What is Alzheimer's disease?

Causes and Symptoms

In 1906, Alois Alzheimer described the pathological correlates of presenile dementia. Once considered rare, Alzheimer’s disease is now recognized as the most common form of dementia, composing 60 to 80 percent of all dementias across different age groups. In the United States, according to the Alzheimer's Association, 5.3 million people had the disease by the summer of 2015. Of those afflicted with the disease, around 5.1 million were sixty-five and older while about two hundred thousand were under the age of sixty-five. The cognitive impairments include agnosia, the loss of perceptual ability regarding the interpretation of sensory perceptions; apraxia, the inability to understand the meaning or appropriate use of things; and dysphasia, the failure to arrange words in a meaningful manner. It is a progressive neurodegenerative disorder that leads ultimately to death. While neurological and psychiatric examination provide an assessment of impairment, definitive diagnosis is arrived at only through autopsy. On the level of observable behavior, in persons affected by Alzheimer’s dementia the symptoms often develop gradually, usually after the age of sixty-five. However, the disease has also been known to develop in younger individuals and to have a more rapid onset. As such, when symptoms such as those noted develop at any age, they deserve immediate medical attention.

On a neurologic and cellular level, Alzheimer’s disease is characterized by a
triad of pathological changes in the brain including senile (or neuritic) plaques,
which consist of extracellular proteinaceous deposits surrounded by dystrophic
neurons; the presence of similar extracellular proteinaceous deposits in the brain
vasculature, termed amyloid (or congophilic) angiopathy; and the presence within
nerve cells of tangled fibrillary protein aggregates, called neurofibrillary
tangles. These pathological hallmarks are accompanied by significant neuronal loss
and brain atrophy, particularly in areas of the brain involved in memory and
cognition, such as the hippocampus and the temporal and prefrontal cortex.
Neuronal loss disproportionately affects nerve cells that use the neurotransmitter
acetylcholine (cholinergic neurons). It should be noted,
however, that the tangles and plaques have also been found postmortem in the
brains of individuals who did not have symptoms of Alzheimer’s disease, so the
association of these hallmarks is not completely understood in all
individuals.

The extracellular protein deposits in the plaques and brain vasculature have distinct optical and staining properties, suggesting that there is significant underlying organization of their constituent molecules. There are many distinct, typically rare, diseases in which proteins are deposited in this organized fashion in various parts of the body. In 1842, these deposits were called waxy degenerations or lardaceous diseases, and in 1854, they were termed amyloid (“starchlike”). The diseases are now termed amyloidoses. Examples of amyloidoses include the familial amyloidotic polyneuropathies and cardiomyopathies, senile systemic amyloidosis, lattice corneal dystrophy, the Dutch and Icelandic variants of hereditary cerebral hemorrhage with amyloid (HCHWA), and the spongiform encephalopathies, such as “mad cow disease” and kuru. In 1922, the most rigorous histological test for amyloid was devised, staining with congo red: amyloid deposits bind congo red (are congophilic) and rotate polarized light rays (are birefringent), resulting in a transition from red stain in bright-field microscopy to an apple-green coloration under polarized light microscopy. Large segments of the proteins that form amyloid deposits have a particular molecular configuration called the beta pleated sheet. These proteins precipitate from solution and aggregate to form organized structures called amyloid fibrils. The parallel alignment of the congo red dye with the organized amyloid fibril results in the optical activity recognized as birefringence.

Because the beta pleated sheet was recognized as the principal configuration of
the molecules in amyloid deposits, the primary component of the amyloid deposits
in Alzheimer’s disease was termed beta (for beta pleated sheet) amyloid protein
(βAP), or Alzheimer’s beta peptide (Aβ).

