© 1992 SCIENTIFIC AMERICAN, INC
Aging Brain, Aging Mind
Late in life the human brain suffers attrition of certain neurons
and undergoes chemical alterations. Yet for many people,
these changes do not add up to a noticeable decline in intelligence
by Dennis]. Selkoe
C
ontemplate senescence, as Shakespeare
did. In As You Like It, his memorable
character Lord Jaques enumerates sev­
en ages of man, concluding with this sad
description:
Last scene of all, that ends this strange
eventful history,
Is second childishness and mere oblivion.
For many of us, as for the melancholy
Jaques, the prospect of aging continues to
conjure images of inexorable, devastating decline, a slow
march toward mindlessness and mortality. But is severe de­
terioration of the brain-and thus the mind-inevitable?
The answer is no. Admittedly, research indicates that as
youth fades, certain molecules and cells in the brain become
increasingly impaired or disappear. Some of the changes can
undoubtedly disrupt cognition if they accumulate past criti­
cal thresholds. Yet studies of human behavior suggest that a
mind-eroding buildup of damage is by no means an auto­
matic accompaniment of longevity.
Older adults who truly lose their minds probably do so be­
cause a specific disease markedly accelerates or is superim­
posed on the aging process. In developed nations, the lead­
ing cause of senile dementia-loss of memory and reason in
the elderly-is Alzheimer's disease. Other causes include the
occurrence of multiple strokes or Parkinson's disease.
Physicians cannot always distinguish between older people
who suffer from minor, relatively stable forgetfulness and
those who are in the early stages of Alzheimer's disease or an­
other progressive, dementing disorder. Ongoing research into
normal aging and into disorders of the mind will both en­
able doctors to make those distinctions and give rise to pal­
liative and preventive therapies. For most students of the ag­
ing brain, the ultimate goal is to enhance the quality of its
function in old age, not necessarily to prolong life, although
the latter could certainly result from the former.
Scientists who study the structural and chemical changes
that typify the aging brain in the absence of disease find that
the changes are heterogeneous, like the brain itself. The brain
consists not only of diverse neurons (the signal-conveying
cells) but also of varied glial cells (which help to support and
repair neurons) and of blood vessels. Certain subsets of cells
and areas of the brain are more prone to age-related damage
than others. And the time of onset and the mix and extent of
GEORGE BERNARD SHAW, who died in 1950 when he was 94,
wrote several plays in his nineties. A major goal of research
into the aging brain is to increase the number of people who
retain mental vigor throughout life.
© 1992 SCIENTIFIC AMERICAN, INC
physical alterations, as well as the effects on
intellect, can differ dramatically from person
to person. In general, however, it seems safe
to say that most of the structural and chem­
ical modifications I will discuss become ap­
parent in late middle life, that is, in the fif­
ties and sixties. Some of them become pro­
nounced after age 70. Because there probably
is no unifying mechanism that underlies all
senescence (age-related dysfunction of cells
and molecules) in the brain, it seems unlike­
ly that investigqtors will find a singular elixir
that will retard or reverse every decline.
Age-associated changes have been most studied in neurons,
which in general do not multiply after birth. As individuals
grow older, their overall number of brain neurons decreases,
but the pattern is by no means uniform. For example, very few
neurons disappear from areas of the hypothalamus that regu­
late the secretion of certain hormones by the pituitary gland.
In contrast, many more nerve cells tend to vanish from the
substantia nigra and locus coeruleus, which are specialized
populations of cells in the brain stem. Parkinson's disease can
decimate some 70 percent or more of the neurons in those ar­
eas, seriously disrupting motor function. Aging alone usually
eliminates many fewer cells, although older individuals who
exhibit mild symptoms reminiscent of Parkinson's disease­
decreased flexibility, slowness of movement and a stooped,
shuffling gait-may have lost up to 30 or 40 percent of the
original complement.
P
arts of the limbic system, including the hippocampus,
undergo variable amounts of cell death as well. (The
limbic system is central to learning, memory and emo­
tion.) Researchers have estimated that about 5 percent of
neurons in the hippocampus disappear with each decade in
the second half of life. This calculation suggests that about
20 percent of neurons are lost in that period. The attrition is
patchy, though; some hippocampal areas show no significant
decline.
Even when neurons themselves survive, their cell bodies
and their complex extensions known as axons and dendrites
(or, collectively, as neurites) may atrophy. Neurons bear a
single axon that relays signals to other neurons, often a dis-
DENNIS J. SELKOE, who in 1988 received the Leadership and
Excellence in Alzheimer's Disease Award from the Nationallnsti­
tute on Aging, is co-director of the Center for Neurologic Diseas­
es at Brigham and Women's Hospital in Boston. He is also profes­
sor of neurology and neuroscience at Harvard Medical School.
Selkoe received his medical degree from the University of Vir­
ginia. This is his second article for Scientific American.
SCIENTIFIC AMERICAN September 1992 135
 CORTEX
Large neurons shrink
Amyloid deposits develop
in extracellular spaces

THALAMUS
Selected neurons shrink or die

BASAL FOREBRAIN
Acetylcholine-secreting neurons
shrink or die

__-- -- --AMyGDALA
Amyloid deposits develop
in extracellular spaces
Neurofibrillary tangles
develop within neurons

NIGRA

Large neurons shrink or die

1,500 Amyloid deposits develop in extracellular spaces

Neurofibrillary tangles develop within neurons

Cil
�
«
II
� 1,000 LOCUS COERULEUS
f-
I Neurons die
Cl
UJ
�
Z
�
500 BRAIN STRUCTURES indicated in boldface are involved in learning, memory and
II reasoning. They normally undergo a number of anatomic changes late in life that
In
can potentially impair cognition. The alterations listed are just a sampling, and
they generally occur only in selected parts of the affected structures. The weight of
the brain declines with age as well (graph). Robert D. Terry and his colleagues at
O Lh��
20 30 40 AGE 60 70 80 90
50 (YEARS)
__ �-L�L--L��
the University of California at San Diego, who collected the data plotted on the graph,
conclude that the decrease sterns in part from shrinkage of large neurons.

