In clinic and research you would use McKhann's et al and Albert et al NIAA criteria (both 2011) for the diagnosis of dementia due to AD. But of course AD starts before dementia develops and there is a concept of AD-MCI. The main differentiation is about whether cognitive changes disrupt the ability for independent living - so yes, it is arbitrary and in fairness in clinic not always that obvious

There are open -access papers on these sites that you'll find useful:

http://www.alzheimersanddementia.com/article/S1552-5260(11)00101-4/abstract

http://www.alzheimersanddementia.com/article/S1552-5260(11)00104-X/abstract

From a biological viewpoint, there is now proof that the pathogenic cascade leading to AD starts years to decades before the clinical onset of the disease, so the boundary between AD-MCI and AD is somehow artificial and may actually hinder any attempt at providing treatment. However, the MCI concept is not completely useless if one considers that there are causes of MCI other than AD, and the clinical syndrome of MCI may lead to different outcomes. The hot issues now would be to classify MCI according to the undelying biology and not on the neuropsychological severity, and to provide prognostic markers of progression potential and pace.

At Wash U we typical use the Clinical Dementia Rating (CDR, Morris, 1993) in lieu of MCI/AD. The CDR represents impairment as being a continuum rather than discrete stages. I think the field in general has shifted to this view rather than earlier notions that MCI is a fundamentally unique stage. As Francesco suggested, a lot of MCI labels may actually be including fronto-temporal dementia, vascular dementia etc. There have been a number of excellent theoretical papers from Cliff Jacks' group about temporal dynamics in AD as well as a couple demonstrations of these theories in work from Randy Bateman and Tammie Benzinger. They do a nice job depicting the temporal relationship between disease pathology and clinical decline.

Thanks everyone for the inputs. MCI can be a precursor of many dementias but when we specifically talk about MCI-AD, I think CDR is a good objective scale. CDR of 0.5 may be taken as MCI where as 1,2,3 may be mild, moderate and severe dementia.But here also a problem arises, a diagnositic criteria of MCI does not require CDR 0.5. As far as my understanding, CDR assesses the presence of clinical cognitive impairment and if present it assesses its severity. The differentiation between MCI and Mild Alzheimer's dementia may the impairment in more than one cognitive domain and ceasing to be independent of daily activities.

In some patients, MCI is a preclinical phase of AD. However, in other patients the cognitive decline of MCI remains stable and other factors may explain the cognitive decline: depression, low educational level. The key question is the deterioration of the ADL, present in the MCI-AD.

Conde-Sala JL, Garre-Olmo J, Vilalta-Franch J, Llinàs-Reglà J, Turró-Garriga O, Lozano-Gallego M, Hernández-Ferrándiz M, Pericot-Nierga I, López-Pousa S. Predictors of cognitive decline in Alzheimer's disease and mild cognitive impairment using the CAMCOG: a five-year follow-up. International Psychogeriatrics, 2012; 24(6): 948-958

I personally believe that the label MCI can be useful only temporary because of two reasons: if there really is a pathology, it will affect ADL and will become a dementia; if there really is not a pathology, it will probably revert to normal.

Besides, there are so many definitions of MCI and so many statistical criteria (1, 1'5 or 2 SD below the mean on one, two or more tests). SO i always follow-up every case diagnosed woth MCI because of the lack of specificity.

I personally do not agree with the education theory. I think that high educational levels do not protect. It's just that you cannot identify with the same results. But a very high score on a test may indicate impairment in a high-educated person. Most importantly, a complex activity that "average" people cannot do may indicate dementia in a high educated person if he/she was able to do it but is not anymore.

I would strongly recommend these very interesting papers by Sachdev et al., Matthews et al. and by Mura et al. about reversion to normal and education on AD.

Best,

Javi

Sachdev PS, Lipnicki DM, Crawford J, Reppermund S, Kochan NA, Trollor JN, et al. Factors Predicting Reversion from Mild Cognitive Impairment to Normal Cognitive Functioning: A Population-Based Study. PLoS ONE 2013; 8(3): e59649. doi:10.1371/journal.pone.0059649

Matthews FE, Stephan BC, McKeith IG, Bond J, Brayne C. Two-year progression from mild cognitive impairment to dementia: to what extent do different definitions agree? J Am Geriatr Soc 2008; 56(8): 1424– 1433.

Mura T, Proust-Lima C, Jacqmin-Gadda H, Akbaraly TN, Touchon J, Dubois B, Berr C. Measuring cognitive change in subjects with prodromal Alzheimer’s disease. J Neurol Neurosurg Psychiatry 2013; doi:10.1136/jnnp-2013-305078

A multicenter study led by Christian Haass and Michael Ewers of Ludwig-Maximilians-Universitaet in Munich has identified a biomarker associated with the activation of an innate immune response to neural damage during early stages of Alzheimer's disease.