J. Wesson Ashford, M.D.,Ph.D

A-Screen (Alzheimer Screening)

Alzheimer Evaluation Forms

Alzheimer Prevention

Alzheimer Hypotheses

Dr. Ashford Lectures

Alzheimer Links

AAGP03 JWA Slides

Hypotheses  of  Alzheimer's  Disease

The Neuroplasticity Theory of Alzheimer's Disease

The most fundamental statement about Alzheimer pathology is that it attacks neuroplastic processes.  At all system levels of function (biological, psychological, social), it is the capacity to store new information that is affected by Alzheimer's disease.  Tracing memory mechanisms to their most basic levels leads to the loci at which Alzheimer pathology affects brain mechanisms.  This hypothesis was first proposed in 1985 ( Ashford & Jarvik; see Ashford, Mattson, Kumar, 1998; Teter & Ashford, 2002 for full discussion).  This hypothesis has recently been rediscovered, eloquently restated, and expanded by others (see Mesulam and Arendt, 2001). This hypothesis has been supported by repeated findings that pathological mechanisms associated with Alzheimer's disease invariably end up being related to learning mechanisms (e.g., acetylcholine, norepinephrine, serotonin pathways, NMDA receptors, synapse counts, tau phosphorylation, Amyloid PreProtein, cerebral cholesterol metabolism; see Ashford, Mattson, Kumar, 1998).  

The neuroplasticity hypothesis also pulls together the tau and amyloid hypotheses with the corollary hypothesis that there are two fundamental cellular memory mechanisms, each attacked by one of  two types of pathology, the first by the amyloid (more closely linked to causation, affecting more diffuse cortical regions including the temporal and parietal lobes), resulting in senile plaques, then, once a critical point is reached, the second by tau hyperphosphorylation, which leads to the neurofibrillary pathology (correlated with dementia severity, initially affecting the hippocampus and medial temporal lobe).  In each case, if the delicate balance between forming new connections and removing connections no longer required is disrupted, Alzheimer pathology may develop.  Amyloid PreProtein processing tips away from an alpha-secretase/beta-secretase balance, to produce excess beta-amyloid and resultant free-radicals.  Tau is excessively phosphorylated to the point that it forms neurofilaments, and then neurofibrillary tangles.  The neurofilaments appear to clog dendrite segments (Ashford et al., 1998), which leads to amputation of the distal portions of dendritic trees, large scale losses of synapses, and the increase of CSF tau.  These late changes correspond with the dementia severity associated with Alzheimer's disease (see Ashford & Schmitt, 2001 for a discussion of modeling of dementia severity).

A central factor in Alzheimer's disease is ApolipoProteinE, which is produced by glial cells and accounts for at least 50% of the Alzheimer's disease that occurs between 60 - 80 years of age.  APOE plays a central role in cerebral cholesterol transport.  Recent evidence has shown that cholesterol metabolism is involved in neuroplasticity.  Epidemiological studies are now implicating cholesterol metabolism in Alzheimer causation.  This chain of causation provides yet another buttress to support the neuroplasticity hypothesis of AD.  Additional evidence suggests that cholesterol is involved in Amyloid PreProtein processing, thus linking the APOE alleles to amyloid production, thought to be central to AD causation, and further supporting the role of this mechanism in neuroplasticity and the general neuroplasticity theory of AD.
Ashford JW, Mortimer JA. Non-familial Alzheimer's disease is mainly due to genetic factors. J Alzheimers Dis. 2002 Jun;4(3):169-77.
Raber J, Huang, Y, Ashford JW, ApoE genotype accounts for the vast majority of AD risk and AD neuropathology. Neurobiology of Aging 2004.
Ashford JW. APOE genotype effects on Alzheime's disease onset and epidemiology. Journal of Molecular Neuroscience 23:155-163, 2004.

Recent evidence supports the hypothesis that acetylcholine, a fundamental neurotransmitter in neuroplasticity, inhibits both senile plaque and neurofibrillary tangle formation (see figure adapted from Fisher, 2000).  This hypothesis suggests that drugs which increase acetylcholine function, such as cholinesterase inhibitors, may slow or stop Alzheimer progression.

Ashford, J.W. and Jarvik, K.L. Alzheimer's disease: does neuron plasticity predispose to axonal neurofibrillary degeneration? New England Journal of Medicine. 5:388-389, 1985.

Ashford, J.W., Mattson, M., Kumar, V. Neurobiological Systems Disrupted by Alzheimer's Disease and Molecular Biological Theories of Vulnerability. In: Kumar, V. and Eisdorfer, C. (Eds.) Advances in the Diagnosis and Treatment of Alzheimer's Disease. Springer Publishing Company: New York, 1998.

Ashford JW, Soultanian NS, Zhang SX, Geddes JW.  Neuropil threads are collinear with MAP2 immunostaining in neuronal dendrites of Alzheimer brain. J Neuropathol Exp Neurol 57:972-8, 1998.

Ashford, JW, Schmitt, FA. Modeling the time-course of alzheimer dementia. Curr Psychiatry Rep. 3:20-8, 2001.

Debates on Alzheimer Theories: Cincinnati, July, 2001 - conference summary,  --  review by Ashford
    - (see also summary of debate position by Ashford and Mortimer, in press, part of the conference)    


Neuroplasticity in Alzheimer’s Disease  Bruce Teter & J W Ashford, Journal of Neuroscience Research, 2002 (286K, pdf)


Arendt T. Alzheimer's disease as a disorder of mechanisms underlying structural brain self-organization. Neuroscience. 2001;102(4):723-65.

Fisher, A. Therapeutic strategies in Alzheimer's disease: M1 muscarinic agonists. Jpn J Pharmacol. 2000 Oct;84(2):101-12.

Mesulam MM. A plasticity-based theory of the pathogenesis of Alzheimer's disease. Ann N Y Acad Sci. 2000;924:42-52.

ADDENDUM:  There is an interesting question about how the brain stores information.  Information appears to be stored in a massive parallel distributed network.  When a broad neural net is activated, about half of the neurons become active.  When learning occurs, the vector weightings of the single neuron units in the net can be conceptualized as storing the new information in a "vector convolution" operation.  When recognition occurs, the mathematical operation in the network can be considered a "vector correlation".  See Ashford, Coburn, Fuster, 1998 for a complete discussion and references.  Accordingly, the storage capacity of the brain is essentially infinite.  However, learning is dependent on the ability of the neurons to be able to continually form new connections (between axons and dendrites at synapses).  If the dendrites become clogged by excess stress (particularly during learning events), as seems to occur in brains affected by Alzheimer's disease (see Ashford, Soultanian, Zhang, Geddes, 1998), then the learning is disrupted and memories are slowly destroyed, more recent ones first, old ones later.

Site constructed and maintained by:  J. Wesson Ashford, M.D., Ph.D.
    Electronic mail address: - or -

Forms and papers provided on this site may be copyrighted by their authors or the publishers.
HTML code copyright by J. Wesson Ashford, M.D., Ph.D., to retain rights for use.
Click For Statistics of Site Use               Last update:  10/13/2003