considered highly conclusive. Moreover, these studies are frequently based on cross-sectional evaluations of individuals of different age groups. The limitation relates to comparing groups that may differ in the sociological impacts of living their respective life periods during different decades. However, the recent trend is to perform longitudinal studies, most of the current ones focusing on longitudinal changes occurring after the age of 60.
However, normal aging is also associated with changes in the neural basis of cognition. Regardless of individual differences, aging influences certain memory types and cognitive fields more than others. In general terms, as indicated by both cross-sectional and longitudinal studies, aging is characterized by considerable reductions in certain capacities:176–179
  • Speed of information processing
  • Working memory
  • Formation of new episodic memories
  • Spatial learning
Other abilities such as emotional processing, short-term memory, autobiographical memory, semantic knowledge, and priming remain relatively intact.174–180 Cumulative knowledge suggests that the identified memory deficits are mainly the consequences of age-related changes in two types of cognitive processes:
  • Disrupted executive functions that eventually exert major consequences on a variety of memory functions. The importance of executive function for memory is mainly related to the controlled processing frequently required during the encoding (particularly when strategic elaboration is required) and retrieval (when an active searching strategy is required) of information. For example, one cognitive process that is particularly dependent on executive processes that are disrupted in aging is the recall of the source of information and temporal details of past episodes.
  • Decay of long-term declarative memory.174
Recent findings suggest that the personal appraisal of the changes that come with aging is an important factor that determines who is not greatly impaired by aging and who deteriorates rapidly. Wellbeing and a positive view of aging seem to act as major protective factors against the detrimental effects of age, not only on brain and cognitive function, but also at a more general level of the organism.181 It seems that, at odds with older adults showing rapid declines, those who are not much impaired in their cognitive abilities may show compensation for brain decline in aging that involves increased recruitment of brain activity during cognitive performance.

12.2.2Neurobiological Mechanisms Associated with Age-Related Cognitive Decline in Humans

There is great interest in understanding the neurobiological mechanisms that underlie memory decline occurring at aging and identifying the factors that determine differential impacts of aging on various cognitive domains and on different individuals. In agreement with the behavioral alterations observed in executive function and declarative memory, neuroimaging studies have shown that age-related cognitive deficits are linked to multiple structural and functional changes in the frontal–striatal circuits, medial temporal lobe (MTL), regions and white matter tracts.174
Thus, the deficits of executive function observed in the nondemented aged population have been associated with alterations in frontal–striatal circuits. A variety of pathophysiological changes that have been reported to occur in frontal–striatal areas in the aged population may account for the reported executive difficulties.173
At the structural level, multiple changes including atrophy of frontal grey matter and striatal volume loss have been reported. Neurotransmitter systems can also experience considerable alteration during the aging process. An age-associated decline in dopamine content, for example, appears to be associated with executive impairments.
Frontal white matter appears to be particularly susceptible to age-related damage (showing diffuse changes and small infarcts), and a link with the degree of cognitive impairment has been established in studies linking behavioral testing with structural magnetic resonance imaging (MRI) evaluating white matter lesions. This latter pathology seems to be related to problems in vascular function (mainly hypertension) that appear to have a special impact on white matter structures supporting frontal–striatal circuits.
On the other hand, the characteristic alterations of long-term declarative memory occurring during aging have been linked to age-related changes in the MTL, including the hippocampus and adjacent regions. The MTL is strongly affected in AD (from its earlier stages), with a number of pathophysiological features characterizing the damage to these structures. These include atrophy, cell loss, and cellular damage, and are consistently associated with marked memory deficits. More specifically, cellular pathology in AD is linked to abnormal extracellular deposition of amyloid protein and intracellular accumulations of tau.182
Substantial evidence supports a key role of deposits (plaques) and soluble forms of amyloid on the triggering of neuronal dysfunction and eventual cell death. Such deposits also lead to neurofibrillary tangles that represent a major pathology in the MTL and eventually spread to associated cortex. In AD, the symptoms progress to the eventual overall impairment as the disease advances. Recent imaging studies suggest that what may account for the memory impairment observed in this disease is the disruption of a network of connections including the MTL and other areas, notably the precuneus, extending into retrosplenial and posterior cingulate cortex.173 In any case, it is important to note that the circuits that degenerate in AD are also vulnerable to normal aging, but the vulnerability is reflected by compromised synaptic communication rather than by neuron death.183
One interesting feature indicated by functional imaging studies of non-demented old individuals is that unique patterns of brain activation distinguish older individuals showing high-performance in cognitive tasks from younger adults.184 A subset of older adults showed increased recruitment of brain areas that has been interpreted as a potential compensatory response to increasing task difficulty.173 They may require the use of additional brain resources to guarantee a certain performance level when other physiological alterations interfere with their cognitive functions. This type of compensatory process has been proposed to play a role in individual differences in cognitive decline during the course of aging.