tance away. Dendrites, which are more
abundant and are found in large branch­
ing arbors, generally receive signals
from other neurons.
Cell-body and neuritic atrophy com­
monly occur with aging in a number of
brain areas important to learning, mem­
ory, planning and other complex in­
tellectual functions. Large neurons, in
particular, shrink in parts of the hip­
pocampus and the cerebral cortex. And
cell bodies and axons can degenerate in
certain acetylcholine-secreting neurons
that project from the basal forebrain to
the hippocampus and diverse areas of
the cortex. Acetylcholine is one of vari­
ous neurotransmitters by which neu­
rons convey signals to one another.
dendrites in some regions of the hippo­
campus and cortex between middle age
(the forties and fifties) and early old
age (the early seventies), followed by a
regression of dendrites in late old age
(the eighties and nineties). These inves­
tigators postulate that the initial den­
dritic growth reflects an effort by viable
neurons to cope with the age-associ­
ated loss of their neighbors. Apparent­
ly, this ability to compensate fails in
very old neurons. Studies of adult rats
show a similar capacity for growth;
longer and more complex dendrites ap­
pear in the visual cortex after the ani­
mals are exposed to visually stimulat­
ing environments.
Such findings are encouraging. They
suggest both that the brain is capable
bodies and neurites, neurons can un­
dergo alteration of their internal archi­
tecture. For example, the cytoplasm of
certain cells of the hippocampus and
other brain areas vital to memory and
learning can begin to fill with bundles
of helically wound protein filaments
known as neurofibrillary tangles. An
abundance of such tangles in these and
other brain areas is believed to con­
tribute to the dementia of Alzheimer's
disease, but the significance of small
amounts in the undiseased brain is less
clear. The development of tangles dur­
ing aging seems to indicate that certain
proteins, particularly those of the cy­
toskeleton, or the internal scaffolding
of the cell, are being chemically modi­
fied in ways that could contribute to

Y
et not all neuronal changes are
necessarily destructive. Some
may represent attempts by sur­
viving neurons to compensate for loss
or shrinkage of other neurons and their
projections. Indeed, Paul D. Coleman,
Dorothy G. Flood and Stephen]. Buell
of the University of Rochester Medical
Center have observed a net growth of
of dynamic remodeling of its neuro­
nal connections, even in the later years,
and that therapy of some kind might
augment this plasticity. On the other
hand, the functionality of the dendrites
that appear in old age has yet to be de­
termined.
In addition to changes in their num­
ber and in the structure of their cell
inefficient signaling by these neurons.
In another internal alteration, neu­
ronal cytoplasm in many parts of the
brain becomes increasingly dotted with
innumerable granules containing lipo­
fUSCin, a fluorescent pigment. This pig­
ment is thought to derive from lipid­
rich internal membranes that have been
incompletely digested. Again, investiga-

136 SCIENTIFIC AMERICAN September 1992

© 1992 SCIENTIFIC AMERICAN, INC
tors disagree over whether lipofuscin
granules harm cells or are mere mark­
ers of longevity.
As is true of neurons, glial cells, which
have a supporting role in brain function,
also become altered. Robert D. Terry of
the University of California at San Diego
and other investigators have established
that the type known as fibrous astro­
cytes increases steadily in size and
number after age 60. Proliferation of
these cells, which are capable of releas­
ing diverse factors that promote neuro­
nal and neuritic growth, is of unknown
consequence. Perhaps it represents an­
other attempt by the brain to compen­
sate for gradual decrements in the
number and structure of neurons.
Meanwhile the areas between neu­
rons are also undergoing change. In hu­
mans, monkeys, dogs and certain other
animals, the extracellular spaces of the
hippocampus, cerebral cortex and oth­
er brain regions commonly accumulate
moderate numbers of spherical depos­
its called senile plaques. These plaques,
which develop very slowly, are primarily
store the genes for virtually all the pro­
teins made in cells. Such work indicated
that the enzymatic machinery designed
to excise and repair faulty nuclear DNA
becomes less efficient late in life and
perhaps in the presence of certain brain
diseases. Scientists have also found evi­
dence that cellular controls governing
genetic activity may be relaxed during
aging. One mechanism may involve sub­
tle elimination of methyl groups (CH3)
from certain parts of DNA molecules
[see "A Different Kind of Inheritance,"
by Robin Holliday; SCIENTIFIC AMERI­
CAN, June 1989].
In recent years, investigators have
begun to suspect that DNA in a special
cellular location-the mitochondria­
may also contribute to senescence of
the brain. Mitochondria are intracellu­
lar "power plants" that provide cells
with critically needed energy. They con­
tain their own snippet of DNA, which
bears instructions for the manufacture
of 13 proteins needed for energy gen­
eration. If mitochondrial DNA slowly
became partially defective, the defects
changes associated with aging includes
the finding that such DNA seems to be
more susceptible to damage than is nu­
clear DNA. One reason may be that the
DNA-repair machinery in the organelles
is less effective than that in the nucleus.
Mitochondrial DNA is also probably sub­
ject to more attack by highly reactive
oxygenated compounds known as free
radicals. Such compounds are a contin­
ual by-product of the reactions through
which mitochondria produce energy.
Free radicals are made during other
cellular processes as well and in re­
sponse to ionizing radiation. They oxi­
dize, or add oxygen atoms to, mole­
cules, thereby altering them.
Furthermore, investigators have dis­
covered that a crucial enzyme encod­
ed by mitochondrial DNA-cytochrome
oxidase-declines with age in rat brains.
Several researchers have also identified
specific deletions in stretches of mito­
chondrial DNA in aged brains and in
patients with certain age-linked brain
disorders, such as Parkinson's disease.

