. Vitamin E and donepezil for the treatment of mild cognitive impairment. N Engl J Med. 2005 Jun 9;352(23):2379-88. Epub 2005 Apr 13 PubMed.


Please login to recommend the paper.


  1. Leon Thal and Ron Petersen presented data at the AAN meeting on 4/13/05 that was published today in the NEJM. It compared the effects of donepezil, vitamin E, and placebo in patients who met clinical criteria for amnestic mild cognitive impairment (MCI). The primary endpoint of the study was worsening of clinical outcome so that diagnosis converted to AD. There was no difference among the groups in primary outcome. However, donepezil-treated subjects performed significantly better than the other groups for a year on several cognitive measures. There was little to no appreciable effect of vitamin E. Further, more than half of the subjects were ApoE4-positive in the study. The effects of donepezil were more pronounced in this group, and the effects were seen for longer (18-24 months). Why the effect is more pronounced in ApoE4-positive subjects is not clear. One possibility is that a greater percentage of these patients actually have AD neuropathology underlying their cognitive change. It appears that the effects of donepezil are actually very similar in amnestic MCI to those previously reported in mild to moderate AD. We clearly have a way to go before we have a therapy that delays the onset and slows the progression of AD. While the effects of donepezil appear to be modifying symptoms and not progression, it is important that one can clearly see a positive effect of a treatment in a large, multicenter trial of patients, most of whom probably were in the earliest symptomatic stages of dementia due to Alzheimer disease. In the future, perhaps proof-of-concept studies on new therapies for AD can take advantage of emerging new imaging and fluid biomarkers so that further information can be garnered from smaller numbers of patients and we can gain more information from subgroup analysis.

    View all comments by David Holtzman
  2. The possible effect of this study on the use of cholinesterase inhibitors
    is a complex issue because the widespread use of donepezil and other
    cholinesterase inhibitors is for established AD, not MCI, and the
    study doesn't speak to that. I believe use for AD will continue,
    although effects there are modest and time-limited, too. The impact of
    the results on the use of cholinesterase inhibitors for MCI is less
    clear. First, MCI is not often recognized clinically, and I doubt the
    term is used very often outside of Alzheimer centers and specialty
    memory clinics. In those settings, to my limited knowledge, these
    medications are widely used, and I'm not sure this study would change
    that, as it does show some limited benefits for donepezil. Given the
    recent FDA MedWatch for galantamine (based on excess deaths in large
    MCI trials), I don't think we can assume that all cholinesterase inhibitors will be
    treated the same. In any case, the big change in practice should be
    for vitamin E, also widely used, both by those with a diagnosis of MCI
    and those with memory changes who aren't seen in memory clinics but
    are making decisions about vitamin E on their own. I would guess that
    practice in academic settings will shift away from vitamin E, but I am
    less confident about supplement use in the broader population.

    View all comments by Deborah Blacker
  3. Petersen, Thomas, Grundman, and colleagues are to be congratulated. They have made a major contribution with the publication of this long-awaited, randomized, placebo-controlled clinical trial, conducted under the auspices of the Alzheimer's Disease Cooperative Study, of the potential benefit of vitamin E or donepezil in slowing the progression of mild cognitive impairment (MCI) to clinically overt Alzheimer disease (AD). The results of the primary analysis revealed a negative study, as after 3 years there was no significant reduction of the time to progression to AD among individuals treated with either vitamin E or donepezil compared with those given placebo. The importance of this study, however, lies both in the demonstration that it is possible to systematically identify and enroll a sizeable number of individuals meeting criteria for amnestic MCI in a multicenter study and in the results from secondary analyses showing that donepezil appears to be modestly beneficial for at least some individuals with MCI. For the first 12 to 18 months of the trial, donepezil-treated individuals had a lower risk of progression to AD than the vitamin E or placebo groups. Moreover, for those donepezil-treated individuals who possessed an apolipoprotein ε4 allele, there was significantly reduced risk of progression to AD even at 3 years. Recognizing that none of the currently available symptomatic therapies for AD (including donepezil) are approved for the indication of MCI, the authors carefully address the therapeutic implications of their findings in the Discussion: "Although our findings do not provide support for a clear recommendation for the use of donepezil in persons with mild cognitive impairment, they could prompt a discussion between the clinician and the patient about this possibility."