The identity of the primary protein component of Alzheimer’s amyloid was deduced
in 1987. Aβ is derived from a much larger protein, the β-amyloid precursor protein
(APP), via two proteases, β-secretase and gamma- (γ-) secretase. When added to
cultures of neurons, Aβ is toxic, and the degree of toxicity is correlated with
the degree to which the Aβ has aggregated. APP is also a substrate for a third
secretase, α-secretase, which cleaves in the middle of the Aβ sequence and thereby
precludes formation of the neurotoxic Aβ peptide. In contrast to Aβ, the
α-secretase-cleaved derivatives of APP are believed to be neurotrophic.

The identification of β-secretase has been a particularly vexing problem for
biologists. A mutation in two amino acids of APP adjacent to the β-secretase site
is known to increase β-secretase activity and thus to increase Aβ production.
However, it is a matter of debate whether this mutation
(lysine and methionine mutated to asparagine and leucine, or, in biochemical
annotation, KM to NL) is due to the increased activity of β-secretase or to an
unrelated “NLase” enzyme. Furthermore, although most Aβ is forty amino acids long,
Aβ is heterogeneous in size, ranging from thirty-nine to forty-three amino acids
in length, and the forty-two amino acid form appears to be particularly
pathogenic. Most of this variation occurs at the β-secretase site, and because of
the variations in observed forms of Aβ, there are different hypotheses regarding
the type of activity that would be expected of a β-secretase. For example, Aβ 40
and 42 might be generated by two distinct secretases. Alternatively, both Aβ 40
and Aβ 42 might be generated from the same enzyme, but the particular cleavage
site might be influenced by local factors: prior to β-secretase cleavage the Aβ
domain of APP resides in the membrane, and factors that influence membrane
thickness, such as cholesterol content, may lead to preferential cleavage at one
of the sites. Still a third possibility is that β-secretase generates a peptide
several amino acids longer than the mature Aβ 40/2 peptide and that a second
enzyme chews back on this “pre-Aβ” to yield the mature Aβ 40/2. Here, too, factors
such as membrane thickness might influence the relative abundance of the Aβ 40 and
Aβ 42 forms.

Nevertheless, β-secretase has recently been identified by several groups as the
beta-site APP cleaving enzyme (BACE), and animals whose expression of BACE has
been “knocked out” produce virtually no Aβ. Although the identity of β-secretase
has long eluded researchers, β-secretase activity is now closely associated and
sometimes identified with expression of the presenilins (see below). However,
since it has never been demonstrated that presenilins have β-secretase-like
catalytic activity, there is still debate regarding whether they are β-secretase
or a requisite component of a secretase complex. β-secretase has been identified
with tumor necrosis factor-alpha converting enzyme (TACE).

APP is a member of a family of proteins, including two amyloid-precursor-like
proteins (APLP1 and 2) for which several possible functions have been proposed,
including the formation of specific brain structures, neurite outgrowth, and
neurobehavioral development. Several mutations have been identified in APP in
association with familial Alzheimer’s disease, and these mutations are believed to
influence the rate of secretase cleavage or to alter the solubility of the Aβ
peptide. A further mutation in APP is associated with hereditary cerebral
hemorrhage with amyloid-Dutch variant (HCHWA-Dutch), a rare disorder with
Alzheimer’s-like cerebrovascular pathology. APP has been localized to chromosome
21, and those afflicted with trisomy 21 (Down
syndrome) suffer many of the same neurodegenerative hallmarks of
Alzheimer’s disease, perhaps through a gene dosage effect.

In addition to mutations in APP, mutations in other genes have been identified as
causative factors in inherited forms of Alzheimer’s disease that appear to affect
APP processing. A tremendous amount of interest has focused upon presenilin 1
(PSEN1), a gene associated with chromosome-14-linked
Alzheimer’s disease. Soon after the discovery of presenilin 1 in 1995, mutations
in another gene, originally designated as STM2, were identified as causative in
chromosome-1-linked Alzheimer’s disease. The amino acid sequences of these
proteins are remarkably similar, and STM2 is now called presenilin 2
(PSEN2). Mutations in either of these genes result in
early-onset Alzheimer’s disease, but presenilin 1 mutations result in a far more
malignant disease. Presenilin 1 mutations often result in an extraordinarily early
onset (third decade of life), and Alzheimer’s plaques are abundant in regions of
the brain, such as the cerebellum, that are unaffected in the sporadic disease.
This results in motor signs, such as myoclonus and seizure, which are absent in
the sporadic disease. Mutations in both presenilin 1 and presenilin 2 result in
increased production of Aβ, especially of a slightly larger and more pathogenic
peptide, Aβ 42. Because these mutations appear to primarily influence the length
of Aβ, their activity has primarily been associated with γ-secretase activity.

