There are two ways to assess changes in brain structure in the senescent brain: in vitro studies of postmortem tissue and in vivo studies of the living brain using techniques such as X-ray CT or MRI. Until about 1990, most in vivo studies used X-ray CT, whereas most current investigations generally employ MRI because of its better spatial resolution and contrast. Studies of structural changes at the cellular and subcellular level require the use of either light or electron microscopy and generally are performed on postmortem tissue that has been chemically fixed.
A. Gross Anatomical Changes
Brain weight is known to decrease with age. Changes begin in the third and fourth decades and show a progressive decline throughout the life span. At age 20, the average weight of the male brain is approximately 1400 g, and by the age of 65 brain weight is approximately 1300 g. Brain weight for females follows a similar trend, although the total weight is 100-150 g less than that of males. Most changes occur after age 55, with a total loss of up to 15% of the peak brain weight by age 90.
Shrinkage of brain tissue or atrophy also occurs with age. Atrophy is clearly seen along the surface of the cerebral hemispheres where the gyri or cortical ridges become progressively more narrow and the spaces between the gyri, referred to as sulci, widen. Mild to moderate cerebral atrophy occurs in a heterogeneous fashion, with frontal, parasagittal, and temporal regions affected more than other areas of the cortex. The cerebral ventricles also dilate with age, typically becoming apparent in the 60s. Severe atrophy is generally associated with disease processes. Changes in brain volume are also associated with aging and begin after age 50, with a 2 or 3% progressive decrease per decade. Until age 50, a decrease in volume is predominantly observed in the gray matter of the cerebral cortex. After age 50, decreases in white matter are greater. Autopsy data suggest that the volume of white matter is decreased in old (75-85 years) compared to young subjects by approximately 11%.
Both cross-sectional and longitudinal aging studies of changes in brain structure have been performed on healthy subjects using either X-ray CT or MRI. Our focus is mainly on the results of the MRI studies because of the better spatial resolution and contrast obtainable with this method. One can manipulate the scanning parameters of a MRI device in several ways, each of which emphasizes signals corresponding to somewhat different features of brain tissue. In this overview, we discuss senescent changes in volumetric measures of gray matter, white matter, cerebrospinal fluid (CSF) space, and a few fairly well-defined brain regions, such as the hippocampus and the basal ganglia. The absence of clearly and easily identifiable landmarks makes it difficult to measure the volumes of specific regions of the neocortex in a rigorous way. We also review some of the findings about age-related alterations in white matter hyperintensities.
a. Some Fundamentals Concerning Structural Brain Imaging Before discussing the specific structural brain imaging findings, it is worthwhile to provide some information about how these types of data are acquired. Both CT and MRI generate images of the brain in the form of slices parallel to one another. With MRI, these slices can be parallel to any plane that one chooses. The slices have a certain thickness and may or may not be contiguous. Recent imaging studies have tended to use quite thin slices (about 1 mm thick).
There are essentially two distinct methods that have been used to evaluate brain volumetrics on CT or MRI images. The first involves determining the volumes of specific brain regions by manually tracing their areas on individual slices. The second method uses some type of computer algorithm to segment automatically each image into specific tissue types (e.g., gray matter, white matter, and CSF). Each method has limitations. The trace method is more subjective, although it enables the tracer to make use of his or her knowledge of neuroanatomy. The most precise results are obtained with thin slices, but the analysis can be quite time-consuming. The segmentation method has the advantage of objectivity, but it can be especially susceptible to the partial volume problem: Even a single pixel (the smallest element in an image) may contain a mixture of brain tissues, and thus it becomes difficult to categorize every element in the image as belonging to one kind of tissue versus a second. In an image in which the blackest pixels correspond to one tissue type and the whitest to a second, gray pixels may correspond either to a third tissue type or may represent the partial volume averaging of the first two types. As investigations of this sort have continued, improvements in both the trace and segmentation methods have been devised, but the types of problems indicated previously persist.
One last technical point needs to be mentioned. Because the size of an individual's brain and its components is related to the size of the subject (and thus, for example, men have on average larger brains than women), almost all studies of brain volumetrics use normalized volumes; for instance, a commonly used volumetric measure is the percentage of the intracranial volume of the subject.
b. Age-Related Volumetric Changes There seems to be essentially uniform agreement that the total amount of brain tissue decreases with advancing age. Also, all investigations have reported an increase in the volume of CSF in the brain of elderly compared to young subjects. When these changes begin and whether the loss of brain tissue corresponds primarily to gray matter, white matter, or both have been contentious issues. Several investigations have indicated that gray matter shows a gradually accelerating decline with age, but after approximately age 50 the white matter of the brain shows a more pronounced age-related decrease and seems to be the main contributor to the age-related loss of brain volume. Also reasonably clear is that these changes are different for men and women. Whereas the increase with age in ventricular CSF is the same for the two sexes, men show a greater increase in peripheral CSF (i.e., subarachnoid CSF) than do women. Peripheral CSF is considered to be a marker of cortical atrophy.
