High carbohydrate diets and Alzheimer’s disease

Abstract

Alzheimer's disease (AD) is a common, progressive, neurodegenerative disease that primarily afflicts the elderly. A well-defined risk factor for late onset AD is possession of one or more alleles of the epsilon-4 variant (E4) of the apolipoprotein E gene. Meta-analysis of allele frequencies has found that E4 is rare in populations with long historical exposure to agriculture, suggesting that consumption of a high carbohydrate (HC) diet may have selected against E4 carriers. The apoE4 protein alters lipid metabolism in a manner similar to a HC diet, suggesting a common mechanism for the etiology of AD. Evolutionarily discordant HC diets are proposed to be the primary cause of AD by two general mechanisms. (1) Disturbances in lipid metabolism within the central nervous system inhibits the function of membrane proteins such as glucose transporters and the amyloid precursor protein. (2) Prolonged excessive insulin/IGF signaling accelerates cellular damage in cerebral neurons. These two factors ultimately lead to the clinical and pathological course of AD. This hypothesis also suggests several preventative and treatment strategies. A change in diet emphasizing decreasing dietary carbohydrates and increasing essential fatty acids (EFA) may effectively prevent AD. Interventions that restore lipid homeostasis may treat the disease, including drugs that increase fatty acid metabolism, EFA repletion therapy, and ketone body treatment.

Full text

High carbohydrate diets and Alzheimer’s disease
Samuel T. Henderson
*
Accera Inc. and Institute for Behavioral Genetics, University of Colorado, 1480 30th Street, Boulder,
CO 80303, USA
Received 30 June 2003; accepted 26 November 2003
Summary Alzheimer’s disease (AD) is a common, progressive, neurodegenerative disease that primarily afflicts the
elderly. A well-defined risk factor for late onset AD is possession of one or more alleles of the epsilon-4 variant (E4) of
the apolipoprotein E gene. Meta-analysis of allele frequencies has found that E4 is rare in populations with long
historical exposure to agriculture, suggesting that consumption of a high carbohydrate (HC) diet may have selected
against E4 carriers. The apoE4 protein alters lipid metabolism in a manner similar to a HC diet, suggesting a common
mechanism for the etiology of AD. Evolutionarily discordant HC diets are proposed to be the primary cause of AD by two
general mechanisms. (1) Disturbances in lipid metabolism within the central nervous system inhibits the function of
membrane proteins such as glucose transporters and the amyloid precursor protein. (2) Prolonged excessive insulin/IGF
signaling accelerates cellular damage in cerebral neurons. These two factors ultimately lead to the clinical and
pathological course of AD. This hypothesis also suggests several preventative and treatment strategies. A change in
diet emphasizing decreasing dietary carbohydrates and increasing essential fatty acids (EFA) may effectively prevent
AD. Interventions that restore lipid homeostasis may treat the disease, including drugs that increase fatty acid
metabolism, EFA repletion therapy, and ketone body treatment.

c2004 Elsevier Ltd. All rights reserved.
Introduction
The clinical course of Alzheimer’s disease (AD)
typically begins in the seventh or eighth decade
and is characterized by disturbances in memory,
language, and spatial skills, all of which worsen as
the disease progresses. Upon autopsy, extensive
neuritic plaques and neurofibrillar tangles are
found in the brain, as well as gross structural
changes, such as loss of neurons in the hippocam-
pus, nucleus basalis and other areas (for overview
see [1]). There are no effective treatments and the
disease invariably progresses until death.
The cause of AD has been the subject of intense
debate. The current favored model is the amyloid
cascade hypothesis, which proposes that peptides
generated from the amyloid precursor protein
(APP) are the causative factor and reducing the
generation or accumulation of these peptides will
treat the disease (for overview see [2]). However,
others have proposed that diet may be the primary
cause. In 1997, William Grant correlated the
amounts and types of foods consumed in different
countries with the prevalence of AD and found a
positive association between both total calories
and total fat and the incidence of the disease [3].
Kalmijn et al. [4] also noted a correlation between
fat intake and dementia in a study of 5386 partic-
ipants in Rotterdam. These important studies
pointed toward a strong environmental component
*
Tel.: +1-303-492-5159; fax: +1-303-492-8063.
E-mail address: samuel.henderson@colorado.edu (S.T. Hen-
derson).
0306-9877/$ - see front matter 
c2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mehy.2003.11.028
Medical Hypotheses (2004) 62, 689–700
http://intl.elsevierhealth.com/journals/mehy