Initially, the mechanisms by which the presenilin mutations might result in
Alzheimer’s disease were unclear. By coincidence, researchers in the field of
programmed cell death, or apoptosis, found a gene they identified
as alpha-1,3- mannosyltransferase (ALG3), which they believed
rescued cells from programmed cell death. ALG3 had significant
identity with a portion of presenilin 2, and the researchers speculated that
mutations in the presenilins might lead nerve cells to an aberrant entry into
apoptosis, resulting in the neuronal cell atrophy and death observed in
Alzheimer’s disease. However, there is considerable debate about whether
necrosis, as opposed to apoptosis, is the primary means of
cell death in Alzheimer’s disease, and the role of the presenilins in apoptosis
remains controversial.

The presenilins had no similarity to any known mammalian proteins but had limited
identity with two proteins found in the nematode worm Caenorhabditis
elegans. One protein with very limited identity is spe-4, a protein
involved in spermatogenesis in the worm. The other protein with more significant
identity (43 percent) is sel-12, a protein involved in the signaling of Notch, a
protein involved in many developmental processes. Attempts to knock out the
expression of presenilin 1 result in severe defects that are lethal to the
late-stage embryo and that
closely resemble the defects obtained with knockout of Notch expression.
Conversely, expression of the normal human presenilin in worm cells is defective
for sel-12 expression and restores Notch signaling, suggesting that both
γ-secretase activity and Notch expression converge in the activity of
presenilin.

Notch signaling requires the proteolysis of a membrane-bound precursor in a manner
that is very similar to the γ-secretase cleavage of APP. While the presenilins do
not resemble any known proteases, a certain class of proteases has an obligate
amino acid residue which, when mutated in presenilin, eliminates both γ-secretase
and Notch cleavage. Also, drugs that are believed to be γ-secretase inhibitors
inhibit both γ-secretase cleavage and Notch cleavage. Thus, the presenilins appear
to be intimately involved with the proteolysis of APP and Notch. Although Notch
activity does not appear to be directly related to Alzheimer’s pathology, there is
a concern that drugs designed to reduce Aβ production by inhibiting γ-secretase
activity might also have an impact on normal Notch signaling, thereby adversely
influencing such things as blood cell maturation.

The presenilins are synthesized by all cells in the body, and they appear to be localized within the cell primarily if not exclusively in the endoplasmic reticulum/golgi apparatus. While Aβ was believed to be derived primarily from the cell-membrane-associated APP through a specific pathway (the endosomal/lysosomal pathway), the localization of the presenilins and their association with γ-secretase activity to the endoplasmic reticulum led to a reevaluation of the cellular site of Aβ generation. There is a consensus now that most Aβ, particularly the more pathogenic Aβ 42, is produced within the cells at the endoplasmic reticulum/golgi apparatus. These results are significant because they suggest that, rather than forming from extracellular circulating Aβ proteins, amyloid plaques may begin as aggregates of Aβ within the cell. Those aggregates may, in turn, lead to neuronal injury and death.