The frontal lobes have been shown to decline in volume with advancing age, with some research groups reporting that the loss of tissue is greater in men than in women. The parietal lobes have also been shown to be reduced in aging, but to a greater extent in women compared to men. Some groups have found no age-related change in the volume of the temporal lobes, whereas others have; the different findings could be attributable to whether or not the subjects under study represented successful (no change in temporal lobe volume) versus typical aging. Some studies have also examined the size of the corpus callosum (the large bundle of nerve fibers connecting the left and right cerebral hemispheres) as a function of age. Several of these have found that the anterior portion of the corpus callosum shows age-related atrophy, which is consistent with the age-related reduction in frontal lobe volume.
Most brain structures that are measurable by manual tracing with MRI have been found to show age-related decreases in volume. These include basal ganglia structures, such as the caudate and lenticular nuclei, and the anterior portion of the thalamus. Many studies have examined the hippocampus and other components of the medial temporal lobe. Most, although not all, investigations have reported a decrease in the size of the hippocampus with advancing age. In one longitudinal study of individuals ranging in age from 70 to 89 years, it was found that the hippocampus decreased in volume by approximately 1.5% per year.
In summary, although there is still much debate in the scientific literature, it seems that most brain structures show a measurable decrease in size with advancing age. However, compared to neurodegen-erative diseases such as Alzheimer's disease, these age-related decreases are small. For example, the study that reported the 1.5% annual rate of hippocampal atrophy also found that this rate was approximately 4% in patients with Alzheimer's disease. Moreover, these changes generally do not begin until the sixth decade. The most notable feature of these data is the increase in variance with advancing age. That is, there are elderly individuals with values within the young normal range, even though some older subjects show significant atrophy. This aspect of aging is not restricted to brain volumetrics, however; as mentioned previously, it essentially typifies most quantitative studies of aging. Finally, it should be noted that the relationship of these measures of increased atrophy to cognitive decline is unclear. Very few investigations have examined the correlation between cognitive decline and brain volumetrics, especially in a longitudinal design.
c. White Matter Hyperintensities With the advent of MRI, abnormal signals were observed in the white matter in a number of neurological diseases known to affect the white matter (e.g., multiple sclerosis and Binswanger's disease), in various dementias, as well as in the elderly. Because these MRI signals are often best observed using scanning parameters that result in their appearing as bright lucencies against a black (low-signal) white matter, they have been termed white matter hyperintensities (WMHs). Two kinds have been distinguished: (i) Periventricular white matter hyperintensities (PWMHs) appear either as frontal or occipital caps of the cerebral ventricles or as a thin lining surrounding the ventricles, and (ii) deep white matter hyperintensitites (DWMHs) are seen as subcortical punctate foci, although larger confluences of foci form with increased severity and often merge with the PWMHs. The neuropathological substrate for these signals can vary depending on the disease. In aging, the PWMH most likely results from the breakdown of the ventricular ependyma, which leads to an increase in the water content of the nearby myelin, demyelination, and reactive gliosis. DWMHs likely correspond to gliosis, demyelination, and atrophy and shrinkage of axons and myelin around blood vessels. Because DWMHs are found in patients with cerebro-vascular disease, vascular dementia, and hypertension, their appearance in the elderly may have an ischemic origin.
The general finding of MRI studies is that WMHs are rare in healthy individuals less than 50 years of age, but their presence increases with advancing age. WMHs are present in a majority of elderly subjects but are generally mild in those individuals who have no cerebrovascular disease; the PWMHs appear as a capping or pencil-thin lining of the lateral ventricles, whereas DWMHs are seen only as diffuse, focal, punctate foci. One meta-analysis of 156 studies showed that the prevalence of WMHs increased from 25% at age 30 to 75% at age 80. This study also found that age and hypertension were the major predictors of the presence of WMHs.