to AD and suggested that dietary modification
might prevent the disease. However, follow-up
studies have failed to confirm this link [5]. This
highlights the difficulties in identifying environ-
mental risk factors in large diverse populations
with many variables, some of which may be omit-
ted or hidden by cultural bias. For example, as-
sumptions on what is considered normal intake of
fat, protein and carbohydrate depends greatly on
where and when you look.
The analysis presented here suggests that AD
results not from high-fat diets, but rather from
high-carbohydrate diets (HC). This view is sup-
ported by the genetic association of AD with the
epsilon 4 allele of the apolipoprotein E gene (E4),
the role of lipids in APP processing, and the role of
insulin/IGF signaling in aging. A molecular model is
presented as well as preventative and treatment
strategies. Furthermore, this analysis supports the
view that AD is similar to type II diabetes, obesity,
and coronary heart disease, in that it results from
the conflict between our Paleolithic genetic ma-
keup and our current Neolithic diet.
Agriculture was abomination
The conflict between our genetic makeup and our
diet is similar to the concept of the “thrifty geno-
type” proposed by James Neel in a landmark work
to explain prevalence of type II diabetes in modern
society. “thrifty” was used to mean “... being ex-
ceptionally efficient in the intake and/or utilization
of food.” [6]. He proposed that pre-agricultural
hunter-gatherers went through cycles of feast or
famine which led to the selection of a metabolism
that would readily store fat, and obesity and type II
diabetes result when this genetic makeup is con-
fronted with the modern abundance of food. An
alternative to this model is that the abundance of
food did not change, rather the type of food did.
In a translation of the classic Indian text The
Ramayana, the adoption of agriculture is depicted
not as a revolution, but as an abomination. “In the
Golden Age, agriculture was abomination... For the
existence of sin in the form of cultivation, the
lifespan of people became shortened.” [7]. Such a
view is consistent with the hypothesis that agri-
culture (The Neolithic Revolution) arose out of ne-
cessity rather than cleverness. Several authors have
argued that present day hunter-gatherers are well
aware of the concepts of agriculture but do not
practice it because it requires too much labor [8].
Instead they propose that humans adopted agri-
culture only when wild game became scarce and
they had no other choice (for overview see [9]). In
fact, Paleolithic hunter-gatherers are likely to have
caused the scarcity of wild game. Recent evidence
suggests that they were extremely productive
hunters, especially of big game, and over hunting
was a major factor in the extinction of mega-fauna
in North America [10] and Australia [11].
To understand the dietary shift brought about by
the Neolithic Revolution it is necessary to recon-
struct the Paleolithic diet. In an earlier work, Eaton
and Konner estimated a plant:animal ratio of 65:35
and a fat:protein:carbohydrate ratio of 21:34:45
[12]. In an updated analysis, Cordain et al. [13]
estimate a much higher fat intake. They conclude
that most hunter-gatherers (73%) derived greater
than 50% of their diet from animal sources, sug-
gesting a reversal of the plant:animal ratio from
65:35 to 35:65. They also propose a macronutrient
range of approximately 40:30:30. Both studies
conclude that protein was a significant part of the
diet, while fat and carbohydrate content varied by
location. Those living at higher latitudes tended to
eat more fat, while those in more tropical latitudes
tended to eat more plant matter. Yet, most hunter-
gatherers ate animal matter when available [13].
Such a large protein intake is consistent with the
tall stature of Upper Paleolithic humans [14], and of
pre-agricultural Native Americans [15]. Male Late
Paleolithic hunter-gatherers are estimated to have
been an average of 177 cm tall, similar to average
male heights in the developed world today [16].
It should be noted that the carbohydrates con-
sumed during the Paleolithic period were very dif-
ferent from the high-glycemic carbohydrates found
in modern diets and from the breads and grains
consumed by Neolithic farmers. Consumption of
plant matter does not necessarily result in a large
intake of carbohydrates. For some plant matter, as
much as 30–100% of the energy is released in the
form of short chain fatty acids produced by hind gut
fermentation of fiber [17]. A good analogy might be
modern primate diets. In an analysis of gorilla diet,
which consists almost exclusively of fruits and
vegetables, a macronutrient profile of fat:pro-
tein:carbohydrate was calculated at approximately
3:24:16, with the remaining 57% of the energy in the
form of short chain fatty acids derived from fiber
[18]. Therefore, Paleolithic diets, rich in animal
products and fruits and vegetables, may have been
a low-carbohydrate diet (~20% of energy).
The diets of Neolithic farmers were of much
poorer quality than Late Paleolithic hunter-
gatherers and this has been implicated in the
overall decline in health during the Neolithic period
(for overview see [19]). For example, average
height of a male Late Neolithic farmer was 161 cm,
690 Henderson

a full 16 cm shorter than a male Late Paleolithic
hunter-gatherer [16]. Stature is known to be
strongly influenced by diet, especially protein in-
take, and is frequently used as a measure of nu-
tritional status [16]. It is likely that the Neolithic
diet was very high in carbohydrates and low in
protein, consistent with depletion of wild game as
a major motivator for the development of agri-
culture. This dependence on grain-based agricul-
ture resulted in a long period of reduced stature in
humans. Only in modern times has average height
returned to Late Paleolithic standards [12].
ApoE4 and agriculture
While the shift to HC diets during the Neolithic Rev-
olution resulted in a general decline in health, it
proved particularly disastrous to carriers of the ep-
silon 4 allele of apolipoprotein E. Currently, the only
well defined genetic risk factor for late onset Alz-
heimer’s disease is allelic variation in the apolipo-
protein E gene (apoE). The main function of the apoE
protein is lipid transport, but as such, it has an im-
pact on a variety of cellular processes. There are
three common allelic variants of apoE: epsilon 2 (E2),
3 (E3) and 4 (E4) (for review see [20]). Possession of
the E4 variant increases the risk of developing AD and
behaves in a dominant dose dependent manner [21].
E4 is also a risk factor for coronary heart disease [22]
and poor recovery from head trauma [23]. Why such
a “deleterious” allele would be selected against in
some populations but not in others may provide an
important clue to the etiology of AD.
The alleles E2, E3 and E4 are not evenly dis-
tributed in all populations. In a meta-analysis of
published apoE allele frequencies, Corbo and
Scacchi noted that E4 is under-represented in
populations with long historical exposure to agri-
culture, and they proposed that E4 may be a thrifty
allele [24]. Populations with the lowest frequencies
of E4 include long time agriculturalists, such as
Greeks (0.068) and Turks (0.079), while popula-
tions with the highest frequencies include long
time hunter-gatherers, such as African Pygmies
(0.407), Papuans (0.368), and Inuits (0.214) [24].
This has been supported by a study of Arab popu-
lations living in northern Israel who had the lowest
E4 frequency ever recorded (0.04) [25]. One in-
terpretation of this distribution is genetic drift due
to migration of populations out of the Middle East
(see Fig. 1). The migration of Neolithic farmers
Figure 1 Distribution of apolipoprotein E epsilon 4 allele (E4), adapted from [24,25]. Frequency of E4 is low (light
regions) in historically agriculture-based societies in the Middle East and in Central America (inset). Arrows indicate
migration of Neolithic farmers from the Middle East along the Mediterranean Sea.
691High carbohydrate diets and Alzheimer’s disease