E
aggregates of a small molecule known could result in production of damaged ven if most nuclear and mito­
as the beta-amyloid protein. Amyloid mitochondrial proteins or in the elimi­ chondrial genes remained unal­
protein also accumulates in scattered nation of such proteins. tered and gave rise to the prop­
blood vessels in these regions and in the The evidence favoring a causal role er amounts of normal proteins, sub­
meninges, the connective tissue that en­ for mitochondrial DNA in some of the sequent modifications of the proteins
velops the brain.
Researchers have not fully deter­
mined which cells give rise to these
protein deposits and what effect the
accumulations have on nearby neurons
in healthy elders. The answers should
emerge soon, however, because the dra­
matically increased deposition of amy­
loid protein in patients with Alzheimer's
disease has pushed such issues to the
forefront of research.
300
250
T
he diverse structural alterations UJ
that occur in the aging brain re-
6
a:
sult from deleterious changes in
the activity or abundance of molecules
u
�
200
that are important to the integrity and I
functioning of cells. One of the most � 150
venerable theories of aging holds that tE
cells throughout the body senesce be-
�
cause defects slowly accrue in their i= 100
DNA, the material from which genes �
are constructed. Genes carry the chem-
ical instructions that inform cells pre-
tE
0
50
cisely how to synthesize proteins. At
some point, the scenario goes, damage
to DNA lowers the quality or quantity
o MIDDLE AGE OLDER VERY OLD ADULTS WITH
of critical proteins (such as certain en­ (FIFTIES) (SEVENTIES) (NINETIES) ALZHEIMER'S DISEASE
zymes) in cells. Or the damage may in­
lllPPOCAMPAL NEURONS (photographs) have been stained in brain specimens from
crease the activity or amount of un­
healthy humans in their fifties, seventies and nineties and from a patient with
desirable proteins (such as those that Alzheimer's disease (left to right). Dorothy G. Flood and Paul D. Coleman of the Uni­
promote the development of cancers). versity of Rochester Medical Center find that the average length of dendritic trees
Until recently, genetic research fo­ (branched shapes) on such neurons increases between middle age and old age in
cused almost entirely on the chromo­ healthy adults, regressing only in late old age (graph). The normal growth may re­
somal DNA in the nucleus, the long flect an attempt by the brain to compensate for destructive age-linked changes. Den­
strands of helical DNA that collectively drites in Alzheimer's patients do not exhibit age-related growth.

SCIENTIFIC AMERICAN September 1992 137

© 1992 SCIENTIFIC AMERICAN, INC
 could lead to molecular mischief in late
Dementia in the U.S. life. Proteins can undergo a number of
distinct chemical modifications, includ­

I (a),
n 1 99 2 the Framingham Study, which repeatedly assesses the health of a ing oxidation of certain of their com­
large group of subjects as they age, estimated the prevalence of dementia ponent amin
o acids, glycosylation (ad­
i n cluding Alzheimer's disease ( b). The Alzheimer's figures differ from dition of carbohydrate side chains) or
those of a survey conducted in East Boston (by a team led by Denis A. Evans of cross- linkin
g (formation of strong chem­
Harvard Medical School), probably because the Framingham group applied a ical bridges between proteins). Such
narrower definition of the disease. The Framingham Study, consi stent with modifications occur normally and en­
other s, also found Alzheimer's disease to be the leading cause of persistent able proteins to carry out their func­
dementia in late life (c).
Some currently treatable causes are listed (d). tions. On the other hand, there is abun­

II 30
« 23.8 dant evidence that as people age, many
proteins ac cumul
ate unhelpful modifi­

!z 20
i=
ifi
w ::;;
UW
cations. For example, the levels of oxi­
dized proteins in skin cells of humans
and brain cells of rats increase progres­
0: 0
WI 10 0.4 0.9
0.. 1-
� o 1.8 75-793.6 sively with age. In very old rats, such
proteins can account for 30 to 50 per­

60-64 65-69 70-74 80-84 85-93

AG E (YEARS)
cent of the total protein content of the
cell. Cells from young adults with pro­
geria, a rare and remarkable syndrome

b 50 47.2 marked by premature aging of many

body tissues, contain levels of oxidized

00
I«
t: �
40
w proteins that approach those found in
healthy 80-year-olds.
Because enzymes are the proteins
� 00
1-
30
Ci that catalyze many of the most impor­
tant chemical reactions in cells, they
ifi ffi
� � 20 18.7 have received the greatest scrutiny.

1I!4�.�1 !D
Several enzymes that synthesize neuro­
WW
10 \�� �
0.. I
N
<i!
o
FR
.
GHA
EAST
BOSTON transmitters or their receptors become
less active as an individual ages. For
some of these enzymes, postsynthetic

�=.65-74y,.-
.- :: 75-84y modification may be partly at fault.
It is a cruel irony that proteases, the
very enzymes responsible for degrad­
AGE (YEARS)
ing oxidized and other proteins, them­
selves undergo oxidation and loss of ac­
c DISTRIBUTION, BY PROBABLE CAUSE* tivity. The problem of damaged proteas­
es, potentially bad enough on its own,
may be exaggerated by a parallel decline
in the enzymes superoxide dismutase
and catalase. These proteins normally
inactivate free radicals and defend
against oxidative damage to diverse
kinds of molecules. At least in rats, they
become increasingly scarce in old age.
John M. Carney of the University of
Kentucky and Robert A. Floyd of the
Oklahoma Medical Research Founda­
tion and their co-workers have recently
provided some of the first evidence that
such oxidation may lead to a loss of
mental function. Comparing aged and
young gerbils, they showed that signifi­
cantly more proteins in the older ani­
mals were oxidized. This increase was
accompanied by a decrease in activity
*Numbers do not add up to 100 percent because of rounding off. of certain critical enzymes. Moreover,
the older anim
als had more trouble than
did the younger ones in navigating a
d SOME TREATABLE CAUSES OF DEMENTIA radial maze.
Medications Certain t u m o r s or infections o f the brain When the scientists injected the aged
Emotional depression Blood clots pressing on the brain
gerbils with a compound (N- tert-butyl­
()(-phenylnitrone) known to inactivate
Vitamin B 1 2 deficiency Metabolic i m balances (including thyroid,
oxygen free radicals and thus decrease
Chronic alcoho l i s m kidney o r liver d i sorders)
oxidation, the levels of oxidized pro­
teins declined, and the activity of the en-

