Time to Dismount

Dr. de la Torre speaks the painful truth about the current state of Alzheimer’s and where future efforts must be directed. Our bodies are, of course, connected systems – as we have one vascular, nervous, and lymphatic system, thus the vascular/neurodegenerative connection must take center stage. To adopt a vascular view of dementia/Alzheimer’s one must also adopt an emerging view of vascular disease, that is inflammation and, more importantly, its antecedents. For those with open minds I direct you to the work of McCully [1], Balin [2] and Miklossy [3]. When the educated reader delves into the works of these experts, an interesting nexus between vascular and neurodegenerative diseases emerges. And the work of Dr. Tanzi at Harvard connects all this research to beta-amyloid plaques in a quite unexpected way [4].

Claude Bernard, arguably the top clinician/research of all time states, “The researcher who does not know what he/she is looking for will not understand what he/she finds."

Stay Well,
Thomas J. Lewis, Ph.D., CEO
RealHealth Clinics

References
[1] Ravnskov U, McCully KS (2009) Review and Hypothesis: Vulnerable plaque formation from obstruction of Vasa vasorum by homocysteinylated and oxidized lipoprotein aggregates complexed with microbial remnants and LDL autoantibodies. Ann Clin Lab Sci 39, 3-16.
[2] Hammond CJ, Hallock LR, Howanski RJ, Appelt DM, Little CS, Balin BJ (2010) Immunohistological detection of Chlamydia pneumoniae in the Alzheimer's disease brain. BMC Neurosci 11, 121.
[3] Miklossy J (2011) Alzheimer's disease - a neurospirochetosis. Analysis of the evidence following Koch's and Hill's criteria. J Neuroinflammation 8, 90.
[4] Soscia SJ, Kirby JE, Washicosky KJ, Tucker SM, Ingelsson M, Hyman B, Burton MA, Goldstein LE, Duong S, Tanzi RE, Moir RD (2010) The Alzheimer's disease-associated amyloid beta-protein is an antimicrobial peptide. PLoS One 5, e9505.

Dr. Jack de la Torre is right in its assertions. Scientific objectivity demands that we abandon the failed amyloid beta cascade (ABC) hypothesis as a causative explanation and as a therapeutic target. It is important to recognize that while ABC may play a role in a small minority of early-onset AD cases linked to genes, the overwhelming majority of AD cases are late-onset cases of the sporadic type. It is in this majority of aging-related cases that ABC is truly “a dead horse” that cannot take us anywhere. Perhaps one first step is to stop calling all these dissimilar early and late onset cases as one disease. Then we can abandon the futile idea that all forms of AD are caused by the same ABC, and focus on what is actually causing the majority of these late-onset AD cases. The older dementia nomenclature was more helpful to make this distinction between early and late onset cases. The early-onset cases were called pre-senile dementia, while the late-onset cases were called senile dementia. They are clearly not the same disease. The common aging-related dementia is more linked to impairment of oxidative energy metabolism in mitochondria, which is in turn linked to vascular risk factors and other metabolic insults to the aging brain. The investigation of aging-related vascular and energetic metabolic factors is a better “new horse” to ride in search for early interventions to this devastating form of senile dementia.

Dr. Jack de la Torre’s concern is that over 100,000 journal articles about Alzheimer’s disease (AD) have failed to provide any substantial help to those with the affliction. Some investigators have suggested that it has been a mistake to pursue so long and so doggedly the amyloid cascade theory of Hardy and Higgins. Perhaps there is some other underlying pathology responsible.

The opposite mistake is ignoring a well confirmed finding. Many other journal papers have overlooked the significance of a link between AD and normal vascular changes in the brain during old age. Here I’m referring to the score of reports (23) showing that cerebral capillary density wanes in those with AD, in aging persons in general, and also in older rabbits, rats, and dogs. This decline occurs not only in the brain but also in muscles (13) and elsewhere in the aging body (kidney & leukocytes)—thus indicating a universal, genetically driven decrease in angiogenesis, which determines capillary density. An age-linked general waning of the microcirculation may account for the many signs and symptoms of older persons, such as muscle weakness, cold intolerance, minor memory lapses, later senile dementia, and possibly AD. Thirty-eight reports of depressed levels of angiogenic factors and the resulting reduced capillary density in aged persons and old animals have been abstracted and summarized in my four journal articles [1-4].