along the Mediterranean Sea is based on historical
records and confirmed by analysis of the Y chro-
mosome [26]. However, such a migration does not
explain the low frequency of E4 found in North
American Mayan populations (0.089) [24]. The Ma-
yan civilization arose in what is present day Mexico
and Guatemala, far from the Middle East and in a
very different ecological environment. Yet, the
Mayans were similar to Middle Eastern farmers in
that they also developed an extensive agricultural
society based primarily on maize. Therefore, con-
sumption of an HC diet, either derived from corn or
wheat, may have selected against the E4 allele.
Why would E4 be deleterious to populations
consuming a HC diet? Possession of an E4 allele is
frequently associated with elevated plasma cho-
lesterol and LDL-cholesterol levels and this asso-
ciation is normally attributed to consumption of a
high-fat, high-cholesterol diet (for overview see
[27]). One interpretation of such a view is that
Neolithic farmers ate a diet much like a modern
Western atherogenic diet, rich in animal fat and
cholesterol, and that E4 was selected against by
widespread prevalence of coronary heart disease.
This seems unlikely. It is more likely that Neolithic
farmers ate little animal matter, were protein
starved, and are better characterized as “...ant-
like armies of largely vegetarian workers.” [9]. In-
stead it can be argued that HC diets and possession
of an E4 allele both suppress lipid metabolism in a
similar manner and, in combination, greatly in-
crease the risk for coronary heart disease and AD.
HC diets, ApoE and lipid metabolism
The effect of HC diets on lipid metabolism is evi-
dent in the fate of triglyceride rich lipoproteins
(TRL), such as chylomicrons and very low density
lipoproteins (VLDL) (for overview see [28]). The
rate of clearance of TRL and the type of cells that
take up free fatty acids (FFA) depends mainly on
the activity of lipoprotein lipases (LPL) and is
strongly influenced by insulin signaling [29]. It is
well recognized that HC diets elevate VLDL levels
and can result in hypertriacylglycerolemia (for re-
view see [30]). This may be due to decreased LPL
activity and fatty acid use by muscle cells [29,31].
For example, Lithell et al. [32] studied 7 men fed
either high carbohydrate (HC, greater than 70% of
calories from carbohydrate) or high fat (HF,
greater than 70% of calories from fat) diets for 3
days. Those consuming the HC diet had statistically
increased insulin levels and decreased lipase ac-
tivities relative to the HF diet. These experiments
are consistent with the proposed fuel use hierarchy
in humans, such that glucose is used preferentially
over fat [33]. In particular, HC diets inhibit the use
of fatty acids and increase the residence time of
TRL (for review see [34]).
Much like a HC diet, the ApoE4 protein increases
TRL residence time by inhibiting lipolysis. ApoE4
binds TRL much more readily than ApoE2 or ApoE3
and will displace ApoCII resulting in decreased LPL
activity (see Fig. 2(I)), for review see [35]). For
example, in a study of young men consuming a high
fat meal (39% fat calories) TRL were elevated in
both E3/E3 and E3/E4 individuals. After 6 h, TRL
returned to post-absorptive values in the E3/E3
individuals, yet remained elevated 50–80% in E3/
E4 individuals [36]. Such elevated TRL have been
observed numerous times in E4 carriers. In a meta-
analysis of different populations, E4 carriers had
significantly higher plasma triglyceride levels than
those with E3 [37]. This decreased LPL activity may
also be the cause of the increased insulin sensi-
tivity observed in E4 carriers [38] possibly due to
lowering of serum FFA levels (see Fig. 2(III)) (for
overview see [39]).
Since E4 and HC diets inhibit lipid metabolism in
a similar manner this may explain the selection
against E4 in long-time agricultural societies.
Symptoms of AD do not typically begin until the
seventh decade so it is unlikely that AD was the
selective force, instead it may have been coronary
heart disease (CHD). The very HC diet of Neolithic
farmers would have raised serum glucose and in-
sulin levels, induced lipogenesis and led to hyper-
triacylglycerolemia. This would be worsened by
possession of an E4 allele. Elevated triglycerides
increase the risk of CHD and this may have selected
against E4 in Neolithic farmers. Importantly, this
same mechanism is likely to be responsible for the
high risk of CHD in modern populations, with or
without an E4 allele.
It should be noted that E4 is not an inherently
damaging allele, it is only deleterious in combina-
tion with a HC diet (which is deleterious on its
own). Populations with little exposure to HC diets
have higher E4 frequencies suggesting it is not se-
lected against in these conditions [24]. Also, E4
may not be a risk factor for AD in all populations,
such as in Nigeria [40,41]. Nigerians eat consider-
ably less high-glycemic carbohydrates than at risk
populations such as the US. For example, in 1999,
Nigerians consumed 20 kg/year of sugar per capita,
compared with 74 kg/year for a typical American
(source Food and Agriculture Organization of the
United Nations statistical database, FAOSTAT).
This may explain the low incidence of AD in Nigeria
[42] despite the relatively high frequency of E4.
692 Henderson

Prior to the development of agriculture, E2, E3 and
E4 may have been neutral alleles that arose when
our human ancestors began to eat more animal
matter, and hence more fat, and this relaxed se-
lection on apoE. The development of agriculture
then imposed a new selection on apoE reducing E4
in Middle Eastern and Mayan populations.
Overview of the etiology of AD
HC diets are proposed as the primary cause of AD
by two basic mechanisms (see Fig. 3 for overview).
The first is disturbed lipid homeostasis within the
CNS, especially decreased delivery of essential
fatty acids (EFA) (see Fig. 3(I)). This compromises
the integrity of cellular membranes, decreasing the
function of membrane proteins such as glucose
transporters and APP. The second is mild chronic
elevated insulin/IGF signaling, which accelerates
cellular damage (Fig. 3(II)). These two mechanisms
contribute to two stages of the disease. Stage I
begins when altered lipid metabolism inhibits the
function of membrane proteins such as glucose
transporters, resulting in decreased glucose utili-
zation and lowered metabolism in susceptible re-
gions of the brain. At this stage no clinical signs of
dementia are evident, yet the disease has begun.
Stage II begins when the inhibition of cellular
function can no longer be compensated for, either
due to excessive cellular damage, or age impaired
loss of homeostatic mechanisms. In stage II, acetyl-
CoA levels are lowered below critical levels, af-
fecting the production of a variety of cellular
components such as cholesterol and acetylcholine
and clinical signs of dementia become evident. The
disturbances in cholesterol metabolism result in
large scale aberrant processing of APP, decreases
in cellular trafficking, and generation of amyloid
beta peptides (Ab). As the disease progresses, the
failure to transport neurotrophin receptors and the
production of increasing amounts of Abultimately
results in large scale cell death and the charac-
teristic pathology of AD.
Stage I – essential fatty acids and
membrane function
Despite the importance of fatty acids in cerebral
neurons little de novo fatty acid synthesis occurs in
the adult brain (for overview see [43]). Most fatty
acids are imported as phospholipids or unesterified
FFA from the plasma through the use of fatty acid
transport proteins (for review see [44]). One im-
portant class of fatty acids required by the CNS are
EFA. For example, docosahexanoic acid (DHA) is
found extensively in phospholipids of neuronal
membranes (for overview see [45]). Inhibition of
lipid metabolism by HC diets may mimic dietary
Figure 2 ApoE4 and a high carbohydrate diet inhibit lipid metabolism. Bold Roman numerals indicate key points in
the model. (I) E4 preferentially binds triglyceride rich particles such as VLDL and chylomicrons reducing ApoCII binding.
(II) Decreased LPL activity inhibits delivery of FFA to astrocytes. (III) Low FFA levels increase insulin sensitivity and
further decrease LPL activity and ketone body transport. (IV) Inefficient delivery of EFA to cerebral neurons inhibits
function of glucose transporters (GLUT). (V) Decreased metabolism lowers acetyl-CoA pools and levels of ATP and
acetylcholine. (VI) Increased insulin signaling inhibits Foxo proteins from entering the nucleus and prevents activation
of stress response genes, such as antioxidant proteins. Abbreviations: LRP – LDL receptor related protein, VLDL – very
low density lipoprotein, HDL – high density lipoprotein, LPL – lipoprotein lipase, FFA – free fatty acid, MCTr –
monocarboxylate transporter, E4 – ApoE4, FATP – fatty acid transport protein, KB – ketone bodies, ACh – acetyl-
choline, ROS – reactive oxygen species.
693High carbohydrate diets and Alzheimer’s disease