12.2.3Memory Deficits in Aged Rodents

Research on experimental animals is essential for gaining insight into what is normal cognitive decline associated with aging and what is pathological. It is also necessary to our understanding of the relative involvement of different factor with age differences in cognition. Most commonly, rodents are used to characterize age-related alterations in memory processes and ascertain the neurobiological processes underlying such cognitive deficits.
Although aged rodents display a variety of cognitive deficits, a large part of the research on this topic has focused on the hippocampus and spatial learning. Before reviewing that issue, we will deal with methodological aspects that are relevant to research in this area, then present a brief discussion of the research carried out in rodents to explore the degree of alteration on frontal lobe functions in these animal species.

12.2.3.1Methodological Aspects of Aging Research in Rodents

Given the relatively short life-spans of rodents (normally 2 to 4 years), they are particularly appropriate for longitudinal studies that are ideal for obtaining aging curves and collecting information about essential factors contributing to developmental decline. However, they are also the exceptions rather than the rules in animal research because they are both expensive and time-consuming. The most frequent approach, as in human studies, is the use of cross-sectional comparisons of groups of animals of different ages, typically including young adults and older individuals.185
The study of aging involves a number of difficulties that are particularly relevant when the focus of research is cognition.186 Aging is generally associated with changes in sensorimotor abilities and motivation, factors that can impact the performances of animals in learning and memory tasks but should be distinguished from putative impairments in cognitive performance.187 Particularly, visual competence can be highly degraded in aging rats, an aspect that should be specially controlled when studying animal performance of tasks with visual components.188
Another factor that requires special attention is that rodents that have been maintained undisturbed in their home cages during the course of their lives may not be appropriate subjects for cognitive testing at old age. Rodents raised in animal houses are normally not confronted with environmental challenges. Therefore, their organisms had no opportunities to adapt and to develop behavioral and physiological strategies relevant for successful performance of many learning and memory tasks.187 One solution proposed to overcome this problem is to raise and house rats in enriched environments.

12.2.3.2Alterations in Frontal Lobe Function in Aged Rodents

Most animal research that has addressed the behavioral alterations associated with frontal lobe dysfunction has been performed on non-human primates. The cognitive deficits observed (deficits in delayed response testing, increased perseveration, difficulties in reversal learning, etc.) were strikingly similar to those reported in aged humans and in young nonhuman primates with frontal lesions.189 However, a more limited number of studies in old rats could also detect similar cognitive impairments that were also comparable to those induced in younger rats by specific frontal lobe lesions. Using different behavioral testing procedures (notably delayed nonmatch to sample), clear evidence was obtained that the temporal organization of memory is significantly disrupted in aged rats, in a similar way as that observed in younger rats with prefrontal cortical damage.190,191
Evidence for impaired cognitive flexibility mediated by prefrontal circuits in aged rats has been provided using an attentional set-shifting task. Barense et al.192 trained young and aged male rats on two problems. The reward was always associated with the same stimulus dimension (for example, they had to link the reward to a particular odor) and a reversal of one problem (for example, they had to make a new association because the reward was predicted by an alternative odor and not by the former odor). Then, a new problem was presented in which the reward was consistently associated with the previously irrelevant stimulus dimension (extradimensional shift or EDS). For example, odors no longer predicted the reward; the digging medium in which the reward was hidden predicted it. Aged rats were significantly impaired on the EDS, although some individual aged rats performed as well as young rats on this phase. Moreover, some aged rats were impaired on the reversal. These deficits of the EDS paralleled those manifested by young rats submitted to neurotoxic lesions of medial frontal cortex. The impairment of rapid reversal learning observed in aged rats was linked to orbitofrontal cortex dysfunction.193