    The data from this trial may be interpreted in multiple ways. Here I offer my perspective, which may be controversial with many investigators who study MCI and certainly differs from the basic premise of the Petersen et al. study, which is that MCI is a transitional state between the cognitive changes of normal aging and very early AD. Simply put, I consider MCI to be a constellation of cognitive changes experienced by older adults that may be caused by many different conditions, some of which are static or even reversible. Many individuals with MCI thus will not progress to dementia; of those who do, AD is not the only dementing illness that will ensue as it is likely that non-AD disorders also may have MCI as a prodrome. One of the many possible causes of MCI, however, is AD. Indeed, our own data (Morris et al., 2001) show that AD is by far the most common cause of amnestic MCI (the MCI subtype studied in the donepezil/vitamin E trial) and that the neuropathologic phenotype associated with MCI in these individuals is histological AD. For these individuals, the symptoms of MCI thus represent the earliest clinical manifestation of the AD process. This subset of MCI is not a risk factor for AD or a transitional state between normal aging and very early AD; it already is very early AD.

    With this perspective, the results from the trial of donepezil/vitamin E in MCI are very consistent with what is known about cholinesterase inhibitor therapy of AD. That is, cholinesterase-treated individuals demonstrate a modest benefit compared with placebo-treated individuals regarding rate of deterioration as measured by cognitive and functional measures, precisely as was shown in the MCI trial. The fact that the donepezil effect was so slight and became nonsignificant after 18 months likely results from a combination of factors, including the small magnitude of effect associated with cholinesterase inhibitors even in overt AD and the possible dilution of the already weak drug effect by the possible inclusion in the sample of MCI individuals who did not have underlying AD. In support of this last point, the MCI individuals in this trial who were most likely to have underlying AD, as indicated by the presence of an ε4 allele, clearly demonstrated a donepezil effect even at 3 years.

    Where does the field now stand? Clearly, there will continue to be debate as to what proportion of individuals with MCI represent underlying AD and whether clinicians can identify this subset. This same trial has shown that the study participants with MCI, in addition to showing the expected amnestic deficits, also have impaired performance (albeit very mild) in non-memory cognitive domains (Grundman et al., 2004), and recent data presented at the annual meeting of the American Academy of Neurology show that they also have mild but measurable functional impairment as regards instrumental activities of daily living (Galasko et al., 2005). The case can be made, therefore, that these individuals already meet established criteria for AD in that they have objective impairments in multiple cognitive domains, including memory, and are functionally impaired. That they indeed have AD is supported by our studies that show that MCI-equivalent individuals demonstrate clear cognitive deterioration over time in a pattern identical to that of AD (Storandt et al., 2002) and, as noted above, have AD neuropathology. Most clinicians, however, are reluctant to diagnose very early AD at the MCI stage, in part because they may not have the time or inclination to obtain evidence from informants about cognitive and functional decline (informant-based clinical methods are more sensitive than neuropsychological measures for detecting early-stage AD, but most clinicians continue to rely on cognitive test results), and in part because they are not yet comfortable with moving the threshold for dementia diagnosis to the MCI stage. Nonetheless, clinicians who do comprehensively evaluate individuals may increasingly recognize the very mild AD clinical phenotype in many persons who meet criteria for amnestic MCI and may choose to offer treatment with currently approved therapies based on the diagnosis of probable early-stage AD. This decision, of course, must be tempered by the at-best modest benefit associated with cholinesterase inhibitors (the cost-effectiveness of these agents even in overt AD is a matter of controversy; see, for example, the AD2000 Collaborative Group, 2004), although often patients and their families opt for the possibility of even slight benefit for a limited period given the otherwise relentlessly progressive deterioration associated with AD.

    The final comment regarding the Petersen et al. MCI trial results is that better drugs than those currently available clearly are needed to treat AD, even in its earliest symptomatic stages. It may be that truly effective therapies will need to be introduced even prior to the MCI stage, as by the time even minimal cognitive impairment can be appreciated, there already is substantial brain damage. In vulnerable brain regions such as the entorhinal cortex, for example, over 50 percent neuronal loss has occurred in individuals at a stage comparable to that of amnestic MCI (Price et al., 2001). Even the MCI stage may be too late to initiate therapy and achieve meaningful clinical responses. What is needed is a dual approach to identify and develop not only disease-modifying therapies that target disease mechanisms, but also antecedent biomarkers that detect AD changes prior to the occurrence of symptoms (i.e., preclinical detection). Both approaches currently are the focus of much research, although much more needs to be done. If these strategies eventually prove successful, however, they offer hope that AD ultimately may become not only treatable but preventable.