deficiencies of EFA which are known to alter the
composition of neuronal membranes, disturb the
activity of membrane proteins [46], and lead to
behavioral defects such as poor performance in
learning tasks (for overview see [47]). This is con-
sistent with the growing evidence that EFA play a
role in AD. Low serum DHA levels have been im-
plicated as a significant risk factor [48,49], and
consumption of fish (a rich source of DHA and EPA)
may prevent the disease [50]. Additionally, altered
lipid metabolism may be responsible for the
extensive membrane deterioration seen in AD
[51,52].
In addition to metabolic changes induced by HC
diets, the development of agriculture has directly
changed the normal dietary balance of EFA. It has
been estimated that Paleolithic hunter-gatherers
ate roughly equal amounts of n6 and n3 fatty
acids [12]. However, the modern Western food
supply is much richer in n6 fatty acids due to the
use of grains both in the diet and as animal feed.
This has greatly altered the ratio of dietary n6to
n3 fatty acids from roughly 1:1 for Paleolithic
hunter-gatherers to 20:1 for a modern diet. The
n6 fatty acids compete for desaturases used by
n3 fatty acids to produce products such as DHA,
essentially lowering their levels (for review see
[53]).
One class of protein known to be effected by EFA
levels are glucose transporters. For example, rats
raised on a n3 deficient diet for three months
exhibit a 30–35% decrease in glucose uptake in the
cortex, hippocampus and SCN compared to ad lib
fed controls, due to inefficient function of glucose
transporters [54]. Such decreases in cerebral glu-
cose utilization are one of the earliest signs of AD
and are evident in at risk populations well before
clinical signs of dementia occur, particularly in E4
carriers [55]. Yet, at this early stage, poor cogni-
tive performance may be masked by recruiting
larger regions of the brain to accomplish mental
tasks [56]. As the disease progresses, inhibition of
glucose use worsens (for overview see [57]), and at
some point declines to where recruitment can no
longer compensate for energy loss (Fig. 3(IV)). This
is the beginning of Stage II.
Stage II – metabolism, cholesterol and
APP
Cerebral neurons are normally considered to derive
acetyl-CoA almost exclusively from glucose. As
glucose utilization worsens it will begin to deplete
neuronal acetyl-CoA pools leading to decreased
synthesis of acetylcholine (Fig. 3(IV)) and the well
recognized cholinergic defects found in AD [58].
Another less obvious, but perhaps more important,
consequence of lower acetyl-CoA levels is altera-
Figure 3 HC diet and Alzheimer’s disease model overview. Light and shaded areas indicate stages of the disease.
Bold Roman numerals highlight major mechanisms of disease progression. (I) HC diet and ApoE4 contribute to de-
creased lipid metabolism in central nervous system, altering the function of glucose transporters and amyloid pre-
cursor protein (APP). (II) Chronic excessive insulin/IGF signaling inhibits the functioning of Foxo proteins thereby
increasing cellular damage.
694 Henderson

tions in cholesterol homeostasis (Fig. 4(II)). The
human brain contains large amounts of unesterified
cholesterol, roughly 25% of the total amount in the
body. Unlike FFA, cholesterol is synthesized de
novo within neuronal cells from condensation of
acetyl-CoA and is part of a complex regulatory
process of cholesterol homeostasis (for review see
[59]). Disturbance in this process has been impli-
cated in several neurological disorders, including
AD. For example, allelic variation in the cyp46 gene
(a cholesterol 24-hydroxylase) has been identified
as a risk factor [60] and decreased cholesterol
levels are found in affected regions of the brain
[61].
One important protein that is sensitive to dis-
turbances in cholesterol homeostasis is APP
(Fig. 4(III)). Early onset AD is frequently associated
with mutations in three genes; APP, presenilin 1
(PS1) and presenilin 2 (PS2). These mutations lead
to aberrant processing of the APP protein and ac-
cumulation of the Abpeptide (for review see [2]).
Recent evidence has suggested that excess cho-
lesterol leads to increased APP cleavage. Diet in-
duced hypercholesterolemia increases the levels of
Aband amyloid deposits in the CNS of transgenic
mouse models of AD [62,63]. Addition of excess
cholesterol to cells in culture increases Abpro-
duction, while depleting cells of cholesterol de-
creases Abproduction (for overview see [64]).
Also, treating animals with cholesterol lowering
drugs (statins) decreases the levels of Abin the
blood [65] and may decrease the risk of developing
AD up to 70% [66]. Yet, most statin drugs do not
cross the blood brain barrier and are predicted to
have a weak, if any, effect on cerebral cholesterol
production (for overview see [67]).
Alternatively, statins may protect against AD by
improving cerebral lipid metabolism. In addition to
inhibition of 3-hydroxy-3-methylglutaryl CoA re-
ductase, statins have other physiologic effects,
such as vasodilatory and anti-inflammatory. Im-
portantly, statins also cause a reduction in circu-
lating TRL by increasing the levels of lipoprotein
lipase while also decreasing apolipoprotein C-III (an
inhibitor of lipoprotein lipase) [68]. Therefore
statins may directly counteract the effects of HC
diets by increasing the activity of LPL.
Abmay not be the only toxic result of aberrant
APP processing. APP is proposed to function as a
membrane cargo receptor for kinesin-I during ax-
onal transport, delivering several cellular factors,
including Bace (beta secretase), Ps1 (Presenilin 1),
and the neurotrophin receptor TrkA [69–71]. Mu-
tations in APP and the presenilins, as well as dis-
turbed cholesterol homeostasis, may lead to
premature cleavage of APP and inhibition of cel-
Figure 4 Lipid homeostasis and APP. Bold Roman numerals indicate key points in the model. (I) High carbohydrate
diets inhibit efficient lipid delivery to the brain, inhibiting function of glucose transporters and lowering acetyl-CoA
pools. (II) Low EFA and acetyl-CoA levels inhibit cholesterol metabolism and membrane function and integrity. (III)
Inability to maintain lipid homeostasis results in improper processing of APP and Abgeneration. Premature cleavage of
APP results in failure to deliver neurotrophin receptors to cell surface and cell death. (IV) Accumulation of toxic Ab.
Abbreviations: EFA – essential fatty acid, MCTr – monocarboxylate transporter, KB – ketone bodies.
695High carbohydrate diets and Alzheimer’s disease