Although the presenilins are intimately associated with γ-secretase activity,
there has not yet been any definitive proof that the presenilins are involved in
proteolytic activity. One would expect, for example, that if the portion of human
APP cleaved by γ-secretase and presenilin were coexpressed in a cell that has no
endogenous γ-secretase activity (such as a yeast cell), that presenilin would
cleave APP, yet this does not occur, and controversy remains over whether
enzymatic activity resides with the presenilins or whether the presenilins are an
obligate component of multiprotein complexes. Recently, a protein that forms a
complex with presenilin and that binds APP was identified. This protein,
Nicastrin, named for the Italian village of Nicastro, where key early studies on
familial Alzheimer’s disease took place, also appears to play an important role in
γ-secretase activity and in Notch processing. It is possible that there are other
key components to this γ-secretase complex that are yet to be elucidated. A number
of additional proteins that interact with presenilin or with APP have been
identified; their role in Alzheimer’s disease pathogenesis is being investigated.
These include calsenilin, Fe65, X11, and BBP-1 (β-amyloid binding protein-1).

Aβ is, therefore, a neurotoxic peptide and a component of two of the predominant
features of Alzheimer’s disease. Since senile plaques occur within specific
regions of the brain, while sparing other regions (especially the cerebellum),
investigators originally believed that this distribution of pathology was due to
regional differences in the production of APP. However, it soon became apparent
that APP was synthesized by virtually all cells and that Aβ is present in
abundance throughout the brain and indeed in other tissues of individuals not
affected by dementia. How, then, does a normal biological molecule become a
pathological agent in the aging brain to cause nonfamilial (sporadic) cases of
Alzheimer’s disease?

It has been proposed that multiple genetic and environmental factors may play a
role in this transformation. Some investigators have proposed that one or more of
the minor components of amyloid plaques may act as “pathological chaperone”
molecules by binding Aβ and altering its solubility. The identification of such
components might lead to new drug research strategies and to new insight into the
mechanisms of amyloid formation. For example, heparin sulfated proteoglycans
(HSPGs) have been identified as components of amyloid deposits, and some
investigators have attempted to block amyloid formation by inhibiting the
interaction of Aβ with HSPGs. Alpha-synuclein was identified as the nonamyloid
component of Alzheimer’s disease plaques (NAC peptide); interestingly, mutations
in α-synuclein have been identified in familial forms of Parkinson’s
disease, and this protein is a major constituent of the
inclusion bodies found in Parkinson’s disease–associated Lewy body dementia.
Aluminum has been identified as a component of the senile plaques found in
Alzheimer’s disease, and aluminum toxicity results in the formation of
neurofibrillary tangles. It has been proposed that aluminum exposure may be a
predisposing factor, but this hypothesis is very controversial. Similarly, zinc
has been implicated in the formation of Aβ plaques, and a drug that chelates
(absorbs) zinc appears to reduce the abundance of plaques in transgenic animal
models of Alzheimer’s disease.

Soluble Aβ may circulate in the serum and cerebral spinal fluid bound to one of
several chaperone proteins, including apolipoprotein E, apolipoprotein J, and
transthyretin. Polymorphisms in these proteins might alter the affinity of these
proteins for Aβ, thereby leaving more free Aβ available to aggregate into amyloid
fibrils. A particular isoform of apolipoprotein E (ApoE4), for example, has an
altered affinity for Aβ, and the inheritance of this isoform is associated with
increased risk for Alzheimer’s disease.

Genes that may play roles in the degradation of proteins may also be involved in
the development of Alzheimer’s disease pathology. For example, polymorphisms in
the cystatin C (CST3) gene may alter the susceptibility to
late-onset Alzheimer’s. Cystatin C is a protease inhibitor, and mutant cystatin
C is an amyloid component in the Icelandic variant of hereditary cerebral
hemorrhage with amyloid (HCHWA-I). Also, the normal process by which cells target
proteins for disposal may be altered in Alzheimer’s disease. Cells normally target
proteins for degradation by conjugating them to the protein ubiquitin, and there
are reports that the ubiquitination pathway is altered in brains afflicted with
Alzheimer’s disease.

Neuritic plaques and less organized forms of plaques (diffuse plaques) are not unique to Alzheimer’s disease but are also found in dementia pugilistica, as a sequelae to severe trauma, are found in Down syndrome, and are coexistent with other diseases, such as Parkinson’s disease. Likewise, as already noted, the amyloid (congophilic) angiopathy is also found in other familial amyloidoses. It is the combination of these cardinal findings with the finding of neurofibrillary tangles that forms the triad of primary pathological signs of Alzheimer’s disease.