138 SCIE NTIF

IC AMERI
CAN September 1 992
© 1992 SCIENTIFIC AMERICAN, INC
zymes increased to the levels character­
istic of young animals. What is more,
the biochemical improvements were ac­
companied by the restoration of maze­
running skill to youthful levels. When
the therapy was stopped, the amount of
oxidized protein and enzyrllatic activity
reverted to that typical of old animals.
Many important nonprotein mole­
cules in the brain also change signifi­
cantly in structure or amount during
aging. There is evidence that the long
chains of carbon atoms making up the
lipids in membranes that envelop cells
and internal organelles undergo chem­
ical modifications. Among these is de­
structive oxidation by free radicals. As
a result, the precise composition of the
membranes can shift, subtly altering
their behavior.
For example, investigators have
documented an age-related de­
crease in the fluidity of the mem­
branes making up synaptosomes,
tiny neuronal vesicles involved in
the storage and release of neu­
rotransmitters. And age-related
changes occur in the lipid compo­
sition of the myelin that sheathes
and insulates axons. Alteration
of myelin can have a measurable
effect on the speed and efficien­
cy with which nerve fibers propa­
gate electrical impulses over long
distances. TISSUE FROM TIIE BRAIN of a 69-year-old man is rid­
dled with the classic lesions of Alzheimer's disease­
senile plaques and neurofibrillary tangles. The plaque
ations and disruptions of the intellect
will remain murky.
We do know that in people who are
free of Alzheimer's disease and other
specific brain-threatening disorders, the
extent of anatomic and physiological
alterations tends to be modest. In many
studies reporting an age-related neuro­
chemical deficit-such as reduction in
the activity of a particular enzyme or in
the levels of selected proteins or RNA
molecules-the measures reported for
elderly adults have ranged from 5 to 30
percent below those in young adults.
The degree of neuronal loss in various
regions of the brain falls into roughly
the same range.
Although a 30 percent loss might
seem quite high, such gradual declines
toms of dementia, a figure that rose to
about 20 percent in the subjects aged
75 to 84. Then the number jumped to
about 50 percent in people older than
85 years (which is twice as high as cer­
tain other estimates). As disturbing as
the data are for people older than 7 5
years, the figures still indicate that a
good many people escape major distur­
bances in cognition in the later years.
Analyses of performance in healthy
old adults lead to a similar conclusion.
For instance, Arthur L. Benton, Daniel
Tranel and Antonio R. Damasio of the
University of Iowa College of Medicine
found that when people in their seven­
ties and eighties remain in good health,
they show only a subtle decline in per­
formance on tests of memory, percep­
tion and language.
One decrement on which many
studies agree is a reduction in the
speed of some aspects of cogni­
tive processing. Hence, septuage­
narians may be unable to retrieve
certain details of a particular
past event quickly-say, the pre­
cise date or place-but they are
often able to recall the informa­
tion minutes or hours later. Giv­
en enough time and an environ­
ment that keeps anxiety at bay,
most healthy elderly people score
about as well as young or mid­
dle-aged adults on tests of men­
tal performance. The more com­

T
he molecular changes I have plex a task is (for example, a mul­
visible in this fragment (large dark-gold sphere) con­
discussed are but a small tistep mathematical problem), the
sists of beta-amyloid protein and, at the periphery,
sampling of those that have more likely it is that an otherwise
damaged axons and dendrites (dark squiggles). The
been found in aged brains of hu­ healthy elder will perform less well
tangles, which are twisted fibers that fill the cyto­
mans and other mammals. In plasm, make a number of cells appear blackened than does a young adult. A mes­
trying to make sense of such al­ (small dark blobs). Plaques and tangles appear in sage of guarded optimism emerg­
terations, scientists immediate­ the brains of healthy aged adults, but usually to a es from many investigations of
ly confront the problem of deter­ much lesser extent and in restricted regions. normal aging: one may not learn
mining whether a phenomenon or remember quite as rapidly dur­
they have documented is the pro­ ing healthy late life, but one may
verbial "cart" or "horse." For instance, often appear to have little practical ef­ learn and remember nearly as well.
there is little doubt that DNA accrues fect on thinking. Indeed, positron emis­ Taken together, then, the physical,
damage over the years. But does the sion tomographic (PET) imaging indi­ epidemiological and psychological find­
damage result, for example, in increased cates that the brains of healthy people ings suggest that mild to moderate de­
oxidation of enzymes, or does oxidation in their eighties are almost as active as creases in memory or speed of intel­
occur first and lead to accumulation of those of people in their twenties. As is lectual processing may be related to a
DNA deficits? The likelihood is that both true of other organs, the brain appears gradual accumulation of standard ana­
sequences can happen. Once many pro­ to have considerable physiological re­ tomic and physiological brain changes
cesses get started, they undoubtedly ex­ serve and to tolerate small losses of that accompany aging. In contrast, de­
acerbate others, triggering a complex neuronal function. mentia apparently arises from more
cascade of events. Epidemiological and psychological speCific and excessive changes in sub­
Equally important is the question of studies paint a similar picture. Estimates sets of neurons and in neural circuits.
what bearing all these diverse anatomic of the prevalence of dementia vary, but In other words, diseases having distinct
and physiological age-linked changes the most dire of them-derived from a causes and mechanisms underlie senile
have on the mind. In many people, the door-to-door study conducted by Den­ dementia. Of course, one might well
answer may be "very little." Until scien­ is A. Evans of Harvard Medical School wonder why people become more prone
tists can record the mental functioning and his colleagues-indicates that as a to various debilitating brain disorders,
of large numbers of healthy people group, almost 90 percent of all people including Alzheimer's disease, as they
close to the time they die and then cor­ older than 65 years are free of demen­ age. In many instances, the answers are
relate such data with structural and tia. Evans and his co-workers reported uncertain.
chemical changes in their brainS, any in 199 1 that fewer than 5 percent of Because Alzheimer's disease is by far
links between speCific physical alter- subjects aged 65 to 75 exhibited symp- the most common cause of severe in-