lular trafficking [72]. Failure to deliver neurotro-
phin receptors would lead to widespread neuronal
cell death (Fig. 4, for review see [73]). In fact,
inhibiting NGF in the brains of mice results in an
age dependent pathology very similar to AD [74].
Stage I/II – insulin/IGF signaling and
aging
HC diets are well known to increase glucose and
insulin levels in humans [31] and this elevated in-
sulin signaling may lead to rapid aging of suscep-
tible tissues. In mammals and lower organisms
there is growing evidence that insulin/IGF signaling
modulates lifespan (for overview see [75]). For
example, reducing the caloric intake of mice and
rats reduces insulin/IGF levels and increases life
span (for review see [76]). More direct evidence
comes from the observation that mice heterozy-
gous for the IGF-1 receptor live 33% longer than
their wild-type littermates [77] and mice lacking
the insulin receptor in fat cells live 18% longer
[78].
The insulin-like signaling pathway shows re-
markable conservation across phyla. In both nem-
atodes and mammals insulin/IGF signaling
negatively regulates the activity of the Foxo family
of transcription factors by sequestration in the
cytoplasm (for review see [79,80]). Activation of
Foxo proteins increases stress resistance and lon-
gevity in mice and nematodes [75]. The long-lived
p66shc()/)) mouse may have increased Foxo ac-
tivation and increased resistance to oxidative
stress [81]. Activation of FKHR (a Foxo protein)
increases expression of stress response genes, such
as Gadd45a, a gene involved in DNA repair [82]. It
has been proposed that insulin/IGF signaling func-
tions, via Foxo proteins, to adjust metabolism and
ultimately lifespan in response to nutritional and
environmental cues [83,84]. Low food availability
will increase the proportion of Foxo in the nucleus
and increase the expression of a variety of stress
resistance genes, resulting in more stress resistant
longer lived individuals. High food availability will
decrease the expression of stress genes, resulting
in less stress resistant shorter lived individuals.
The mammalian brain is well supplied with in-
sulin receptors where insulin appears to signal
abundant food and not trigger glucose uptake as it
does in muscle and fat [85,86]. For example,
chronic infusion of insulin into the brains of ba-
boons reduces food intake [87], while inhibition of
the insulin receptor in the brains of mice in-
creases food intake [88,89]. Therefore, the strong
increases in postprandial glucose and insulin levels
induced by HC diets may continuously signal that
nutrients are plentiful, exclude Foxo from the nu-
cleus, and accelerate aging of susceptible neurons
(Fig. 3(VI)). This condition will be exacerbated in
E4 individuals who are more insulin sensitive.
Treatment and prevention
This hypothesis suggests several treatment and
preventative measures that may be beneficial for
AD and other disorders resulting from what can be
collectively called the “Neolithic Syndrome”.
Such treatment may be especially effective in
combination.
The Paleolithic prescription
A modified “Paleolithic prescription” [90] may
prevent AD. The Paleolithic prescription proposes a
change in diet and activity to a level more similar
to our Late Paleolithic ancestors, and emphasizes
reducing fat and increasing dietary fiber as the keys
to better health [90,91]. However, the inhibition of
lipid metabolism by HC diets may be the most
detrimental aspect of modern diets. Therefore,
reducing dietary intake of high-glycemic carbohy-
drates and increasing protein, fiber and fat would
be preferred. Similar diets appear to reduce the
risk of AD [92]. Since HC diets are proposed to be
the primary cause of AD regardless of apoE geno-
type, such a diet would generally reduce the risk of
AD. However, this diet is predicted to be particu-
larly beneficial to carriers of apoE4, and suggests
that individuals should “eat right for your apoE
type”. Dietary change would be the preventative
treatment of choice, since it would not only lower
the incidence of AD, but many other harmful con-
ditions. Yet such a change would require dramatic
decreases in carbohydrate intake (to < 30% of daily
caloric intake) and would be difficult to implement
without drastic changes in dietary thinking.
EFA repletion diet
Increasing evidence has implicated consumption of
fish (a source of EFA) as protective against AD [50].
For individuals in Stage I or II, an EFA repletion
regime, consisting of high doses of EFA, may re-
plenish EFA in neuronal membranes and prevent
and/or treat the disease [93]. In particular, ele-
vation of n3 EFA may allow for more efficient
696 Henderson