In 1993, Dr. de la Torre advanced the view that “AD is a vascular disorder with neurodegenerative consequences” [5]. He has published widely on relevant risk factors, which include physical inactivity, obesity, smoking, etc. [6]. These factors reflect life style decisions, which lead over many decades to hyperlipidemia, hypertension, diabetes, atherosclerosis, carotid artery pathology, etc. These conditions involve primarily arteries and arterioles, i.e., the ‘macro’ vascular system.

In contrast, my focus has been on the microcirculation, i.e., the ‘micro’ vascular bed. A genetically determined, age-linked decline in capillary density throughout the body is the basis of the angiogenesis hypothesis, which proposes that many of the hallmarks of old age (listed above) are due to a deficiency condition, i.e., declining levels of angiogenic factors, much like declining levels of various hormones with age.

While preventive measures such as a healthy lifestyle can reduce the risk factors or their impact on the body, little can be done to reverse the life-long accumulated damage to the walls of most of the macro vascular system. To date, atherosclerotic pathology cannot be corrected. In contrast, a reduced capillary density (the micro vascular bed) may be corrected (at least in part) by administration of recombinant angiogenic factors, as demonstrated in animal studies. Again, pathologic changes in the walls of larger vessels (arterioles) cannot readily be reversed, but the number of small vessels (capillaries) can be increased with recombinant angiogenic factors.

The angiogenesis hypothesis does not exclude other factors contributing to AD (e.g., amyloid deposits, neurofibrillary tangles, etc.) or ignore pathology in the larger vascular system and even changes in capillaries (e.g. kinking, etc.). But the hypothesis proposes that an important underlying pathology is the diminished number of capillaries naturally occurring in the aged.

Michael de Montaigne, 1533-1592, wrote about new discoveries, which here could include new hypotheses. Whenever a new one is reported to the scientific world, the members say first, “It is probably not true.” Later when the validity of the proposition has been demonstrated, they say, “Yet, I may be true, but it is not important.” Finally, when sufficient time has elapsed, they say, “Yes, surely the discovery is important, but it is no longer new and has been known all along.”

I believe that the angiogenesis hypothesis is an important new concept for geriatric medicine and warrants the widest possible dissemination in the relevant literature to invite its scrutiny and, if substantiated, to elicit its exploitation with therapeutic angiogenesis.

References
[1] Ambrose CT (2013) Alzheimer’s disease: The great morbidity of the 21st century. Am Sci 101, 194-274.
[2] Ambrose CT (2012) Neuroangiogenesis: a vascular basis for Alzheimer's disease and cognitive decline during aging. J Alzheimers Dis 32, 773-788.
[3] Ambrose CT (2015) A therapeutic approach for senile dementias: neuroangiogenesis. J Alzheimers Dis 43, 1-17.
[4] Ambrose C (2015) Muscle weakness during aging: a deficiency state involving declining angiogenesis. Ageing Res Rev 23, 139–153.
[5] de la Torre JC, Mussivand T (1993) Can disturbed brain microcirculation cause Alzheimer's disease? Neurol Res 15, 146-153.
[6] de la Torre JC (2010) Vascular risk factor detection and control may prevent Alzheimer's disease. Ageing Res Rev 9, 218-225.

Determining cause and effect is the essence and the bane of the life of science. But the problem with our manipulative sciences is that you can just about will any result. In the case of AD, expressing excessive levels of amyloid in the mouse models, you will inevitably force a change, behaviorally and pathologically. Toss enough amyloid in a dish of cells, and yes, you will force a change. I remember the gene-jocks inserting the human protease receptor in HEK293 cells to assess compound activity in their high throughput assays. They were having issues trying to distinguish calcium signaling between the transfected and parental cell lines. Turns out, I showed them that the HEK293 cells expressed the receptor without the transfection; furthermore, the inserted receptor failed to engage in its secondary messaging molecules. This impulse to jump into artificial systems may have helped move some compounds through the pipeline, but in the AD field, it has taken us off course, and now we’re so ‘committed’ to forcing our will on a single amyloid deposition hypothesis, that we seemed to have lost perspective.