12.2.3.3Alterations in Medial Temporal Lobe–Hippocampal Function in Aged Rodents

Due to the great interest in understanding the mechanisms underlying hippocampal dysfunction at aging, a large number of studies focused in characterizing the performance of aged rodents in spatial learning tasks. For reviews see References 194 through 196. Age-related spatial learning deficits were reported, for example, in the radial-arm maze. Aged rats were slower than younger adult rats in learning to this task,197–199 an effect that is clearly dependent on the requirement to develop a spatial strategy since aged rats were shown to be impaired in nonspatial reference memory versions of the radial-arm maze.200
Consistent deficits in learning, memory, and the acquisition of new response solutions have also been found in aged rodents trained in the Barnes circular platform task,201 in which animals learn to identify which of 18 holes distributed along the perimeter of a circular platform allows them access to a tunnel to escape eventually from exposure to light.151,201 Similar age-related deficits have also been reported in the Morris water maze spatial learning task.202–205 Aged rats normally take longer to learn the location of the hidden platform, while they show no signs of impairment when trained in a cued platform version.203,206
An assessment of hippocampal-dependent spatial learning and memory capabilities of healthy aged rodents revealed striking individual differences.207–209 For example, the water maze task revealed the existence of important individual differences in spatial memory abilities within old rats.152,207,210–212 While some animals show clear deficits in spatial memory, others perform similarly to younger animals and represent a very interesting tool for investigating the neurobiological substrates of cognitive aging (see below).

12.2.4Aging and Structural and Functional Plasticity

Based on the well reported individual differences in cognitive aging, one of the most popular strategies in current research is to first characterize aged animals in a learning task to subsequently investigate neurobiological correlates of the observed learning and memory deficits.
A pioneer study showed in aged rats (22 to 24 months) a correlation between the degree of decline in performance in learning and place navigation tasks and brain energy metabolism (evaluated as regional glucose utilization) in 5 of 45 brain regions examined: dentate gyrus, medial septum-diagonal band area, hippocampal CA1, hippocampal CA3, and prefrontal cortex. Learning impairments in the aged rats were related to the extent of decrease in glucose utilization in restricted areas of the limbic system.213