function of glucose transporters and the APP
protein.
Ketone body treatment
While increasing fatty acid metabolism may help
prevent the disease, by the time clinical dementia
is diagnosed (Stage II) irreparable damage may
have occurred and reversal will be difficult. One
strategy that might be effective is direct elevation
of acetyl-CoA levels using ketone bodies (KB). In-
creasing acetyl-CoA levels will provide a substrate
for acetylcholine and cholesterol synthesis and can
be used in the TCA cycle [94]. A simple way to el-
evate plasma KB levels is through consumption of
medium chain triglycerides, which are readily me-
tabolized to KB. We have found that exogenous
administration of medium chain triglycerides in-
creased cognitive performance in early stage non-
E4 AD patients [95].
Increasing fatty acid metabolism
Drugs that increase the use of fatty acids, espe-
cially in glia, may be beneficial for AD. This may
explain the beneficial effects of statins (as dis-
cussed) and non-steroid anti-inflammatory drugs
(NSAIDS) [96]. NSAIDS function, in part, as PPAR-
gamma agonists. Increasing PPAR-gamma activity
increases the expression of genes associated with
fatty acid metabolism such as FATP (for review see
[97]). Other drugs may have similar effects. Fibrate
drugs, such as Bezafibrate, ciprofibrate, fenofi-
brate and Gemfibrozil may also prove beneficial.
Fibrates act as PPAR-alpha agonists and like statins
they increase lipoprotein lipase, apoAI and apoAII
transcription and reduce levels of apoCIII, thereby
increasing lipid availability to the brain [98].
Conclusion
AD is a devastating neurodegenerative disorder
that will reach epidemic proportions in the next 50
years. While tremendous progress has been made
in our molecular understanding of the disease, no
effective treatments exist. Much of the current
research centers on modulating the processing of
the APP protein and correcting the imbalance be-
tween Abproduction and clearance. This approach,
while promising, has many drawbacks. Altering the
processing of APP may affect other proteins such as
Notch and is technically difficult [99]. Here it is
argued that the primary event leading to the de-
velopment of AD is consumption of an evolution-
arily discordant HC diet. This hypothesis predicts
that relatively simple preventative measures, such
as lowering the consumption of starchy carbohy-
drates and increasing EFA in the diet will be ef-
fective. Yet, in practice this may be difficult
without sufficient public awareness. Other treat-
ments may also be effective, such as ketone body
therapy, EFA repletion diets, and statin drugs.
Hopefully, in the future more research will focus
on the role of diet in AD.
Acknowledgements
This work is dedicated to Florence Tomlins Hen-
derson. I wish to thank members of the Johnson lab
for critical reading of the manuscript and active
discussion. I am grateful to Dr. Thomas Johnson for
allowing me the time and freedom to pursue my
interests.
References
[1] Sisodia SS, Martin LJ, Walker LC, Borchelt DR, Price DL.
Cellular and molecular biology of Alzheimer’s disease and
animal models. Neuroimaging Clin N Am 1995;5:59–68.
[2] Selkoe DJ. Alzheimer’s disease: genes, proteins, and
therapy. Physiol Rev 2001;81:741–66.
[3] Grant WB. Dietary links to Alzheimer’s disease. Alzheimer’s
Disease Rev 1997;2:42–55.
[4] Kalmijn S, Launer LJ, Ott A, Witteman JC, Hofman A,
Breteler MM. Dietary fat intake and the risk of incident
dementia in the Rotterdam Study. Ann Neurol
1997;42:776–82.
[5] Engelhart MJ, Geerlings MI, Ruitenberg A, et al. Diet and
risk of dementia: Does fat matter?: The Rotterdam study.
Neurology 2002;59:1915–21.
[6] Neel JV. Diabetes mellitus: A thrifty genotype rendered
detrimental by progress? Am J Hum Genet 1962;14:353–62.
[7] Mehta NG. Did agriculture reduce human lifespan? Nature
2001;409:131.
[8] Gowdy JM. Limited wants, unlimited means: a reader on
hunter-gatherer economics and the environment. Washing-
ton, DC: Island Press; 1998.
[9] Harris M. Cannibals and Kings: the origins of cultures. New
York: Random House; 1977.
[10] Alroy J. A multispecies overkill simulation of the end-
Pleistocene megafaunal mass extinction. Science
2001;292:1893–6.
[11] Roberts RG, Flannery TF, Ayliffe LK, et al. New ages for the
last Australian megafauna: continent-wide extinction about
46,000 years ago. Science 2001;292:1888–92.
[12] Eaton SB, Konner M. Paleolithic nutrition. A consideration
of its nature and current implications. N Engl J Med
1985;312:283–9.
[13] Cordain L, Miller JB, Eaton SB, Mann N. Macronutrient
estimations in hunter-gatherer diets. Am J Clin Nutr
2000;72:1589–92.
697High carbohydrate diets and Alzheimer’s disease