SCIENTIFIC AMERICAN September 1992 1 39

© 1992 SCIENTIFIC AMERICAN, INC
tellectual decline in the elderly, let me
briefly review the latest research into
why it develops, why it becomes in­
creasingly prevalent late in life and
how it might eventually be treated. For­
tunately, progress in this area has been
truly remarkable of late [see "Amyloid
Protein and Alzheimer's Disease," by
Dennis]. Selkoe; SCIENTIFIC AMERICAN,
November 1991].
beta-APP mutations in inherited Alz­
heimer's disease, now makes it clear
that amyloid protein deposition can
serve as a seminal event in some if not
all cases of Alzheimer's disease.
No one is sure how the initially in­
ert protein leads over a long time to
the extensive structural and biochem­
ical changes in axons, dendrites, neu­
ronal cell bodies and glial cells that rav­
ish the minds of Alzheimer's victims.
person who can walk, talk and eat but
cannot make sense of the world.
In spite of recent progress, many
central problems need to be addressed.
How do mutations in the beta-APP gene
lead to accelerated amyloid deposition
compared with the slow rate encoun­
tered in normal aged humans? Why is
such deposition largely confined to the
brain, when virtually all tissues synthe­
size the amyloid precursor? Which cells

U
ntil very recently, the question of
what causes Alzheimer's disease
had to be answered, "We don't
know." But rapid advances from many
laboratories analyzing the chemistry
and molecular biology of the beta-amy­
loid deposits has now led to identifi­
cation of the first specific molecular
cause of this complex and devastating
disorder. In 199 1 studies by Alison M.
Goate and John A. Hardy and their col­
leagues at St. Mary's Hospital Medical
School in London, and subse­
quently other research teams, es­
tablished that particular genetic
mutations are at fault in at least
some instances.
These DNA mutations occur
within the gene that gives rise to
the beta-amyloid precursor pro­
tein, or beta-APP. This precursor
includes within it the beta-amy­
loid protein that constitutes both
senile plaques and vascular amy­
loid deposits. The normal func­
tions of beta-APP have not yet
been revealed, but those of us
who study Alzheimer's disease
have found that the precursor is
made by most cells of the body.
We know, too, that the mutated
version somehow leads to accel­
erated extracellular and vascular depOSi­
tion of the beta-amylOid segment. Some
mutations may lead to more or faster
One possibility is that the protein it­
self remains unreactive, but as it col­
lects over many years, it attracts oth­
er types of molecules to the deposits.
These other molecules may then dam­
age surrounding neurons and glia. An­
other hypothesis suggests that after the
amyloid protein reaches critical concen­
trations, it directly damages surround­
ing neurons and glia or makes them
more vulnerable to subtle injurious pro­
cesses that can occur in the brain.
PET SCANS OF BRAINS from a healthy aged adult
(left) and a patient with Alzheimer's disease (right)
differ markedly. The excess of dark shading in the
second image shows that brain activity is impaired.
In any case, the work of several neu­
roanatomists-including Damasio, Brad­
ely T. Hyman and Gary W. Van Hoesen
actually secrete the devastating amy­
loid fragments? Why do some neurons
in certain brain regions, such as the hip­
pocampus, show striking reaction to
the presence of amyloid protein, yet
other areas, such as the cerebellum, ex­
hibit little or none? Most important,
how can the terrible destruction be
blocked?
As these questions are being pursued,
so is the problem of how to bolster and
protect the aging mind. The fact that no
single compound is likely to block
all the potential ravages of great
longevity is underscored by re­
sults of many clinical trials of
vitaminS, minerals and various
other compounds thought to "en­
hance" biochemical reactions in
the brain or increase blood flow.
These substances have yielded lit­
tle or no cognitive improvement
in either demented or functional
elderly people.
One reasonable "home remedy"
would be to stay physically fit.
Robert E. Dustman and his col­
leagues at the University of Utah
and other investigators have dem­
onstrated that older subjects who
regularly do aerobic exercise per-
form better on cognitive tests
than do sedentary individuals of the
same age with low aerobic fitness.