12.2.4.1Structural and Neurochemical Alterations

The literature contains controversy as to whether normal aging is accompanied by a loss of neurons214,215 because the most recent findings seem not to confirm earlier reports indicating such cell death. However, consensus is greater on the view that alterations in relevant neurocircuits may underlie age-related cognitive deficits.183
Human and monkey studies reported regressive changes with age in dendritic arbors and spines of cortical pyramidal neurons in specific regions and layers of the frontal lobe.216–218 Evidence of degeneration in the PFC was found both in old monkeys and humans, as indicated by drastic alterations in the morphology of terminal dendrites and reduction of synaptic and spine densities.183
Synaptic alterations are believed to be associated with changes in the expression levels of glutamate receptors, with available evidence indicating decreases in N-methyl-D-aspartic acid (NMDA; particularly the NR2B subunit) and -amino-5-hydroxy-3-methyl-4-isoxazole propionic acid (AMPA) receptors in older individuals.219,220 In addition, degeneration of myelinated axons in both deep cortical layers and white matter has been reported to correlate with sensory and cognitive capabilities in old animals.221
At the MTL, the hippocampus is the brain area more deeply studied. Using unbiased stereological methods, Geinisman et al.222 reported a decrease in the number of axospinous synapses in the mid-molecular layer of the dentate gyrus of aged rats (28 months) that was hypothesized to underlie reductions in the amplitude of excitatory postsynaptic potentials and the decline in functional synaptic plasticity detected in the dentate gyrus of senescent rats.
The cholinergic and monoaminergic systems that project from the basal forebrain and brainstem also displayed functional impairments in aging.223 Interestingly, signal transduction pathways seem to be differentially regulated in the aged hippocampus and PFC. Whereas activation of the cAMP/protein kinase A (PKA) pathway has been proposed as a mechanism for improving age-related hippocampus-related cognitive deficits, agents that increase PKA activity impair — instead of improving —prefrontal cortical function in aged rats and monkeys with prefrontal cortical deficits. Conversely, PKA inhibition was shown to ameliorate prefrontal cortical cognitive deficits.224 These findings further illustrate the complexity and difficulty in understanding the mechanisms affecting cognitive function in the aged brain.

12.2.4.2Functional Alterations

There is controversy in the literature as to whether aged animals show deficits in hippocampal LTP.196 In general terms, age-related LTP-induction deficits are mainly found when the induction protocols involve low-intensity stimulation, but no consistent alterations are observed when high-intensity and robust stimulation is applied.196,201 Moreover, the threshold for LTP induction is increased in aged rats, which may be related to the greater difficulties displayed by aged rats to encode memories. As to LTP maintenance, whenever high-intensity stimulation has been used, age-related maintenance deficits appear at late recording time points,196,201 LTP maintenance deficits have been correlated with impaired performance in hippocampus-dependent learning tasks, including the Barnes circular platform task,151
As to long-term depression (LTD) and depotentiation, in contrast to LTP, these are more readily produced in aged than in adult rats.196 A recent study224 investigated whether LTD in area CA1 is related to individual differences in learning abilities in the outbred Long-Evans rat strain. Young rats exhibited larger NMDAR-dependent LTD (NMDAR-LTD) than the aged animals (24 months), and no differences were found between the aged unimpaired and the aged impaired groups. When an NMDAR-independent form of LTD (non-NMDAR-LTD) was examined, the aged unimpaired group showed significantly larger non-NMDAR-LTD than either the young or the aged impaired groups.
The authors also found a significant correlation between the magnitude of non-NMDAR-LTD and learning abilities in aged, but not in young, rats. This study suggests that high-performing aged rats maintain the ability to generate LTD through mechanisms different from those used by young adults, whereas aged animals that fail to make a switch to the mechanisms that mediate LTD will be impaired in learning performance.
Interestingly, variability in escape and spatial learning in the water maze in the aged unimpaired (outbred male Wistar rats 28 to 30 months old), but not in aged impaired (selected from a large pool based on water maze escape performance over a 9-day period) group was correlated with variability in short-term and long-term potentiation.152

12.2.4.3Aged Hippocampus and Place Cells

Recent evidence indicates that the older hippocampus may also be slower to switch between cognitive maps and that such failure to switch between hippocampal maps in time may account for their impaired spatial performance.225 Spatial abilities in rodents have been largely related to hippocampal neurons called place cells that encode spatial information defined by visual landmarks226 or by self-motion cues.227 A cognitive map of an animal’s environment would be formed by a population of place cells activated by multiple cues on that particular environment.228 Rosenzweig et al.225 found that the ability of rats to find a reward in a particular environment is correlated with the ability of place cells to switch between two different cognitive maps, one based on self-motion cues that are unrelated to the task and another based on relevant landmark cues. Interestingly, old rats were impaired relative to young adult rats, both in switching from the irrelevant to the relevant map and in finding the reward.