The relation between severity of WMHs and deficient cognition is less clear. As indicated previously, aging leads to measurable deficits in certain cognitive functions. Individuals at all ages with essential hypertension have also been shown to have impaired cognition in several domains compared to normotensive control subjects, even if the hypertension has been well controlled by drugs. Elderly hypertensive subjects have a greater amount of brain atrophy than do normotensive age-matched controls. Also, as stated previously, the extent of WMHs is greater in hypertensives than in controls. However, even in healthy subjects free of cerebrovascular risk factors, it has been reported that increased volume of WMHs is associated with increased ventricular volume and deficient cognitive function, especially on neurop-sychological tests sensitive to frontal lobe dysfunction. Interestingly, there seems to be a correlation between elevated systolic blood pressure, even in the normal age-related range, and WMH burden.
B. Microscopic Changes in Neuronal Structure and Neurotransmitter System Integrity
There are a number of cellular changes that occur in the gray matter of the brain with aging. Although these changes are more dramatic in disease states, there are also degenerative processes associated with healthy aging. One of the most fundamental and currently unresolved issues is whether or not there is a loss of neurons in the cerebral cortex of the aged brain. Early studies suggested that cell loss did occur with age, but recent studies have shown a decrease in the number of large cells and an increase in the number of small cells, suggesting that shrinkage of the cell body, as opposed to a decrease in number, is associated with senescence.
Some areas of the brain do show a decrease in neuronal number, including the hippocampus, thalamus, putamen, cerebellum, and subcortical nuclei such as the substantia nigra, locus coeruleus, nucleus basalis of Meynert (NBM), and inferior olive. The hippocam-pal region is of interest because of its purported role in encoding new memories; the earliest neuropathologi-cal changes in Alzheimer's disease occur in parts of this structure and in surrounding tissue. Parts of the basal ganglia (putamen, globus pallidus, and portions of the thalamus), along with the cerebellum and substantia nigra, are key components of the neural system involved with regulating movement. Neuronal death ofthe dopaminergic cells in the substantia nigra causes Parkinson's disease; several symptoms of this degenerative disease increase their frequency in the aged. Some studies have reported neuronal loss in the NBM, the source of the cholinergic projection to the cerebral cortex. This nucleus shows a significant loss of neurons in Alzheimer's disease.
There are also degenerative changes that occur within the nerve cell body, including accumulation of lipid products, vacuoles, inclusions, and abnormal protein within the cytoplasm. Pigment accumulation or lipofuscin occurs at different rates within different areas of the brain. The large neurons of the precentral gyrus are particularly predisposed to lipofuscin deposits, which are composed of lipids, proteins, and carbohydrates. Neuromelanin, which results from peroxidation of the lipofuscin granules, also occurs in the neurons of the substantia nigra and locus coeruleus. Granulovascular degeneration, which results in the accumulation of cytoplasmic vacuoles, is common in aging. Lewy bodies, commonly found in Parkinson's disease, occur in a small number of healthy aged subjects, whereas colloid inclusions or fine granular material in the cisterns of rough endo-plasmic reticulum are often found in the senescent brain. Marinesco bodies, composed of fine granules and filaments in a lattice-type network, also commonly occur with advancing age. Neurofibrillary tangles (abnormal fibrous protein accumulation in the cytoplasm), which are one of the key pathological markers of Alzheimer's disease, occur within parts of the hippocampus and amygdala, several subcortical nuclei, and some regions of the cerebral cortex (primarily in the entorhinal cortex in the temporal lobe) but with a density far less than that found in patients diagnosed with Alzheimer's disease.
A few studies have found senescent changes in the dendritic system of neurons, although this kind of analysis is sensitive to the type of fixation used to preserve the brain. These changes include a decrease in dendritic number, which usually begins with dendrites farthest away from the cell body. A decrease in the number of synaptic terminals is also seen. A proliferation of existing dendrites can be observed and is thought to be a compensatory mechanism for those that are lost. These changes can be seen after age 60. Neuroaxonal dystrophy, or enlargement of the distal ends of axons, occurs predominantly in some nuclei in the medulla and increases in frequency with age.
Neuropathologic changes that occur in the neuropil, or the area surrounding cells, include the development ofneuropil threads, senile plaques, and Hirano bodies. Neuropil threads consist of the abnormal protein seen in neurofibrillary tangles, but these proteins are located in the neuronal process surrounding amyloid plaque cores and are rare in aging. Senile plaques are composed of amyloid protein, degenerating neuronal processes, and reactive glial cells. There are different types of plaques and these are classified by the organization of amyloid and the presence or absence of degenerating cell processes. In normal aging, senile plaques may exhibit dystrophic neurites, but these neurites usually lack the paired helical filaments seen in Alzheimer's disease. These plaques are commonly found in the frontal and temporal cortex and in the hippocampus. Hirano bodies are spindle or rod shaped structures found in the neuropil surrounding neurons and are commonly seen in senescent brains.