amyloid accumulation than others. Has­
tened depOSitions, in turn, could partly
explain why some people show symp­
toms earlier than others.
Research on Down's syndrome has
contributed importantly to the new un­
derstanding. Individuals with that syn­
drome are born with three copies of
chromosome 21 (where the beta-APP
gene is located) rather than the normal
two copies. They also invariably develop
myriad senile plaques and neurofibril­
lary tangles in the fourth and fifth de­
cades of life. Neuropathological exami­
nation of Down's patients who have
died early in life has revealed that a few
amorphous deposits of amyloid protein
can begin to appear in the teens, de­
cades before full-blown senile plaques
and neurofibrillary tangles and clinical
signs of dementia develop. That critical
finding, together with the discovery of
and their co-workers at the University
of Iowa and John Morrison of the Mount
Sinai School of Medicine and his col­
leagues-has shown that a buildup of
amyloid protein, combined with the for­
mation of neurofibrillary tangles and
other structural changes in neurons and
their extensions, contributes to a pro­
gressive disconnection of neuronal cir­
cuits serving memory and thinking.
Over years the limbic system and the
association cortices, which are vital to
organizing mental processes, appear to
become increasingly out of touch with
other neuronal areas. This disconnec­
tion helps lead to the impaired memo­
ry, judgment, abstraction and language
that is all too familiar in Alzheimer's
patients. Because most motor and sen­
sory functions are spared until rather
late in the disease, the changes give
rise to the classical, tragic picture of a
I
would additionally advise against
ingesting agents, including alcohol,
that interfere with the activity of
the nervous system. Similarly, I would
urge physicians to be cautious when
prescribing for elderly patients medica­
tions that act on the brain. Extensive
experimental and clinical evidence has
shown that people older than about 60
years are particularly sensitive to benzo­
diazepines (such as the sedative Val­
ium) and many other depressants and
stimulants of the central nervous sys­
tem. Compared with young adults, old­
er people suffer a greater decline in rea­
sOning while such drugs are in their sys­
tem, are affected longer and react to low
doses more strongly. These undesirable
effects on cognition are even more pro­
nounced in those who already suffer
from a dementing illness.
Researchers continue to engage in

140 SCIENTIFIC AMERICAN September 1992

© 1992 SCIENTIFIC AMERICAN, INC
READER Si ERVICE DIR ECTORi Y
The advertisers listed below are making additional information available, free. Just circle the
corresponding number on the adjacent card, detach and mail . Offer expires December 15, 1992.

1. ACADEMIC PRESS DICTIONARY OF SCIENCE AND TECHNOLOGY Thelargestoriginaldictionaryofanyt)pe

produced in the US sinre the 1960's, this volume features 124,000 definitions, encompassing every field of scienre today. (800-321-5068)

2. AMERITECH Circle #2 for more information on Ameritech's commitment to improving the quality of life in the
Great Lakes Region it serves.

3. ARC SCIENCE SIMULATIONS An interactive, visual model of the dynamic sky and solar system through time.

4. AUDIO-FORUM LEARN A FOREIGN LANGUAGE ON YOUR OWN! Cassette/book courses in 76 languages

from Audio-Forum. Free catalog.

5. BIO-IMAGING TECHNOLOGIES, INC. The new measure for drug and device development. For business inquiries
call 609-883-2000.

6. BOOKS ON TAPE, INC.

7. BRAINMAKER Find patterns and hidden relationships in data and learn from the past. Complete neural network
development system. California Scientific Software. 800-284-8112, 916-478-9040.

8. CALCULUS BY AND FOR YOUNG PEOPLE by Don Cohen. Book, worksheets and two videotapes.
Telephone 1-217-356-4555 FAX 1-217-356-4593

9. COURVOISIER COGNAC Discover the pleasures of cognac and enhance your knowledge. Send for a FREE
booklet outlining the most "Frequently Asked Questions on Cognacs & Courvoisier."

10. DAVIS INSTRUMENTS The world's smallest and most affordable weather stations designed for use anywhere.
PC option too! Free catalog. (800-618-3669).

11. EDUCALC MAIL STORE The EduCALC catalog offers the latest in calculators, Palmtop computers and Informators
at the best discount prices available. Comprehensive selection of books, peripherals and software.

12. EMI BRONZE REPRODUCTIONS Remington, Russell, Chiparus, Moreau, Degas. Cast in the highest quality for
discerning bronze collectors. (800-777-8126)

13. EPIC SYSTEMS NEURALYST FOR EXCEL Neuralyst integrates neural network technology into Microsoft Excel
(PC or Macintosh), providing for "intelligent" spreadsheet data analysis capabilities.

14. GLOBAL INTUITION NETWORK Promotes research and applied use of intuition in decision making. Free information.

15. HOECHST-ROUSSEL PHARMACEUTICALS, INC. SomeIVilJ.e, New Jersey 1-800-451-4455.

16. HORSTMANN SOFTWARE DESIGN CORPORATION ChiWriter scientifidmultilingual word proressor for the IBM PC
and compatible. Enables the editing of scientific and foreign language text in completely WYSIWYG fashion. (800-736-8886)

17. HUGHES COMMUNICATIONS, INC. Brochure describes services offered in C-band, Ku-band and high-powered
DBS satellite communications for corporate networks, cable, broadcasting and other applications.

18. ISCAN, INC. World's leading manufacturer of eye movement monitoring systems for medical research, human factors
and advanced man/machine interface applications.

19. JOHN NUVEEN & CO., INC. To learn more about tax-free investments, send for a free informative brochure which
will be sent promptly with a prospectus.

20. MAZDA MOTOR OF AMERICA To find out more about the New Mazda and our whole new line of exciting cars
and trucks, call 1-800-639-1000.

21. MEDIA MAGIC Offers a free 96 page catalog of unique educational and entertaining Software, Videotapes, and
Books, about Computers in Science & Art.

22. NILES & ASSOCIATES Professors, students and researchers! Organize your references and create bibliographies
instantly with EndNote Plus. Call 1-800-554-3049.

23. PHARMACEUTICAL MANUFACTURERS ASSOCIATION To learn more about the critical role of pharmaceutical
research, call for a free "Good Medicine" brochure. 1-800-538-2692

24. PRESCIENCE For information on Theorist® and Expressionist®, an easy to use equation typesetter, call 415-543-2252.
25. RAND JOURNAL OF ECONOMICS A leader in the field of microeconomics, a focus on industrial organization,
contracts, and regulation.