Neurofibrillary tangles are intracellular aggregates of an unusually modified (hyperphosphorylated) form of the structural protein tau (τ). Neurofibrillary tangles are not unique to Alzheimer’s disease but occur in several other neurodegenerative diseases, such as amyotrophic lateral sclerosis (ALS)–Parkinsonism dementia complex as a consequence of measles infections of the central nervous system (subacute sclerosing panencephalitis), in the rare spongiform encephalopathies (including Creutzfeldt-Jakob disease and kuru), and in frontotemporal dementia. The number of neurofibrillary tangles correlates well with the severity of impairment.

There has been a long-standing debate regarding the significance of the
pathological findings in Alzheimer’s disease. The debate has focused on whether
Aβ-associated pathology or the tau-protein-associated pathology is the primary
lesion in the disease, thus dividing investigators jocularly into "βaptist" and
"τaoist" camps.

There has been a general consensus that Aβ plays a central causative role in the
disease. According to this “amyloid hypothesis,” the excessive production, rapid
deposition, or aberrant metabolism of Aβ results in the formation of toxic
aggregates of Aβ, which in turn result in injury to neurons, neuronal death, and
cognitive impairment. The tau pathology manifested as neurofibrillary tangles is a
consequence of Aβ effects. Consistent with this view is the fact that all genes
that have thus far been associated with familial Alzheimer’s disease play a role
in Aβ processing or solubility.

However, others would argue that amyloid plaques in Alzheimer’s disease are an epiphenomenon, that they are a result of the neurodegenerative process rather than its cause. According to this “tau hypothesis,” neurofibrillary tangles are the central pathological finding in the disease, and Aβ plaques are a relatively inert consequence of normal aging or of neurofibrillary tangle-associated neuronal degeneration. Consistent with this viewpoint is the strong correlation between cognitive impairment and neurofibrillary tangle burden. In contrast, amyloid plaques may be found in the brains of individuals who are not affected by dementia, and total plaque burden (the number of plaques found in a given brain section) does not correlate well with impairment. Indeed, it has been difficult to demonstrate significant impairment in transgenic animal models of Alzheimer’s disease (animals that overproduce human APP in the brain), even in the presence of significant amyloid deposits in the brain. It is noteworthy that those animals fail to develop neurofibrillary tangles. However, there have been no direct links between the genetics of Alzheimer’s disease and tau protein, although a mutation has been identified in tau associated with a non-Alzheimer’s dementia (frontotemporal dementia).

There has been a significant convergence of opinion that Aβ is central to
Alzheimer’s disease pathogenesis. This is due to reports that total brain amyloid
protein burden (a measure of the total Aβ protein, including protein from plaques)
and not plaque number correlates well with severity of dementia.

Treatment and Therapy

There are presently no therapies available for Alzheimer’s disease that target the
putative underlying mechanisms for the genesis and progression of the disease,
although there are extraordinary efforts under way throughout the world to address
these mechanisms. Treatment options at present are primarily palliative. This
means that they address the symptoms, perhaps providing some temporary relief, but
not providing any cure.

Acetylcholinesterase inhibitors such as donepezil (Aricept), rivastigmine (Exelon), galantamine (Razadyne), and tacrine (Cognex) have shown efficacy in terms of temporarily slowing progression of the disease. These agents all inhibit an enzyme (acetylcholinesterase) that degrades the neurotransmitter acetylcholine. Memantine (Namenda, Ebixa, and Axura), an N-methyl-D-aspartate (NMDA) receptor antagonist, may have a small benefit in individuals with moderate-to-severe Alzheimer's disease. Clinical trials using an agent that mimics the actions of acetylcholine (nicotine) ended abruptly due to side effects that included severe anxiety.

