Alzheimer's disease (AD) is the most common form of dementia in late life, affecting approximately 10 per cent of individuals of 65 years of age, with the prevalence doubling every five years up to the age of 80, above which the prevalence is 40 per cent [1
]. In 2008, in the United States alone, there were more than 5 million people with AD, which is a 10 per cent increase from the previous (in 2000) prevalence estimate of 4.5 million (Figure , adapted from Hebert et al
]). Age-related mild cognitive impairments may affect two to three times as many individuals [3
]. By 2050, the number of elderly people with AD in the USA could range from 11 million to 16 million, which strongly calls for strategies to prevent or delay the onset of the disease.
Figure 1 Incidence of AD cases in the USA. State-specific projections up to 2025 (adapted from Hebert et al.).
AD is a neurodegenerative disorder characterised by global deficits in cognition, ranging from memory loss to impaired judgment and reasoning [5
]. Clinical diagnosis per se
is often uncertain and clinical assessment requires multiple examinations and laboratory tests over time. Despite thorough clinical exams, the frequency of unrecognised dementia in the community ranges from 50 per cent to 90 per cent of cases [6
]. Insidious onset and progressive impairment of memory and other cognitive functions make the initial stages of AD difficult to distinguish from so called 'normal ageing'. In order to develop prevention treatments for AD, it is necessary to identify persons who are still cognitively normal (NL), but are either at very high risk for developing the disease or are in an early, pre-symptomatic stage of the disease. Such individuals are most likely to benefit from therapies which are instituted when the potential for preservation of function is the greatest, well before irreversible synaptic and neuronal injury.
The difficulty in studying the genetics of complex, age-associated disorders such as late onset AD (LOAD) using traditional clinical case identification measures has resulted in an increasing emphasis on studying endophenotypes of AD, which provide more specific targets for genetic studies. In 2007, it was shown that NL children of mothers affected with LOAD express a biological phenotype characterised by progressive reductions in brain glucose metabolism in the same brain regions as clinical AD patients, as measured on 2-[18
F]fluoro-2-deoxy-D-glucose positron emission tomography (18
F-FDG PET) [7
]. By contrast, children of AD-affected fathers and of parents with no dementia did not show metabolic abnormalities [7
]. The genetic basis for the selective maternal transmission of metabolic deficits in AD is not known.
The present review first summarises known genetic mechanisms involved in the early- and late-onset forms of AD, and in vivo brain imaging studies that provide a link between genetics, pathology and pathophysiology in AD. It then presents recent findings of progressive reductions in brain glucose metabolism in adult children of mothers affected with LOAD, and discusses possible genetic and epigenetic mechanisms involved in maternally inherited brain hypometabolism. Finally, it examines other neurological disorders with known or suspected maternal transmission, and discusses the role of imaging in identifying endophenotypes in preclinical AD.
Although many imaging methods are available, this review will focus on the two most widely utilised brain imaging techniques in both clinical practice and research studies in AD -- magnetic resonance imaging (MRI) and PET. Various MRI and PET studies in AD have been published recently [8
]. Briefly, these methods have long been used in AD for the detection of structural brain changes and volume reductions due to neuronal loss (ie atrophy) on MRI, and reductions of cerebral metabolic rate of glucose (CMRglc) on 18
F-FDG-PET. Both modalities accurately discriminate AD from controls, predict decline from normal cognition to dementia and correlate with disease progression and histopathological diagnosis [8
]. In addition, 18
F-FDG-PET has the advantage of being particularly useful for distinguishing AD from other forms of dementia [10
]. Overall, AD patients present with brain atrophy and hypo-metabolism in the parieto-temporal, posterior cin-gulate, medial temporal and prefrontal cortex, while the primary motor and visual areas, cerebellum, thalamus and basal ganglia are relatively spared [8
In addition to traditional imaging techniques, a turning point in AD was accomplished with the recent development of PET tracers for amyloid-beta (Aβ) plaques, which allow examination of Aβ deposits in vivo
. Among different amyloid PET ligands, the best characterised is the N
C]2-(4'-methylaminophenyl)-6-hydroxybenzoth-iazole, also known as Pittsburgh compound-B (11
]. AD patients consistently show 11
C-PIB uptake in amyloid-rich brain regions, particularly in the frontal, parieto-temporal, posterior cingulate, precuneus and occipital cortices, as well as the thalamus and striatum, whereas the control population mostly showed non-specific binding restricted to the white matter [8
]. A combination of imaging modalities may help to make an earlier diagnosis, and to define the nature and extent of AD pathology among pre-symptomatic individuals.