26. SCHERING-PLOUGH RESEARCH INSTITUTE A research-based company engaged in the discovery, development,
manufacturing and marketing of pharmaceutical and health care products worldwide.
27. SA MEDICINE A cost-effective and efficient way for practitioners to keep current in the ever-changing field of medicine.

28. STANFORD GRAPHICS FOR WINDOWS Delivers stunning presentation graphics, plus powerful data
analysis, specifically designed for the business, statistical and technical user.

29. TOURISM CANADA Canada's Official Travellers Guide. Everything you need to know plus telephone numbers.
Order your free copy today. 1-800-VlSIT92.

30. WOLFRAM RESEARCH! MATHEMATICA Mathematica is a system for doing numerical, symbolic and graphical
computation used both as an interactive calculation tool and as a programming language. CalI1-800-441-MATH.

If the card is missing, write on a postcard the name and corresponding number of the advertiser(s) you would
like information from and send to: Scientific American, September 1992, P.o. Box 5147, Pittsfield, MA 01203-9827.

© 1992 SCIENTIFIC AMERICAN, INC

 a BEFORE b DURING C PLACEBO d LAST DAY OF
12
TREATMENT TREATMENT
15 CONDITION TREATMENT

::J
(I)...J-
Q..-z ' (I)
:J:2w 9 � a:
oa:f-
a:wo
� 1i?10
(!J Q..a:
...J(I)
Q..
6
u.a:
O
W

Do
>-Wu. (!J
z...JO a:
00:2
tIl:2.,.;
w�
tIlZ
:2Z 5
[]
a:O
3 :J:J

D DO
a:
<3�(!J za:
�
0 YOUNG OLD 3 7 14 o YOUNG OLD
DAYS

MEASURES OF PROTEIN CHEMISTRY in the brains of gerbils
suggest that certain age-related changes may be reversible.
John M. Carney of the University of Kentucky and Robert A.
Floyd of the Oklahoma Medical Research Foundation find that
the number of carbonyl groups on brain proteins (a marker of
protein oxidation) in old gerbils (blue bar in a) is typically
higher than in young gerbils (brown). But the levels decline
in response to a drug that inactivates certain oxidizing chem­
icals (b). Geriatric gerbils are also less efficient than young­
sters at negotiating a radial maze (c), such as the one shown
in the photograph. After treatment, the number of errors
drops (d). The apparent improvement in short-term memory
encourages hope that antioxidants may one day help to some­
what protect the minds of aging humans.

lively debate over whether maintaining
or increasing mental activity can protect
against cognitive decline late in life. Un­
fortunately, rigorous data on the sub­
ject remain elusive.
The value to brain function of dietary
restriction-a well-publicized potential
antidote to aging-is similarly unclear.
A nutritionally balanced but very low
calorie diet has been shown to delay
many age-related diseases and to in­
A more palatable alternative (both lit­
erally and figuratively) to serious di­
etary denial might one day be pro­
longed administration of antioxidants,
such as vitamin E. This vitamin has
been shown to extend longevity and re­
tard some age-associated systemic dis­
eases in rodents, but benefits for the
aging human brain have not been
proved.
old people. An estimated three million
Americans are 85 years or older to­
day, and that number may double by
the year 2020. Discovery of ways to
block age-related disorders of higher
cortical function without necessarily
prolonging life will undoubtedly enable
many of the elderly to remain indepen­
dent and enjoy life well beyond the
eighth decade. Successful aging of the
body and mind will bring about pro­

crease the life span in a variety of low­
er mammals. In some studies, rats fed
restricted diets exhibited fewer neuro­
chemical changes in their brains late in
life than did their better-fed counter­
parts, and they were more successful
in old age at learning to find their way
through a maze.
Similarly, Alan Peters and his col­
leagues at the Boston University School
of Medicine maintained rats on a very
low calorie regimen that enabled them
to live for as long as four years, per­
haps a year longer than usual. The
team found that the animals lost neu­
rons and developed various age-related
neuronal and glial alterations later in
life than did control animals. On the
other hand, the eventual occurrence of
such alterations suggests that calorie
control may delay, but will not prevent,
senescence of the brain.
No one really understands the mech­
anism by which marked calorie restric­
tion leads to prolonged survival in test
animals. Nor does anyone know the ex­
tent to which it might retard cognitive
impairment in humans. It is evident,
however, that to have any benefit, the
approach would probably have to be
practiced throughout much of life. More­
over, sudden and severe nutritional de­
privation in the elderly could lead to
symptoms of dementia, making calorie
restriction a riSky undertaking if it is at­
tempted without professional guidance.
F
or now, the most rational ap­
proach to developing treatments
for cognitive failure late in life is
to decipher the molecular mechanisms
underlying dementing diseases and
then to design drugs that block one or
more critical steps. In Alzheimer's dis­
ease, for example, it appears likely that
therapies will ultimately be directed at
inhibiting the enzymes that liberate the
beta-amyloid protein from its precur­
sor, blocking the delivery of the amyloid
protein into cerebral tissue or prevent­
ing the inflammatory and neurotoxic
responses that the protein apparent­
ly initiates. Such treatments might also
prove helpful for combating mild to
moderate forgetfulness in some old­
er people who do not have full-blown
dementia. This seems likely because
the amyloid plaques and neurofibrillary
tangles that characterize Alzheimer's
disease do form in areas important for
memory and learning during healthy
aging, albeit to a much smaller de­
gree. Several therapies are also under
study for preventing Parkinson's dis­
ease and for preventing or treating
stroke [see "Stroke Therapy," by Justin
A. Zivin and Dennis W. Choi; SCIENTIFIC
AMERICAN, July 1991].
During the next three decades, rigor­
ous molecular and clinical examination
of brain aging will become more com­
mon as the developed nations confront
a huge surge in the numbers of very
found economic and sociological con­
sequences that will require great cre­
ativity and vigor to address. Fortunate­
ly, at that point, society will possess
a valuable resource to help solve those
problems: the sharp minds and ac­
cessible wisdom of many of its oldest
citizens.
FURTHER READING
NEURON NUMBERS AND DENDRITIC Ex­
TENT IN NORMAL AGING AND ALZHEI­
MER'S DISEASE. P. D. Coleman and D. G.
Flood in Neurobiology of Aging, Vol. 8,
No. 6, pages 521-545; November/De­
cember 1987.
THE RETARDATION OF AGING AND DIS­
EASE BY DIETARY RESTRICTION. Richard
Weindruch and Roy L. Walford. Spring­
field, Ill., Charles C. Thomas, 1988.
RNA AND PROTEIN METABOUSM IN THE
AGING BRAIN. Caleb E. Finch and David
G. Morgan in Annual Review of Neuro­
science, Vol. 13. Edited by W. M. Cowan
et al. Annual Reviews Inc., 1990.
LONGEVITY, SENESCENCE, AND THE GE-
NOME. Caleb E. Finch. University of Chi­
cago Press, 1990.
THE MOLECULAR PATHOLOGY OF ALz­
HEIMER'S DISEASE. Dennis ]. Selkoe in
Neuron, Vol. 6, No. 4, pages 487-498;
April 1991.
BRAIN AGING AND ALZHEIMER'S DISEASE:
"WEAR AND TEAR" VS. "USE IT OR LOSE
IT." D. F. Swaab in Neurobiology of Ag­
ing, Vol.
12, No. 4, pages 317-324; July/
August 1991.

