Dementia due to Alzheimer’s disease (AD) is a neurodegenerative disorder associated with cognitive and functional decline. The underlying pathological process begins at least a decade before any clinical symptoms [
1] and is characterized by neuronal cell death, extracellular deposits of amyloid-ß (Aβ) and intracellular formation of fibrillary aggregates of abnormally phosphorylated tau [
2]. Activity in the hippocampus, as measured by functional magnetic resonance imaging (fMRI), is among the key emerging neuroimaging markers that allow an improved AD risk prediction [
3]. In AD, hippocampal activity is typically decreased during memory tasks due to hippocampal atrophy. However, several fMRI studies in the prodromal stage of AD (i.e., in Mild Cognitive Impairment, MCI) have found increased hippocampal activity during memory tasks that was associated with memory deficits, subsequent cognitive decline and faster clinical progression [
4‐
6]. Pattern separation, a process thought to critically depend on the hippocampus, has been particularly used as a task to show increased hippocampal activity in patients with MCI [
7,
8]. Pharmacological treatment of hyperactivity in patients with MCI, or in mice with increased levels of Aβ, significantly reduced activity in the hippocampus and improved memory performance in pattern separation tasks [
7]. Although one may expect that reducing brain activation in a given area will lead to a drop in performance, there is emerging evidence that hyperactivity in the hippocampus has a negative, rather than a positive, impact on cognition [
9]. Thus, reducing excess hippocampal activity may present a promising therapeutic target. Pharmacological interventions, however, are prone to side effects such as headache, diarrhoea, or sleep disturbances. In addition, an elderly population is likely to be on other medication so - apart from the possible side effects - drug interaction may be a problem. An alternative approach may be real-time fMRI neurofeedback. With real-time fMRI neurofeedback, participants train to voluntarily ‘control’ region specific brain activity [
10,
11]. The training is accomplished by continuously measuring brain activity in real-time, and providing feedback to the participant about the ongoing activity in the targeted brain area [
11].
The current study aims to test whether real-time fMRI neurofeedback is capable of reducing hippocampal hyperactivity in patients with MCI and, in addition, whether the reduction of hyperactivity will be associated with an improvement in memory performance. Comparable to previous studies, we will deploy pattern separation tasks to assess hippocampal activity as well as memory performance (e.g., [
7,
8]). As there is evidence for an association between Aβ and hippocampal hyperactivity [
9], we will determine Aβ levels in blood. In addition, we will examine variables that may predict neurofeedback success and identify the most successful regulation strategies.