There are numerous substances in the brain which play a critical role in neural transmission. The class of neurotransmitters that we discuss are called neuromo-dulators; their effect on neurotransmission has a longer time course than classical transmitters such as glutamate and GABA, and they seem to affect the responsiveness of target neurons to other inputs. These substances, which originate from midbrain and brain stem nuclei, include acetylcholine, dopamine, serotonin, and noradrenaline. They play a central role in many of the neurological and psychiatric illnesses that are common in the elderly, including Alzheimer's disease (acetylcholine), Parkinson's disease (dopa-mine), and depression (serotonin).
a. Cholinergic System The NBM in the basal forebrain provides most of the cholinergic innervation of the cerebral cortex. One study found that cells within the NBM increase in size until age 60 and then begin to atrophy, particularly in posterior regions of the nucleus. Acetylcholine (ACh) is the primary neurotransmitter produced by cells within the NBM. There is little change in acetylcholinesterase content, an enzyme responsible for the breakdown of ACh, in elderly subjects. There are also minimal changes in high-affinity choline uptake, which is the rate-limiting step in ACh production. However, age-related changes in ACh receptors are observed. There are two principal cholinergic receptor types: muscarinic and nicotinic. A 10-30% reduction in muscarinic receptor density is seen in the cerebral cortex and striatum. The cortex and hippocampus exhibit a decrease in nicotinic receptors, whereas the thalamus shows a decrease in nicotinic and an increase in muscarinic receptor density.
b. Dopaminergic System Dopamine innervation of the cortex, limbic system, and basal ganglia originates from the ventral tegmentum and the substantia nigra. There are substantial changes to neurons of the substantia nigra with age. After age 65, there is a progressive decline in cell number within this region. The remaining cells exhibit decreased nucleolar volume and mild accumulation of neurofibrillary tangles, Lewy bodies, and neuromelanin. Studies that have examined the effects of aging on dopamine have shown that the density of presynaptic D1 receptors decreases, whereas the density of postsynaptic D1 receptors increases in the striatum with age. The striatum also shows a decrease in pre- and postsynaptic D2 receptor density.
c. Noradrenergic System Noradrenaline is produced by locus coeruleus neurons found in the brain stem. There is a progressive loss of noradrenergic neurons from the brain stem beginning from age 3040. Noradrenergic neurites can be found in senile plaque formations in aging. A decrease in tyrosine hydroxylase, which is needed for the production of dopamine and noradrenaline, is also observed in elderly subjects. The cerebral cortex contains both a-and b-adrenergic receptors. Adrenergic a2 receptors significantly decrease with age. Loss of b-adrenergic receptors occurs in a more heterogeneous fashion based on cortical region. Receptors in the frontal lobe show no decrease in number, whereas those in the precentral, temporal, occipitotemporal, and cingulate cortical regions exhibit a linear decline with age.
d. Serotonergic System The raphe nuclei in the midbrain supply the serotonergic innervation of the brain. The primary metabolite resulting from the breakdown of serotonin, 5-hydroxyindoleacetic acid, does not decline with age. However, there is a decrease in imiprimine binding with aging; the imiprimine binding site is a presynaptic marker for the reuptake of serotonin. Two types of serotonergic receptors (Si and S2) have been found to decline with age. S1 receptors demonstrate up to 70% reduction in number; S2 receptors density is decreased by 20-50% in elderly subjects. Functional imaging studies using markers for the S2 receptor have found decreases in receptor density in occipital, parietal, temporal, and frontal lobes of the brain. Of these regions, the most significant decreases in density are observed in frontal and temporal cortices.
C. Summary: Anatomical Changes with Age
There is a general decrease in brain weight with advancing age, and macroscopic atrophy in the form of increased sulcal and ventricular size becomes clearly evident. The volumes of a number of brain structures are reduced in old versus young brains. The most prominent of these include the hippocampal formation, the frontal lobes, and a number of subcortical structures in the basal ganglia and the nuclei that are the source of the neuromodulatory transmitters (e.g., NBM and the substantia nigra). White matter changes, which are associated with vascular problems such as hypertension, are also more common in the elderly.
At the cellular level, the aging brain exhibits many neuropathologic changes, but these changes are diffuse and relatively modest in relation to those associated with neurodegenerative diseases. Decreased overall size and volume of gray and white matter occur relatively early in the life span. Degenerative processes that affect the gray matter of the brain appear to begin later and include diffuse accumulation of abnormal products within and around neurons and changes in the ultrastructure of neuronal processes. Neuromodu-latory neurotransmitter systems show declines in concentration and/or receptor densities in aging brains, but these declines are generally mild to moderate in degree.
Was this article helpful?