[14] Formicola V, Giannecchini M. Evolutionary trends of stature
in upper Paleolithic and Mesolithic Europe. J Hum Evol
1999;36:319–33.
[15] Steckel R, Prince JM. Tallest in the World: Native Ameri-
cans of the Great Plains in the Nineteenth Century. AER
2001;91:287–94.
[16] Angel LJ. Health as a crucial factor in the changes from
hunting to developed farming in the eastern Mediterranean.
In: Cohen MN, Armelagos GJ, editors. Paleopathology at the
origins of agriculture. Orlando: Academic Press; 1984. p.
51–73.
[17] Cummings JH. Short chain fatty acids in the human colon.
Gut 1981;22:763–79.
[18] Popovich DG, Jenkins DJ, Kendall CW, et al. The western
lowland gorilla diet has implications for the health of
humans and other hominoids. J Nutr 1997;127:2000–5.
[19] Cohen MN, Armelagos GJ, editors. Paleopathology at the
origins of agriculture. Orlando, FL: Academic Press; 1984.
[20] Mahley RW, Rall Jr SC. Apolipoprotein E: far more than a
lipid transport protein. Ann Rev Genomics Hum Genet
2000;1:507–37.
[21] Corder EH, Saunders AM, Strittmatter WJ, et al. Gene dose
of apolipoprotein E type 4 allele and the risk of Alzheimer’s
disease in late onset families. Science 1993;261:921–3.
[22] Davignon J, Gregg RE, Sing CF. Apolipoprotein E polymor-
phism and atherosclerosis. Arteriosclerosis 1988;8:1–21.
[23] Nicoll JA, Roberts GW, Graham DI. Amyloid beta-protein,
APOE genotype and head injury. Ann N Y Acad Sci
1996;777:271–5.
[24] Corbo RM, Scacchi R. Apolipoprotein E (APOE) allele
distribution in the world. Is APOE*4 a ‘thrifty’ allele? Ann
Hum Genet 1999;63(Pt 4):301–10.
[25] Bowirrat A, Friedland RP, Chapman J, Korczyn AD. The very
high prevalence of AD in an Arab population is not explained
by APOE epsilon4 allele frequency. Neurology 2000;55:731.
[26] Semino O, Passarino G, Oefner PJ, et al. The genetic legacy
of Paleolithic Homo sapiens sapiens in extant Europeans: a
Y chromosome perspective. Science 2000;290:1155–9.
[27] Ordovas JM. The genetics of serum lipid responsiveness to
dietary interventions. Proc Nutr Soc 1999;58:171–87.
[28] Murray RK, Granner DK, Mayes PA, Rodwell VW. Harper’s
biochemistry. 25th ed. New York: McGraw-Hill; 1999. p.
927.
[29] Campos H, Dreon DM, Krauss RM. Associations of hepatic
and lipoprotein lipase activities with changes in dietary
composition and low density lipoprotein subclasses. J Lipid
Res 1995;36:462–72.
[30] Parks EJ, Hellerstein MK. Carbohydrate-induced hypertria-
cylglycerolemia: historical perspective and review of bio-
logical mechanisms. Am J Clin Nutr 2000;71:412–33.
[31] Ginsberg H, Olefsky JM, Kimmerling G, Crapo P, Reaven GM.
Induction of hypertriglyceridemia by a low-fat diet. J Clin
Endocrinol Metab 1976;42:729–35.
[32] Lithell H, Jacobs I, Vessby B, Hellsing K, Karlsson J.
Decrease of lipoprotein lipase activity in skeletal muscle
in man during a short-term carbohydrate-rich dietary
regime. With special reference to HDL-cholesterol, apoli-
poprotein and insulin concentrations. Metabolism
1982;31:994–8.
[33] Hellerstein MK. No common energy currency: de novo
lipogenesis as the road less traveled. Am J Clin Nutr
2001;74:707–8.
[34] Hellerstein MK. Carbohydrate-induced hypertriglyceride-
mia: modifying factors and implications for cardiovascular
risk. Curr Opin Lipidol 2002;13:33–40.
[35] Weisgraber KH. Apolipoprotein E: structure–function rela-
tionships. Adv Protein Chem 1994;45:249–302.
[36] Bergeron N, Havel RJ. Prolonged postprandial responses of
lipids and apolipoproteins in triglyceride-rich lipoproteins
of individuals expressing an apolipoprotein epsilon 4 allele.
J Clin Invest 1996;97:65–72.
[37] Dallongeville J, Lussier-Cacan S, Davignon J. Modulation of
plasma triglyceride levels by apoE phenotype: a meta-
analysis. J Lipid Res 1992;33:447–54.
[38] Craft S, Asthana S, Schellenberg G, et al. Insulin metabo-
lism in Alzheimer’s disease differs according to apolipo-
protein E genotype and gender. Neuroendocrinology
1999;70:146–52.
[39] Randle PJ. Regulatory interactions between lipids and
carbohydrates: the glucose fatty acid cycle after 35 years.
Diabetes Metab Rev 1998;14:263–83.
[40] Osuntokun BO, Sahota A, Ogunniyi AO, et al. Lack of an
association between apolipoprotein E epsilon 4 and Alzhei-
mer’s disease in elderly Nigerians. Ann Neurol
1995;38:463–5.
[41] Kalaria RN, Ogeng’o JA, Patel NB, et al. Evaluation of risk
factors for Alzheimer’s disease in elderly east Africans.
Brain Res Bull 1997;44:573–7.
[42] Hendrie HC, Ogunniyi A, Hall KS, et al. Incidence of
dementia and Alzheimer disease in 2 communities: Yoruba
residing in Ibadan, Nigeria, and African Americans residing
in Indianapolis, Indiana. Jama 2001;285:739–47.
[43] Edmond J. Essential polyunsaturated fatty acids and the
barrier to the brain: the components of a model for
transport. J Mol Neurosci 2001;16:181–93, discussion 215-
21.
[44] Zhou L, Nilsson A. Sources of eicosanoid precursor fatty
acid pools in tissues. J Lipid Res 2001;42:1521–42.
[45] Agranoff BW, Hajra AK. Lipids. In: Siegel GJ, editor. Basic
neurochemistry. New York: Raven; 1993.
[46] Bourre JM, Pascal G, Durand G, Masson M, Dumont O,
Piciotti M. Alterations in the fatty acid composition of rat
brain cells (neurons, astrocytes, and oligodendrocytes) and
of subcellular fractions (myelin and synaptosomes) induced
by a diet devoid of n3 fatty acids. J Neurochem
1984;43:342–8.
[47] Youdim KA, Martin A, Joseph JA. Essential fatty acids and
the brain: possible health implications. Int J Dev Neurosci
2000;18:383–99.
[48] Kyle DJ, Schaefer E, Patton G, Beiser A. Low serum
docosahexaenoic acid is a significant risk factor for Alzhei-
mer’s dementia. Lipids 1999;34(Suppl):S245.
[49] Conquer JA, Tierney MC, Zecevic J, Bettger WJ, Fisher RH.
Fatty acid analysis of blood plasma of patients with
Alzheimer’s disease, other types of dementia, and cogni-
tive impairment. Lipids 2000;35:1305–12.
[50] Morris MC, Evans DA, Bienias JL, et al. Consumption of fish
and n3 fatty acids and risk of incident Alzheimer disease.
Arch Neurol 2003;60:940–6.
[51] Nitsch RM, Blusztajn JK, Pittas AG, Slack BE, Growdon JH,
Wurtman RJ. Evidence for a membrane defect in Alzheimer
disease brain. Proc Natl Acad Sci USA 1992;89:1671–5.
[52] Prasad MR, Lovell MA, Yatin M, Dhillon H, Markesbery WR.
Regional membrane phospholipid alterations in Alzheimer’s
disease. Neurochem Res 1998;23:81–8.
[53] Simopoulos AP. Essential fatty acids in health and chronic
disease. Am J Clin Nutr 1999;70:560S–9S.
[54] Ximenes da Silva A, Lavialle F, Gendrot G, Guesnet P,
Alessandri JM, Lavialle M. Glucose transport and utilization
are altered in the brain of rats deficient in n3 polyun-
saturated fatty acids. J Neurochem 2002;81:1328–37.
[55] Reiman EM, Caselli RJ, Chen K, Alexander GE, Bandy D,
Frost J. Declining brain activity in cognitively normal
apolipoprotein E epsilon 4 heterozygotes: a foundation for
698 Henderson