Because the amyloid lesions in Alzheimer’s disease may elicit a limited
inflammatory response in the brain, which in turn may be a major component of the
neurotoxic effects of Aβ, nonsteroidal anti-inflammatory drugs
(NSAIDs), such as aspirin and ibuprofen, have been proposed as agents having
possible ameliorative effects.

Certain retrospective studies have shown a negative correlation between the use of
estrogen replacement therapy (ERT) and both the age of onset and severity of
Alzheimer’s disease in postmenopausal women, suggesting that ERT may have a
protective function in these subjects.

Cholesterol may influence the production of Aβ by altering the characteristics of cell membranes, and Aβ is transported by lipoproteins. Thus, lipid metabolism may have an impact on the development of Alzheimer’s disease.

Perspective and Prospects

According to the US Centers of Disease Control and Prevention, as of 2013 Alzheimer’s disease was the sixth leading cause of death among Americans. In 2014, the annual cost of Alzheimer's reached $214 billion in the United States, making Alzheimer's the most expensive condition in country, according to the Alzheimer's Association. The disease is projected to affect between eleven million and sixteen million individuals in the United States by 2050.

The pharmaceutical manufacturer Lilly patented the first orally active β-secretase
inhibitor, and many such drugs are currently in development. It is possible that
some of these drugs may suffer liabilities due to alterations of Notch signaling.
Nevertheless, they will provide key information regarding the validity of the
amyloid hypothesis. With the recent cloning of β-secretase, drug discovery efforts
surrounding β-secretase will likely be greatly facilitated.

Biopharmaceutical research has also focused on targeting γ-secretase for
Alzheimer's disease; however, these efforts are complicated by the fact that
attempts to control γ-secretase activity affect other physiologically critical
protein substrates mediated by γ-secretase.

Another potential target for drug intervention in Alzheimer’s disease is the
inhibition of Aβ aggregation. Although a number of agents have been reported that
inhibit aggregation, they are all characteristically large molecular weight
peptides that have limited access across the blood/brain barrier. Thus, although
this approach may work in principle, the development of a useful drug from this
approach may be severely limited by issues regarding access of the drug to the
brain.

Neurofibrillary tangles are an additional feature of Alzheimer’s disease that may serve as a target for drug development. Since neurofibrillary tangles are composed of hyperphosphorylated tau protein, inhibitors of tau phosphorylation might be developed as useful therapeutic entities.

It has been reported that experimental vaccination of transgenic animals showing
features of Alzheimer’s disease such as amyloid plaques with Aβ peptide results in
the reduction of amyloid burden. The exciting implications of these experiments
are that Alzheimer’s disease might one day be treated or even prevented by
vaccination. A vaccine called CAD106 targets beta amyloid, a protein fragment that
forms the amyloid plaques, a significant biomarker of Alzheimer's disease.

Other approaches to therapeutics for Alzheimer’s disease include the development
of neurotrophic factors that may enhance neuron survival. Research into
stem
cells has provided evidence for the presence of neuronal stem
cells within the brain. Research into the mechanisms of stem cell migration and
development might yield drugs that would enhance the recruitment of such primitive
stem cells into dystrophic areas of the brain.

Because early detection of Alzheimer’s disease increases the effectiveness of
current drugs and treatments and helps keep some of the disease’s more devastating
symptoms at bay, research is focused on ways to detect the disease definitively. A
Canadian study released in 2002 notes possible connections between scores of
verbal memory tests and the likelihood of developing the disease. The study
examined patients’ performance on a variety of psychological tests to gauge the
tests’ reliability in detecting preclinical Alzheimer’s disease. The researchers
discovered that verbal memory tests—for example, recalling the categories of words
or being able to remember terms for a short period of time—were highly accurate in
determining which individuals went on to develop the disease. These findings seem
to support what other researchers have found: the evident decline in verbal memory
in elderly individuals in the one- to two-year period prior to the development of
Alzheimer’s disease symptoms.

