Some current theories:

It looks like the mechanisms (or combination that interact as the mechanism) that create Alzheimer's involve amyloid-beta (part of plaques) and tau (the protein that forms tangles in the cells). The tau protein serves a stabilizing function with respect to the internal organization and structure of nerve cells. If the tau proteins gets out of whack, they become threads that get twisted around one another into "neurofibrillary tangles" and then fail to help stabilize the cells. So, scientists study how the amyloid-beta is deposited as plaques and how the modification and clumping of tau into tangles takes place. There are enzymes that facilitate "cell suicide" or apoptosis that have their effect by cutting proteins. These enzymes (caspases) seem to cut the tau, and the amyloid may activate those enzymes. One could argue that these events are the results of a mechanism such as oxidation stress, and, according to a few scientists who go even further, postulate that the plaques work to guard against that stress, maybe analogous to scar tissue. More likely, there could be a kind of disease-accelerating feedback loop with the beta amyloid affecting the mitochondria in a way that increases the beta amyloid. The destruction taking place within the brain cells affects the plaque in a way that increases the destruction.
The treatment efforts to reduce or counteract beta amyloid are further along than those designed to directly reduce damage to tau proteins, and the accumulation of amyloid-beta looks to be earlier in the chain of events than the tau tangles.

Of course, that leaves the issue of what initiates the causal chain resulting in Alzheimer's. Presumably, an adequate explanation will account for the full range of disease-related changes (plaque deposition, tangles, loss of cells and pattern of cell deaths, and changes in metabolism. Considerable attention is being given to insulin.* RJ Castellani (et al**) argue that oxidative stress precedes amyloid deposits and the lesions targeted for treatment are "compensatory phenomena... manifestations of cellular adaptation". The legitimate attention being given to oxidative stress might be more widespread among serious scientists if that factor wasn't so overly hyped in the nutriceutical and supplements fields - the den of thieves, longevity fanatics, and weight-obsessed loons.

Even when the set of actions that culminate in AD has been clarified, a lot of trial and error (such as the first vaccine attempt) will be necessary to devise an active treatment. The fundamental disease-causing mechanism might be clarified backwards by treatment success using an agent that is based upon one theory. Proof about the mechanism should speed the process dramatically. However, there may be enough fundamental understanding to luck out soon with an active prevention or damage-reversing agent. In either of those cases, the demands for the earliest possible treatment and for early mass screening will expand enormously.

Some of the most interesting summaries concerning different hypotheses about the causes of Alzheimer's have resulted from formal debates. This format is very exciting.

The genetics issues are very complicated, since the rare early-onset cases of Alzheimer's appear to have much more straight forward single gene inheritance while the more frequent later-onset cases may have 4 or more genes that increase the risk of Alzheimer's. Of course, the genetics researchers believe that genetics is the only way to find out how the disease occurs and, thus, the only way to develop a real cure. No trial and error or backwards clarification here, just a long, long road.

Genetics Update, Philadelphia, July 2004: Research was reported by Brenda Plassman, Ph.D., at Duke University about rates of Alzheimer's in fraternal and identical twins. 20% by late 70s in fraternals (genetic siblings) and 40% in identicals. THis indicates a strong environmental component.

Earlier Formal research summaries, such as this one concerning genetics, offer important research specifics:

Genetic Clues to Targets to Therapies for Alzheimer's Disease
Peter St George-Hyslop - University of Toronto, Toronto, Canada
"The Latest Links: Genes and Alzheimer's Disease." National Institute on Aging
Neuroscience and Neuropsychology of Aging - November 10, 2001 San Diego, California

Genetic epidemiologic studies have indicated that about 40 percent of the population variance in risk for Alzheimer disease (AD) is attributable to genetic factors. Molecular genetic studies focusing on pedigrees with autosomal dominant inheritance have identified four genes that confer susceptibility to AD: presenilin 1 (PS1), presenilin 2 (PS2), b-amyloid precursor protein (bAPP), and apolipoprotein E (ApoE). Analysis of the biochemical effects of these genes indicates that they all influence the proteolytic processing of bAPP, causing overproduction and accumulation of the neurotoxic derivative, amyloid b-peptide (Ab), while one effect of ApoE e4 is to reduce the clearance of Ab. This commonality suggests that accumulation of neurotoxic Ab is the central event in most forms of inherited and noninherited AD. This accumulation then triggers a series of downstream events, including the misprocessing and accumulation of tau protein in neurofibrillary tangles (which is likely to be neurotoxic itself and to aggravate the problem).
Consequently, three therapeutic targets can be envisioned:

  1. blocking Ab production by inhibiting one or both enzymes involved in
    cleaving Ab from the bAPP precursor (b-secretase and g-secretase);
  2. inhibiting the aggregation of Ab monomers into toxic protofibrils; and
  3. augmenting the removal of Ab.

One such strategy is to induce an immune response to Ab by active or passive immunization. Early studies of two independent transgenic mouse models of AD reveal that immunization against Ab does reduce the learning deficit (Janus et al., Nature 408:979-982, 2000) and the neuropathologic abnormalities. This suggests that vaccination and other strategies directed at removing Ab might be effective, alone or in combination, in blocking the disease process in humans.

* Rasgon N, Jarvik L. Insulin resistance, affective disorders, and Alzheimer’s disease: review and hypothesis. J Gerontol A Biol Sci Med Sci. 2004;59A:178–183.
Steen E, Terry BM, J Rivera E, Cannon JL, Neely TR, Tavares R, Xu XJ, Wands JR, de la Monte SM. Impaired insulin and insulin-like growth factor expression and signaling mechanisms in Alzheimer's disease - is this type 3 diabetes? J Alzheimer's Dis. 2005 Feb;7(1):63-80.
de la Monte SM, Wands JR. Review of insulin and insulin-like growth factor expression, signaling, and malfunction in the central nervous system: Relevance to Alzheimer's disease. J Alzheimer's Dis. 2005 Feb;7(1):45-61.

** Neuropathology of Alzheimer disease: pathognomonic but not pathogenic. Acta Neuropathol, 2006, 111(6):503-9.