using positron emission tomography to efficiently test
treatments to prevent Alzheimer’s disease. Proc Natl Acad
Sci USA 2001;98:3334–9.
[56] Bookheimer SY, Strojwas MH, Cohen MS, et al. Patterns of
brain activation in people at risk for Alzheimer’s disease. N
Engl J Med 2000;343:450–6.
[57] Swaab DF, Lucassen PJ, Salehi A, Scherder EJ, van Someren
EJ, Verwer RW. Reduced neuronal activity and reactivation
in Alzheimer’s disease. Prog Brain Res 1998;117:343–77.
[58] White P, Hiley CR, Goodhardt MJ, et al. Neocortical
cholinergic neurons in elderly people. Lancet
1977;1:668–71.
[59] Dietschy JM, Turley SD. Cholesterol metabolism in the
brain. Curr Opin Lipidol 2001;12:105–12.
[60] Papassotiropoulos A, Streffer JR, Tsolaki M, et al. Increased
brain beta-amyloid load, phosphorylated tau, and risk of
Alzheimer disease associated with an intronic CYP46 poly-
morphism. Arch Neurol 2003;60:29–35.
[61] Mason RP, Shoemaker WJ, Shajenko L, Chambers TE,
Herbette LG. Evidence for changes in the Alzheimer’s
disease brain cortical membrane structure mediated by
cholesterol. Neurobiol Aging 1992;13:413–9.
[62] Howland DS, Trusko SP, Savage MJ, et al. Modulation of
secreted beta-amyloid precursor protein and amyloid beta-
peptide in brain by cholesterol. J Biol Chem
1998;273:16576–82.
[63] Refolo LM, Malester B, LaFrancois J, et al. Hypercholester-
olemia accelerates the Alzheimer’s amyloid pathology in a
transgenic mouse model. Neurobiol Dis 2000;7:321–31.
[64] Pappolla MA, Smith MA, Bryant-Thomas T, et al. Choles-
terol, oxidative stress, and Alzheimer’s disease: expanding
the horizons of pathogenesis (1). Free Radic Biol Med
2002;33:173–81.
[65] Friedhoff LT, Cullen EI, Geoghagen NS, Buxbaum JD.
Treatment with controlled-release lovastatin decreases
serum concentrations of human beta-amyloid (A beta)
peptide. Int J Neuropsychopharmacol 2001;4:127–30.
[66] Wolozin B, Kellman W, Ruosseau P, Celesia GG, Siegel G.
Decreased prevalence of Alzheimer disease associated with
3-hydroxy-3-methyglutaryl coenzyme A reductase inhibi-
tors. Arch Neurol 2000;57:1439–43.
[67] Wolozin B. Cholesterol and Alzheimer’s disease. Biochem
Soc Trans 2002;30:525–9.
[68] Schoonjans K, Peinado-Onsurbe J, Fruchart JC, Tailleux A,
Fievet C, Auwerx J. 3-Hydroxy-3-methylglutaryl CoA reduc-
tase inhibitors reduce serum triglyceride levels through
modulation of apolipoprotein C-III and lipoprotein lipase.
FEBS Lett 1999;452:160–4.
[69] Kamal A, Stokin GB, Yang Z, Xia CH, Goldstein LS. Axonal
transport of amyloid precursor protein is mediated by
direct binding to the kinesin light chain subunit of kinesin-I.
Neuron 2000;28:449–59.
[70] Gunawardena S, Goldstein LS. Disruption of axonal trans-
port and neuronal viability by amyloid precursor protein
mutations in Drosophila. Neuron 2001;32:389–401.
[71] Kamal A, Almenar-Queralt A, LeBlanc JF, Roberts EA,
Goldstein LS. Kinesin-mediated axonal transport of a
membrane compartment containing beta-secretase and
presenilin-1 requires APP. Nature 2001;414:643–8.
[72] Lynch C, Mobley W. Comprehensive theory of Alzheimer’s
disease. The effects of cholesterol on membrane receptor
trafficking. Ann N Y Acad Sci 2000;924:104–11.
[73] Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal
development and function. Annu Rev Neurosci
2001;24:677–736.
[74] Capsoni S, Ugolini G, Comparini A, Ruberti F, Berardi N,
Cattaneo A. Alzheimer-like neurodegeneration in aged
antinerve growth factor transgenic mice. Proc Natl Acad
Sci USA 2000;97:6826–31.
[75] Guarente L, Kenyon C. Genetic pathways that regulate
ageing in model organisms. Nature 2000;408:255–62.
[76] Masoro EJ. Caloric restriction. Aging (Milano) 1998;10:
173–4.
[77] Holzenberger M, Dupont J, Ducos B, et al. IGF-1 receptor
regulates lifespan and resistance to oxidative stress in
mice. Nature 2003;421:182–7.
[78] Bluher M, Kahn BB, Kahn CR. Extended longevity in mice
lacking the insulin receptor in adipose tissue. Science
2003;299:572–4.
[79] Kaestner KH, Knochel W, Martinez DE. Unified nomencla-
ture for the winged helix/forkhead transcription factors.
Genes Dev 2000;14:142–6.
[80] Birkenkamp KU, Coffer PJ. Regulation of cell survival and
proliferation by the FOXO (Forkhead box, class O) subfamily
of Forkhead transcription factors. Biochem Soc Trans
2003;31:292–7.
[81] Nemoto S, Finkel T. Redox regulation of forkhead proteins
through a p66shc-dependent signaling pathway. Science
2002;295:2450–2.
[82] Tran H, Brunet A, Grenier JM, et al. DNA repair pathway
stimulated by the forkhead transcription factor FOXO3a
through the Gadd45 protein. Science 2002;296:530–4.
[83] Kimura KD, Tissenbaum HA, Liu Y, Ruvkun G. daf-2, an
insulin receptor-like gene that regulates longevity and
diapause in Caenorhabditis elegans. Science 1997;277:
942–6.
[84] Henderson ST, Johnson TE. daf-16 integrates developmental
and environmental inputs to mediate aging in the nematode
Caenorhabditis elegans. Curr Biol 2001;11: 1975–80.
[85] Havrankova J, Roth J, Brownstein M. Insulin receptors are
widely distributed in the central nervous system of the rat.
Nature 1978;272:827–9.
[86] Wickelgren I. Tracking insulin to the mind. Science
1998;280:517–9.
[87] Woods SC, Lotter EC, McKay LD, Porte Jr D. Chronic
intracerebroventricular infusion of insulin reduces food
intake and body weight of baboons. Nature 1979;282:
503–5.
[88] Obici S, Feng Z, Karkanias G, Baskin DG, Rossetti L.
Decreasing hypothalamic insulin receptors causes hyper-
phagia and insulin resistance in rats. Nat Neurosci
2002;5:566–72.
[89] Bruning JC, Gautam D, Burks DJ, et al. Role of brain insulin
receptor in control of body weight and reproduction.
Science 2000;289:2122–5.
[90] Eaton SB, Shostak M, Konner M. The Paleolithic prescrip-
tion: a program of diet & exercise and a design for living.
New York: Harper & Row; 1988,.
[91] Cordain L. The Paleo diet: lose weight and get healthy by
eating the food you were designed to eat. New York: Wiley;
2002.
[92] Engelhart MJ, Geerlings MI, Ruitenberg A, et al. Dietary
intake of antioxidants and risk of Alzheimer disease. Jama
2002;287:3223–9.
[93] Conner WE, Anderson GJ, Lin DS. Dietary n3 fatty acid
deficiency and its reversibility. In: Mostofsky DI, Yehuda S,
Salem N, editors. Fatty Acids. Totowa: Humana Press;
2001. p. 177–92.
[94] Hasselbalch SG, Madsen PL, Hageman LP, et al. Changes in
cerebral blood flow and carbohydrate metabolism during
acute hyperketonemia. Am J Physiol 1996;270:E746–51.
[95] Reger MA, Henderson ST, Hale C, et al. Effects of beta-
hydroxybutyrate on cognition in memory-impaired adults.
Neurobiol Aging 2004;25:311–4.
699High carbohydrate diets and Alzheimer’s disease

[96] Zandi PP, Breitner JC. Do NSAIDs prevent Alzheimer’s
disease? And, if so, why? The epidemiological evidence.
Neurobiol Aging 2001;22:811–7.
[97] Gelman L, Fruchart JC, Auwerx J. An update on the
mechanisms of action of the peroxisome proliferator-
activated receptors (PPARs) and their roles in inflammation
and cancer. Cell Mol Life Sci 1999;55:932–43.
[98] Staels B, Dallongeville J, Auwerx J, Schoonjans K, Leiters-
dorf E, Fruchart JC. Mechanism of action of fibrates on
lipid and lipoprotein metabolism. Circulation 1998;98:
2088–93.
[99] Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer’s
disease: progress and problems on the road to therapeutics.
Science 2002;297:353–6.
700 Henderson