    . Mild cognitive impairment represents early-stage Alzheimer disease. Arch Neurol. 2001 Mar;58(3):397-405. PubMed.

    . Mild cognitive impairment can be distinguished from Alzheimer disease and normal aging for clinical trials. Arch Neurol. 2004 Jan;61(1):59-66. PubMed.

    . Detailed assessment of activities of daily living in moderate to severe Alzheimer's disease. J Int Neuropsychol Soc. 2005 Jul;11(4):446-53. PubMed.

    . Rates of progression in mild cognitive impairment and early Alzheimer's disease. Neurology. 2002 Oct 8;59(7):1034-41. PubMed.

    . Long-term donepezil treatment in 565 patients with Alzheimer's disease (AD2000): randomised double-blind trial. Lancet. 2004 Jun 26;363(9427):2105-15. PubMed.

    . Neuron number in the entorhinal cortex and CA1 in preclinical Alzheimer disease. Arch Neurol. 2001 Sep;58(9):1395-402. PubMed.

    View all comments by John Morris
  4. The Alzheimer's Disease Cooperative Study Group MCI trial results regarding donepezil are very encouraging and should be used to stimulate a primary prevention trial. The New England Journal report did not offer a molecular mechanism to explain donepezil's effect on disease progression. Nitsch, Buxbaum, and colleagues demonstrated a possible mechanism in 1992, showing that activation of muscarinic M1 and M3 receptors results in increased nonamyloidogenic processing of APP (1). Abraham Fisher and co-workers have recently shown that this is due to stimulation of α-secretase activity (Sorrento 2005). We and others have shown (2,3,4) that systemic therapy with cholinergic agents, including both M1 agonists and acetylcholinesterase inhibitors, decreases cortical and CSF Aβ in vivo. Therefore, it is likely that the disease-slowing effect of cholinergic agents is due to their antiamyloidogenic properties. If so, then therapy would be much more likely to be successful if started prior to the MCI stage of AD, as cortical Aβ deposition is already heavy and extensive in MCI. A primary prevention trial of donepezil is warranted and urgently needed.


    . Release of Alzheimer amyloid precursor derivatives stimulated by activation of muscarinic acetylcholine receptors. Science. 1992 Oct 9;258(5080):304-7. PubMed.

    . Reduction of cortical amyloid beta levels in guinea pig brain after systemic administration of physostigmine. Neurosci Lett. 2001 Sep 7;310(1):21-4. PubMed.

    . Reduction of cerebrospinal fluid amyloid beta after systemic administration of M1 muscarinic agonists. Brain Res. 2001 Jun 29;905(1-2):220-3. PubMed.

    . The selective muscarinic M1 agonist AF102B decreases levels of total Abeta in cerebrospinal fluid of patients with Alzheimer's disease. Ann Neurol. 2000 Dec;48(6):913-8. PubMed.

    View all comments by Thomas Beach
  5. 1. It is scientifically and nutritionally ridiculous to give vitamin E without simultaneously replacing peroxidized long-chain omega-3 and omega-6 polyunsaturated essential fatty acids in critical brain synapses (and in mitochondrial membranes, to prevent uncoupling and aqueous oxidation). Soderberg showed a major (50 percent) loss of DHA in Alzheimer cortex at postmortem, while Corrigan et al., 1991 saw some definite improvement in AD cases given vitamin E along with omega-6 fatty acids. Why are the vitamin E enthusiasts so unaware of the need to replace the very same essential fatty acids that are presumably being lost all the time, due to the imagined vitamin E deficiency? I would suggest giving vitamin E again, accompanied by a moderate dose of fish oil and evening primrose oil. 2. A half or more of all AD cases appear to have depression, which suggests a lifelong history of anxiety disorder, and which may lead to other pathology, like cortical Lewy bodies. The typical cognitive problems of anxious people, which worsen in midlife, respond very quickly to Inositol powder supplement (personal observations). One should take a proper history from MCI cases, detecting chronic anxiety, for which Inositol is the only scientific treatment available. Interestingly, Inositol regulates not only serotonin and noradrenaline neurotransmission, but also cholinergic function, including the M1 and M3 receptor pathways that inhibit amyloidogenesis. Anyone for Inositol? 3. Sadly, it may be too late to treat MCI, if the underlying AD pathology has indeed progressed further than expected. But it is never too early to prevent Alzheimer's, if one knows what causes the whole thing, which is chronic ingestion of unprotected linoleic acid in steam-deodorized, vitamin E depleted polyunsaturated food oils (Kalmijn et al., 1997).