From the very first IHC slide I observed under the microscope, I could not reconcile the amyloid hypothesis from the simple unbiased observation of intraneuronal Aβ42, and the lack of the apparently toxic Aβ42 piling along the neuronal membranes. How did the field go so off course that extracellular plaques arise from neuronal deposition? Seeing intraneuronal Aβ is not an original observation because neurons make it, but this aged-observation placed intraneuronal amyloid as the source of extracellular deposition; however, intraneuronal Aβ has natural functions in the processes of learning and memory. Can you possibly imagine neurons making excessive amounts of amyloid, pushing it out into the synaptic spaces that deposited to from plaques (diffuse and then neuritic/dense-core) to then kill themselves? Does that sound logical? Just about every AD neuron I’ve observed (perhaps tens of thousands) does not have a decoration of amyloid anywhere on its cell body cell membrane.

I, like Jack, subscribe to the understanding that vascular issues are truly the first domino in the pathological sequences. I extend this model linking the cerebrovascular issues to a dysfunctional BBB. All of the cells of the vascular system, including the endothelial cells, smooth muscle cells and pericytes are degenerating, and all seems to have ‘amyloid’ on their faces.

Since amyloid, specifically Aβ42, is normally present in very low concentrations in the brain as evident by the exceptionally high binding affinities to several neuronal receptors, it is logical to characterize Aβ42 as the heavily-regulated currency in the brain. Any changes in the BBB, such as endothelial damage, can lead to unregulated excessive amounts of Aβ into the brain (including smooth muscle cells and pericytes) that overwhelms the neuronal receptors to become engorged with amyloid to die, which then kicks off another pathological spiral of cell death independent of amyloid and deposition. Seems logical, even back in 1999, but now that clinical proof has caught up to us disproving the amyloid hypothesis in its original and modified form, time to smell the roses and move on.

The harsh reality is that AD is not reversible, is not a ‘deposition-based’ disease, and the only hope for AD patients is us to prevent further decline. Prevention can only occur before the clinical presentation of AD and a clear understanding of the pathological processes leading to AD on where to target, which at this time continues to be the vascular system.

But who are ‘they’ anyhow, is it you? Who makes that call to ‘move-on’, to ‘dismount’? Is it the few VPs, or the village of scientists? If the latter, well, it didn’t work the first time, so I can’t see the village making the change, as Jack noted, thousands of papers, and with no hope. Is it the bias in the funding agencies to self-serve apparent KOLs to ‘keep the lights/and egos on’ on dated hypotheses? Maybe, as no other hypotheses seemed to have risen to the recognition of funding, publishing, and clinical trials as the amyloid deposition hypothesis. If the VPs are those to make that change, how can they change course without acknowledging failure? Perhaps it’s the laymen pharma board members that can make change. However, chances are that they may be spoon-fed entertaining snippets of data to show progress. Hard to say, but like much of our capitalist society, if you fail, others will fill the void, and the field needs to fill the void by “others”.

Portions of this blog are illustrated on YouTube: https://www.youtube.com/watch?v=_NTaGjQow1c

JL de la Torre’s thoughtful piece raises an issue scientific freedom, and another of the ethics of knowledge, issues that deserve to be addressed.

Scientists can be stubborn in our views, when evidence gathers against an idea that we favour. Philosophers have attributed that stubbornness to respect for an idea we have grown up with; to fear of conceptual vacuum, should a prevailing understanding be dislodged; to fear of being in error should an idea in which we have invested fail; or to mere partiality, a fondness for what seems like a neat idea, often our own.

Whatever its source, this stubbornness is not new. The physicist Max Planck wrote in his scientific autobiography (1948): A new scientific truth does not become accepted by its opponents declaring themselves convinced and informed, but because the opponents gradually die out, and the next generation grows up familiar with the new truth, from the start [1].

de la Torre expresses a similar impatience, at the continued adherence of many scientists studying dementia to the amyloid cascade hypothesis of its pathogenesis, an adherence that requires them to ignore mounting evidence that the cause of the dementia is vascular.

So the question of scientific freedom arises: when only some of us prefer a new idea with its new evidence, what if anything is to be done? We can plead our case, of course and design new tests of the competing ideas. But historians of science have long observed that there is no experiment conclusive to any controversy, no new idea that is irresistible. The shift from old idea to new can seem painfully slow, especially to those convinced of the new.

