Commentary
Causes or consequences of inflammation and pathological signs of
Alzheimer disease
Allen D. Roses*
Glaxo Wellcome Research and Development, Research Triangle Park, NC 27009, USA
Received 21 March 2000
1. Introduction
An open peer commentary is an unusual editorial mech-
anism for providing insight from invited “experts” who are
thought to know a lot about particular subjects, in this case
inflammation and Alzheimer disease (AD). In fact, such
commentaries are usually written to present a broader con-
text for readers. Inflammation is not an area of high com-
petence for me, but I have a minority (perhaps unique!)
viewpoint on the etiology and pathogenesis of AD.
As a physician, I view disease in general, and AD in
particular, through history, symptoms, and signs, not as a
neuroscientist studying pathologic signs of disease in end-
stage brain tissue. A patient’s illness begins with history,
and develops with complaints (symptoms) and evidence of
illness (signs). Frequently etiology (cause) and pathogenesis
(process) are implied from associated symptoms and signs.
The documentation of lifelong genetic polymorphisms or
mutations underlying early- and late-onset AD produces a
different view of disease etiology, with long-term gradual
pathogenesis occurring much earlier in time than the end-
stage necropsy tissues that are so intensely studied [10,12].
The pathological footprints left by the march of disease are
frequently mistaken for the trajectory that the disease pro-
cess forced the brain to take [11,15]. Therefore, the presence
of inflammation-related proteins, amyloid deposition,
plaque formation, neurofibrillary changes, and the loss of
brain mass are all signs or pathological footprints (conse-
quences) of underlying, unmeasured, small metabolic vari-
ations that occur over decades, producing symptomatic dis-
ease with varying ages of observable onset.
I hasten to add that treatments may be directed effec-
tively at clinical symptoms, or to delay the development of
associated neuropathology signs, including inflammatory
processes. Whether inflammation and neuritic plaques
cause AD is a very different question than whether delaying
the process of neuritic plaque formation will slow the pro-
gression of AD symptoms. However, statistical data con-
cerning the amelioration of symptoms should not be con-
fused as evidence supporting those processes as etiological
cause. One can subscribe to the thesis of Rogers et al.
regarding the role of inflammatory processes, or believe that
the formation of extracellular amyloid and neuritic plaques
cause AD. Both assign causation to pathological signs
loosely associated with the progression of symptoms. The
genetic data suggest that the disease reflects the long-term
consequences of small metabolic variations, which can be
measured in subjects who can be predicted to be at greater
risk. Thus, one only has to review the published data relat-
ing amyloid density to APOE genotype to view amyloid
deposition as a function of genetics and time. In the original
reports of the association of APOE and AD, higher density
plaques were found in tissue from APOE-4/4 patients than
from APOE-3/3 patients [8,14]. Despite controversial non-
confirmations along the way, all one has to read is the
current experimental literature to conclude that APOE is
necessary for amyloid deposition, and that the density re-
lates to the APOE genotype—just as in the original reports
[4,5]. However, it is also clear than the density of amyloid
deposition is unrelated to the course of disease once it starts
regardless of the APOE genotype [8]. Inflammatory con-
comitants of the plaque appear to be secondary as well [6].
Debates in AD (and many other diseases) frequently use
the “chicken or egg” analogy. Actually, all chickens came
from a fertilized egg that contained DNA from both a
rooster and a hen. A chicken is the product of a genetically
determined process, which can be influenced by environ-
mental factors—just as people. If we view AD as a heter-
ogeneous collection of disease-causing mutations and mul-
tiple genetic variants that can influence the susceptibility to
developing the disease as a function of age, then whether
* Tel.: ϩ1-919-483-7086; fax: ϩ1-919-315-6013.
E-mail address: adr69412@glaxowellcome.com (A.D. Roses).
www.elsevier.com/locate/neuaging
Neurobiology of Aging 21 (2000) 423–425
0197-4580/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved.
PII: S0197-4580(00)00142-1