The search for causes of Alzheimer’s dementia remains a topic of interest in the effort to develop prevention strategies. For instance, recent research has demonstrated that isoflurane, a common anesthetic, can lead to the development of amyloid protein in cultured neuronal cells and to cell death. Research following this discovery may lead to recommendations to avoid the use of such anesthetics, in favor of others, as a way of preventing exposure of vulnerable persons to what may be a causative factor in Alzheimer’s dementia.

In 2013, scientists at the National Institutes of Health discovered that blocking the activity of regulator protein CD33 may stem the disease process of Alzheimer's.

In 2015, as part of the Alzheimer's Association International Conference in Canada, a panel presented research on biomarkers that had been conducted to determine whether analyzing chemical compounds in saliva and cerebrospinal fluid could help doctors to make earlier diagnoses of Alzheimer's. That same year, the National Institutes of Health approved millions of dollars in funding for the creation of an Alzheimer's Disease Research Center at Stanford's University School of Meidicine in California.

Bibliography

Brioni, Jorge D., and
Michael W. Decker, eds. Pharmacological Treatment of Alzheimer’s
Disease: Molecular and Neurobiological Foundations. New York:
Wiley 1997. Print.

Dickerson, Bradford, and Alireza Atri,
eds. Dementia: Comprehensive Principles and Practices.
Oxford: Oxford UP, 2014. Print.

Dysken, M. W. "Effect of Vitamin E and
Memantine on Functional Decline in Alzheimer Disease: The TEAM-AD VA
Cooperative Randomized Trial." JAMA 311.1 (2014): 33–44.
Print.

Esiri, Margaret M.,
Virginia M.-Y. Lee, and John Q. Trojanowski, eds. The Neuropathology
of Dementia. 2nd ed. New York: Cambridge UP, 2004.
Print.

Gauthier, Serge, ed.
Clinical Diagnosis and Management of Alzheimer’s
Disease. 3rd ed. Boca Raton: Taylor, 2006. Print.

Goldman, Bruce. "Stanford-Based Alzheimer's Disease Research Center to Be Launched." Stanford Medicine. Stanford Medicine, 6 July 2015. Web. 5 Aug. 2015.

Kunkle, Frederick. "Saliva Seen as Possible Diagnostic Tool for Alheimer's Disease." Washington Post. Washington Post, 19 July 2015. Web. 5 Aug. 2015.

Mace, Nancy L., and
Peter V. Rabins. The Thirty-Six-Hour Day: A Family Guide to Caring
for Persons with Alzheimer Disease, Related Dementias, and Memory
Loss. 5th ed. Baltimore: Johns Hopkins UP, 2012.
Print.

Ropper, Allan H., Martin A. Samuels, and
Joshua P. Klein. Adams and Victor's Principles of
Neurology. 10th ed. New York: McGraw-Hill, 2014. Print.

Rowland, Lewis P., ed.
Merritt’s Textbook of Neurology. 12th ed. Philadelphia:
Lippincott, 2010. Print.

"Suppressing Protein May
Stem Alzheimer’s Disease Process." NIH.gov. US Natl. Inst.
of Health, 25 Apr. 2013. Web. 21 May 2013.

Sutton, Amy L., ed.
Alzheimer’s Disease Sourcebook. 5th ed. Detroit:
Omnigraphics, 2011. Print.

"2015 Alzheimer's Disease Facts and Figures." Alz.org. Alzheimer's Assoc., 2015. Web. 5 Aug. 2015.

Wasco, Wilma, and
Rudolph E. Tanzi, eds. Molecular Mechanisms of Dementia.
Totowa: Humana, 1997. Print.

Whitehouse, Peter J.,
Konrad Maurer, and Jesse F. Ballenger, eds. Concepts of Alzheimer
Disease: Biological, Clinical, and Cultural Perspectives.
Baltimore: Johns Hopkins UP, 2003. Print.

Winblad, Bengt, et al. "Safety, Tolerability,
and Antibody Response of Active Aβ Immunotherapy with CAD106 in Patients
with Alzheimer's Disease: Randomised, Double-Blind, Placebo-Controlled,
First-in-Human Study." Lancet Neurology 11.7 (2012):
597–604. Print.