The question whether Alzheimer's disease is a vascular dementia or a proteinopathy, or even, as Robert Whitehouse has argued thoughtfully in these blogs, whether there is a disease entity that deserves so famous a name – this cannot be settled by an authority or by a learned society, or by a popular vote. Any attempt would breach the freedom of the individual to disagree.

Patience is then required. Those of us who are young enough must settle down to watch the process unfold; those of us who are older may suffer to die wondering. The only proper action is to do as de la Torre has done, to plead the case and go back to the laboratory.

That said – and this is the second issue, the ethical issue raised by de la Torre’s piece - our freedom to think as we will is nevertheless constrained, this writer believes, by at least two issues that arise outside science. Most importantly, people are suffering, and many now cognitively normal are destined to suffer, the tragedy of dementia; the sooner we understand it the sooner there will be done whatever can be done. Those who resist new ideas against the evidence may be condemning a generation or two to a personal suffering they might have avoided, had scientists been less conservative. It is a moral challenge; human suffering demands that we all think clearly, suppress our fears and put aside our partialities. That suffering constrains our scientific freedom.

Second, the taxpayers of research-committed nations, by electing governments that support research, become part a social compact that commits part of their taxes to research. They do so for the large part willingly, trusting the scientists and our grant review committees and research councils. The taxpayers have a reasonable expectation that the funds will be spent on the ideas best based in evidence. The scientists’ part in that compact is to spend those funds – other people’s money – wisely, to think clearly, suppress our fears and put aside our partialities, even if our favourite theory may fall by the scientific wayside.

For the record, I think that de la Torre is correct; that the dementia we call Alzheimer's disease is a small-vessel vascular dementia; that it is past time to let the proteinopathy ideas go. With others I have proposed that the dementia is driven by vascular ageing [2], specifically by the impact of the pulse, made increasingly intense by aortic stiffening, on the cerebral vasculature, producing the small vessel pathology that several researchers, de la Torre prominent among them, have long suspected to be the cause of the dementia. If we are right, then understanding is finally at hand of the link between age and dementia (in the hardening of the aorta), and a previously unrecognised limit to longevity (how long the brain’s vasculature can withstand the pummeling it receives from the pulse) has been recognised. It is also an idea rich with therapeutic options; but, properly, the idea will have to battle it out in the field.

I may be too old ever to learn how the idea will fare; the process of assessment will likely outlive me - the fate of those of us who propose new ideas late in our allotted years. But I hope more than trust that its assessment will not be delayed by the fears and partiality of others. We all have a duty – a non-scientific duty – to understand our own frailties, and deal rigorously with the evidence before us.

How goes the process so far? Well, it’s early days but, so far, I fear that Planck understood this aspect of human nature very accurately.

References
[1] Eine neue wissenschaftliche Wahrheit pflegt sich nicht in der Weise durchzusetzen, dass ihre Gegner überzeugt werden and sich als belehrt erklären, sondern vielmehr dadurch, dass die Gegner allmählich aussterben and dass die heranwachsende Generation von vornherein mit der Wahrheit vertraut gemacht ist. Planck, M. Wissenschaftliche Selbstbiographie. 1948 Johann Ambrosius Barth Verlag, Leipzig; p. 22. His impatience is sometimes expressed in a briefer translation: Science does progress, one grave at a time.
[2] Stone J, Johnstone DM, Mitrofanis J, O'Rourke M (2015) The mechanical cause of age-related dementia (Alzheimer's disease): the brain is destroyed by the pulse. J Alzheimers Dis 44, 355-373.

Jack de la Torre's post "Time to dismount" should be used by as many as possible health practitioners in the field of Alzheimer's disease as an attention grabber and hopefully consciousness awakening.

The name "Alzheimer's disease" itself is misleading, non only for patients, but for young doctors and for the population in general: fortunately we can keep the acronym AD convincing ourselves that we are speaking about "Alzheimer's dementia", which might be a better name than "disease" for a syndrome. What we keep doing as AD doctors and teachers is to tell every day to our patients and students that the cause of the disease is not completely known, that a causal treatment does not exist and yet we spend hours in explaining complicated figures of pathophysiology mechanisms and good reasons for participating in drug clinical trials. What if a cure for AD did not exist because there is no single mechanism? If this question is a point of agreement for readers of the field, why then not look at the possibility that more mechanisms might exists at the same time? The logical consequences of such a theory would be a) that AD is not a disease but a syndrome in the original ethimologic sense, i.e. of more factors "running" (dromos) together (syn); b) that different peripheral mechanisms probably share a common pathology; c) that there is a striking possibility that tens, if not hundreds of studies have been conducted already, have lead evidence converging to a common AD pathology factor but comprehensive attention has not been given to the fact; d) that no drug treatment is to be found; and that e) AD as a syndrome or process is mostly resembling two other AD-related conditions, i.e., aging and redox imbalance - I still like to name the latter oxidative stress in the 1993 revised meaning by its pioneer and definer Helmut Sies.