142 SCIENTIFIC AMERICAN September 1992

© 1992 SCIENTIFIC AMERICAN, INC
SECRETS from a Lifetime of Practice . • •

Very up-to-date mini-textbooks

in Q & A format
Each book in the Secrets Series is a practical,
clinical guide that add r e s s e s key que s t i o n s
commonly raised by teachers during rounds , i n
the O R , in the c l i n i c , a n d on o r a l exam s .
S ECR ETS are manageable, readable textbooks
for students and residents, and succinct reviews
and references for practitioners.

MEDICAL SECRETS
Edited by ANTHONY ZOLLO, J r . , M . D .
Q. How can I obtain these
The "Secrets" approach is applied in this book to review
most of what a student should k now about general SECRETS?
internal medicine. Organized by subspecialty areas of
internal medicine, the book i s an ideal review for
examinations, rounds, and clinical d i scussions .
A. Send your order today.
• June 1 99 1 , 550 pages, il lust rated

OB/GYN SECRETS CRITICAL CARE SECRETS

Edited by H E LEN L. F R E D E R I C K S O N , M . D . Edited by P O L LY E. PA R S O N S , M . D .
and LOU I S E W I L K I N S - H AU G , M . D . , Ph. D . and J E A N I N E P . W I E N E R- K RON I S H , M . D .

T h i s m i n i-textb o o k of obstetrics a n d gynecology i n
q u e s t i o n and answer fo rmat encompasses g e n e r a l
gynecology, reproductive endocrinology a n d infertility,
gynecologic oncology, general obstetrics, maternal
complications, the fetus, the placenta, and labor and
delivery. • M ay 1 99 1 , 308 pages, illustrated
The fu ndamentals required to care for critically ill
p a t i e n t s - w h e t h e r i n t h e i n t e n s ive care u n i t , the
operating room, the recovery room, or the emergency
room-are presented in questions and answers designed
both to provide factual information and to stimulate
discussion. • February 1 992, 480 pages, illustrated

SURGICAL SECRETS, 2nd edition PEDIATRIC SECRETS

Edited by C H A R L E S M. A B E R N AT H Y , M . D . Edited by R I C H A R D A. P O L I N , M . D .
and A L D E N H . H A R K EN , M . D . and M A R K F. D I T M A R , M . D .

The vanguard book of the " Secrets " series that achieved Here is a collect ion of answers, tips and "secrets" from
best-seller status is now in its second edition, and is even the experts on a wide range of d iagnostic and thera­
better. The text has been revised and updated and peutic problems that pediatricians face. This book
includes a new Atlas of Technical Tips that contains provides both the right question and its answer to help
graphic, original artwork depicting methods of tying students, residents, and practitioners hone their skills
sutures, handling instruments, and other basics. for practice and board exams.
• July 1 99 1 , 330 pages, illustrated • 1 989, 447 pages, illust rated

r-----------------------------------------------------------------------------------,

o CRITICAL CARE SECRETS, $32.95 U . S . ; $37.95 elsewhere. Name ___________________________________

480 pages, illustrated , February 1 992.

Street __________________________________

o MEDICAL SECRETS, $ 3 3 . 9 5 U . S . ; $ 3 8 . 9 5 elsewhere.

5 5 8 pages, illustrated , 1 99 1 . City/ State/ Z i p ____________________________

o OB/GYN SECRETS, $ 2 7 . 9 5 U . S . ; $ 3 2 . 9 5 elsewhere.

308 p ages, illustrated , 1 99 1 . I enclose payment: 0 Check 0 Visa 0 M asterCard

o PEDIATRIC SECRETS, $30.95 U . S . ; $ 3 5 . 9 5 elsewhere. Credit Card # ______________________________

450 pages, i llustrated, 1 98 9 .

E x p i ration Date __

o SURGICAL SECRETS,
__________________________ _

2nd Edition, $ 2 8 . 9 5 U . S . ; $ 3 3 . 9 5
elsewhere. 330 pages, i llust rated , 1 99 1 . Signature ________________________________

HANLEY & BELFUS, INC., Medical Publishers 2 1 0 South 1 3t h Street, P h iladelphia, PA 1 9 1 07 2 1 5-546-7293

© 1992 SCIENTIFIC AMERICAN, INC