    . Essential fatty acids in Alzheimer's disease. Ann N Y Acad Sci. 1991;640:250-2. PubMed.

    . Total homocysteine and cognitive decline in a community-based sample of elderly subjects: the Rotterdam Study. Am J Epidemiol. 1999 Aug 1;150(3):283-9. PubMed.

    . Fatty acid composition of brain phospholipids in aging and in Alzheimer's disease. Lipids. 1991 Jun;26(6):421-5. PubMed.

    View all comments by Robert Peers
  6. Petersen and colleagues have recently presented the results of a superb study in which they showed convincing evidence that in subjects who met the clinical criteria for amnestic mild cognitive impairment (MCI), donepezil, a cholinesterase inhibitor, or vitamin E, an exogenous antioxidant, were without effect on primary outcomes. However, while donepezil was associated with a lower rate of progression to Alzheimer disease (AD) during the first 12 months, vitamin E had no appreciable effect (1).

    The results of the study raise some important questions: Is vitamin E an appropriate therapeutic approach for MCI or AD? In other words, should vitamin E stay or should it go? Does this trial refute the “oxidative hypothesis” of AD? Before we try to answer these questions, it is important to point out that this hypothesis does not necessarily predict that vitamin E (or other antioxidants) will ameliorate the human disease, but that oxidative damages/events play a significant role in the development of AD. A corollary to the hypothesis is that some appropriate antioxidant intervention, at some appropriate dosage, in appropriately selected patients over an appropriate time interval has the potential to improve the prognosis.

    How many of those criteria were fulfilled in the published study? While there is no doubt that vitamin E is a potent antioxidant, the amount used in the study is high, and because this is not a short trial, we do not know if any of the subjects included in the trial had increased indices of oxidative stress before starting the therapy. This is a very crucial point because in healthy people it is known that high and even ultra-high doses of vitamin E are insufficient to decrease basal levels of oxidative stress (2). Furthermore, there is a large amount of literature showing that the levels of different markers of oxidative stress in MCI subjects consistently overlap with those of elderly controls, suggesting that a subgroup of these patients do not have oxidative stress and, as a consequence, probably will not benefit from an antioxidant therapy (3-5). Taken together, these data strongly support the concept that quantitative measurements of oxidative stress are needed for the selection of MCI (or AD) subjects (and any other clinical setting) in clinical trials with antioxidants. This approach will allow us to select the appropriate population who will potentially benefit from this therapeutic strategy. Thus, in future clinical trials with “antioxidants” it will be extremely important to obtain measures of oxidative stress before and after the therapy. By doing so, we could then attempt to demonstrate if any correlation exists between the therapy, the reduction of such parameters (markers), and the clinical outcomes. Another important point is the compliance with the therapy. Unfortunately, no data are presented on the circulating plasma levels of vitamin E before and after the supplementation therapy. This is a crucial piece of information if one wants to conclude that the vitamin was ineffective.

    In summary, with these caveats in mind, the results with vitamin E are far from being clear-cut. Only when all of these requirements are satisfied, will we prove in a conclusive way whether or not an early intervention with this (or other) antioxidant(s) can really delay the onset of AD.


    . Vitamin E and donepezil for the treatment of mild cognitive impairment. N Engl J Med. 2005 Jun 9;352(23):2379-88. Epub 2005 Apr 13 PubMed.

    . Effects of vitamin E on lipid peroxidation in healthy persons. JAMA. 2001 Mar 7;285(9):1178-82. PubMed.

    . Increase of brain oxidative stress in mild cognitive impairment: a possible predictor of Alzheimer disease. Arch Neurol. 2002 Jun;59(6):972-6. PubMed.

    . Evidence of increased oxidative damage in subjects with mild cognitive impairment. Neurology. 2005 Apr 12;64(7):1152-6. PubMed.

    . Oxidative DNA damage in peripheral leukocytes of mild cognitive impairment and AD patients. Neurobiol Aging. 2005 May;26(5):567-73. PubMed.

    View all comments by Domenico Pratico

Make a Comment

To make a comment you must login or register.