We have to face that these consequences have enormous implications which are the reasons why health practitioners don't dismount from the "single molecular mechanism" horse. First of all, one needs to describe, organize the evidence, find a consensus, produce guidelines, and counseling. After the evidence, single molecule clinical trials would have to be stopped. AD MDs and researchers like me dismounted for instance from the single antioxidant trial horse long time ago, but in a recent review, I listed the tens of still ongoing trials with single antioxidants against AD which are hard to have a scientific rationale. A screening action at a population level should be started to identify vascular risk factors which are in fact the problem in AD pathophysiology; health practitioners in the field of AD should clearly explain to their cognitively impaired patients that their memory problems are a way to get older and implement a networking task force to early identify and aggressively treat vascular problems at any age, the sooner the better. Simultaneously, students should be made clear of the dramatic importance of disease prevention (in the curriculum of studies and exams), and health insurances worldwide should be once for all made aware of the fact that improving AD prevention by financial support of vascular screening and management will be enormously less expensive than the costs of anti-AD drug reimbursement and disability costs.

I am perfectly aware of the majority of the thoughts of the readers, sustainability, implementation, feasibility, etc. I don't want to use the time to be dedicated to my patients to start a probably utopic change campaign for AD management; therefore, I am grateful to Jack de la Torre for the possibility that many sharing his opinion will know each other and start working together and go back to my patients and students for the everyday pragmatic fight against misleading thoughts on cognitive impairment with and without dementia.

After many, probably too many, years of amyloid hypothesis, it’s time to be brave and test other pathogenetic hypothesis. It doesn’t mean that amyloid is not an important aspect in neurodegeneration and expression of the disease but its position in the cascade of events leading to AD needs to be reconsidered. de la Torre suggests a major role of vascular factors and this proposal is supported by the observation that the main risk factors for AD are vascular factors. Nevertheless, we cannot forget that the strongest risk factor for AD is aging, which, in turn, is characterized by reduction of energy (mitochondrial failure?), altered metabolism, immunosenescence. Furthermore many alterations of the CNS are reflected peripherally in AD. So, should we start to think of AD as a systemic disease, especially in old age subjects? Enlarging our horizon of research to embrace the concept that the failure of different but tightly connected systems is the cause of AD is tantalizing but also confusing, since our science requests to simplify as much as possible a complex problem. Fortunately we have now many theoretical models and bioinfomatic tools whose application could be helpful for reading and finding a sense in the complexity of disease biology. A more strict collaboration among different fields of research should be implemented, also to avoid that a theory becomes a dogma, reducing possibility to look beyond or somewhere else.

In the chapter titled Prevention is the Cure, Siddhartha Mukherjee describes in his epic work, “The Emperor of All Maladies: A Biography of Cancer,” [1] the decades of struggle to accept and take appropriate measures to control the undeniable association between tobacco smoking and lung cancer. In the 1950s, Richard Doll and Bradford Hill in the UK [2,3] and Ernst Wynder and Evarts Graham in the USA [4] demonstrated unequivocally that, in comparison with controls, history of smoking occurred in all cases of lung cancer. These pivotal studies in two cohorts at both sides of the Atlantic were dismissed as unimportant. Turning to a behavior that has become familiar to science, instead of addressing this epochal discovery the scientific community focused the controversy on the meaning of the word “cause.” Epidemiologist Sir Austin Bradford Hill solved the issue by providing the mathematical features that strengthen the concept of causality but it took more than 30 years and untold numbers of cases of lung cancer and other complications of smoking—including stroke—to finally insert cigarette warnings and limit tobacco advertisement in the USA.

