Functional neuroimaging findings in healthy middle-aged adults at risk of Alzheimer's disease.

It is well established that the neurodegenerative process of Alzheimer's disease (AD) begins many years before symptom onset. This preclinical phase provides a crucial time-window for therapeutic intervention, though this requires biomarkers that could evaluate the efficacy of future disease-modification treatments in asymptomatic individuals. The last decade has witnessed a proliferation of studies characterizing the temporal sequence of the earliest functional and structural brain imaging changes in AD. These efforts have focused on studying individuals who are highly vulnerable to develop AD, such as those with familial genetic mutations, susceptibility genes (i.e. apolipoprotein epsilon-4 allele), and/or a positive family history of AD. In this paper, we review the rapidly growing literature of functional imaging changes in cognitively intact individuals who are middle-aged: positron emission tomography (PET) studies of amyloid deposition, glucose metabolism, as well as arterial spin labeling (ASL), task-dependent, resting-state functional magnetic resonance imaging (fMRI) and magnetic resonance spectroscopy (MRS) studies. The prevailing evidence points to early brain functional changes in the relative absence of cognitive impairment and structural atrophy, although there is marked variability in the directionality of the changes, which could, in turn, be related to antagonistic pleiotropy early in life. A common theme across studies relates to the spatial extent of these changes, most of which overlap with brain regions that are implicated in established AD. Notwithstanding several methodological caveats, functional imaging techniques could be preferentially sensitive to the earliest events of AD pathology prior to macroscopic grey matter loss and clinical manifestations of AD. We conclude that while these techniques have great potential to serve as biomarkers to identify at-risk individuals, more longitudinal studies with greater sample size and robust correction for multiple comparisons are still warranted to establish their utility.