The field of dementia has also had its share of changes in nomenclature (including the recent banning of the time-honored term “dementia”). The central controversy has been the role of cerebrovascular disease (CVD) in Alzheimer’s disease (AD). In 1997, David Snowdon and colleagues demonstrated in the Nun Study [5] that the presence of lacunes in the basal ganglia, thalamus, or deep white matter increased 20.7 times the odds of clinical dementia, often in the presence of fewer neuropathological lesions of AD. The authors concluded: “CVD may play an important role in determining the presence and severity of the clinical symptoms of AD.”

During the past two decades, a solid body of clinical, brain imaging, epidemiological and neuropathologic evidence has accumulated confirming the critical role of CVD in AD and other dementias of the aged. Recently, data from the National Alzheimer’s Coordinating Center in the United States [6] demonstrated the presence of vascular pathology in 80% of 4,629 brains from patients with neuropathologically confirmed AD. Lesions included large-vessel disease with atherosclerosis in the arteries of the circle of Willis and its branches resulting in large territorial infarcts, small-vessel disease with arteriolosclerosis and small infarcts including lacunes and multiple microinfarcts, ischemic periventricular leukoencephalopathy and brain hemorrhages. Notably, cerebral amyloid angiopathy was present in less than half of the brains (41%). Summaries of the extensive literature available on the interaction between AD and CVD are available from the American Heart Association/American Stroke Association [7] and the International Society for Vascular Cognitive Disorders [8].

The universal failure of therapies destined to control tau and beta amyloid deposition has been received with widespread disappointment. A more positive attitude has been recommended by epidemiologists who have suggested that the from the public health viewpoint, early and comprehensive treatment of all vascular risk factors for stroke is required to delay the onset and slow down the progression of AD. Treatment of elevated total blood homocysteine (tHcy) with B-group vitamins has been demonstrated in a recent clinical trial to effectively stop cortical brain atrophy in patients with Mild Cognitive Impairment from early AD. Along the same lines, we recently found that MTHFR gene mutations may result in elevated tHcy and may be an early marker of late-onset AD.

The World Health Organization and the G8 Dementia Summit recently identified dementia and AD prevention as a major public health priority. Three large European multidomain prevention trials have been launched and the results are awaited with interest [9] because control of vascular risk factors appears to be resulting in a decrease in incidence of dementia in the elderly [10]. Trust the epidemiologists, they seem to know better.

References
[1] Mukherjee S (2010) The Emperor of All Maladies: A Biography of Cancer. Scribner, New York.
[2] Doll R, Hill AB (1950) Smoking and carcinoma of the lung; preliminary report. Br Med J 2, 739-748.
[3] Doll R, Hill AB (1956) Lung cancer and other causes of death in relation to smoking; a second report on the mortality of British doctors. Br Med J 2, 1071-1081.
[4] Wynder EL, Graham EA (1950) Tobacco smoking as a possible etiologic factor in bronchiogenic carcinoma; a study of 684 proved cases. J Am Med Assoc 143, 329-336.
[5] Snowdon DA, Greiner LH, Mortimer JA, Riley KP, Greiner PA, Markesbery WR (1997) Brain infarction and the clinical expression of Alzheimer disease. The Nun Study. JAMA 277, 813-817.
[6] Toledo JB, Arnold SE, Raible K, Brettschneider J, Xie SX, Grossman M, Monsell SE, Kukull WA, Trojanowski JQ (2013) Contribution of cerebrovascular disease in autopsy confirmed neurodegenerative disease cases in the National Alzheimer's Coordinating Centre. Brain 136(Pt 9), 2697-2706.
[7] Gorelick PB, Scuteri A, Black SE, et al. (2011) Vascular contributions to cognitive impairment and dementia: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 42, 2672-2713.
[8] Sachdev P, Kalaria R, O'Brien J, et al. (2014) Diagnostic criteria for vascular cognitive disorders: a VASCOG statement. Alzheimer Dis Assoc Disord 28, 206-218.
[9] Sindi S, Mangialasche F, Kivipelto M (2015) Advances in the prevention of Alzheimer's disease. F1000Prime Rep 7, 50.
[10] Schrijvers EM, Verhaaren BF, Koudstaal PJ, Hofman A, Ikram MA, Breteler MM (2012) Is dementia incidence declining?: Trends in dementia incidence since 1990 in the Rotterdam Study. Neurology 78, 1456-1463.