Mild Cognitive Impairment (MCI) is a condition in which patients show an objective evidence of impairment in one or more cognitive domains but a spared independence in daily functional abilities. The cognitive domain which results to be the most impaired defines the clinical subtype. Specifically, the amnestic subtype of MCI (aMCI), which is the most frequent, is characterised by deficits in episodic memory and visuospatial abilities. Current literature reports that prompt and focused rehabilitation provided early in the disease course might slow down the development of the neurodegenerative condition. The coupling between cognitive and aerobic trainings is the most incisive non-pharmacological treatment, and its effects have been demonstrated to have an impact on global cognitive functions, quality of life, aerobic capacity, and mood in the elderly with aMCI. However, the best modality for coupling aerobic and cognitive trainings (i.e., whether sequential-SEQ: aerobic training followed by cognitive training; or simultaneous-SIM: cognitive training during aerobic training) is still unknown. The aim of the present study is to determine which modality, among SEQ or SIM, is the most incisive training in patients with aMCI. Using a Randomized Controlled design, the effect of this combined aerobic-cognitive non-pharmachological training in both modalities will be evaluated on cognitive changes assessed with standard and computerized neuropsychological batteries and in functional activity during a task-based functional Magnetic Resonance Imaging (fMRI) scan.

This study will be the first one aimed at investigating the effects of a combined aerobic and cognitive training in the two different paradigms (SIM and SEQ) on cognitive functioning and brain functional MRI activity in patients with aMCI. Previous studies on aMCI patients reported that the combination of cognitive and aerobic training is more effective than the single treatments alone on both clinical and MRI outcome measures. Specifically, with the proposed training we expect to obtain ameliorative effects in terms of episodic memory, visuospatial abilities, cerebral activity and aerobic capacities, with a consequent benefit on quality of life and mood in both experimental groups; in addition, we also expect a reduction of IL-6 and TNF-alpha expression levels, and an enhancement of the BDNF.

The combined aerobic-cognitive training (in SIM and SEQ modalities) will have a duration of 3 months, with 24 sessions scheduled twice a week for 70 minutes each.

The first session will last 10 minutes more compared to the others in order to provide participants with all information concerning the cognitive task that they will perform during (in SIM condition) or after the aerobic training (in SEQ condition).

Aerobic and cognitive exercises will be the same for both groups (SIM and SEQ) and will be characterized as follows:

• Aerobic training: 30 minutes of cyclette at 40-59% heart rate reserve (HRR) or 64-76% maximum heart rate (corresponding at Rating of Perceived Exertion by Borg -BORG RPE= 12-13 -moderate intensity-).16 The cardiac frequency will be monitored by pulse oxymeter.
• Cognitive training: the entire training will last for 30 minutes and will consist of two cycles of a memory task (lasting 10 minutes) and a visuospatial task (5 minutes).

For the SIM condition, patient will perform the cognitive training sited on the cyclette using an adapted computer screen and a joystick.

For the SEQ condition, patient will perform the cognitive training sited on a desk using a computer and a joystick.

The aerobic training will be performed on a cyclette. For the cognitive training, the cognitive exercises will train two cognitive domain 1) memory and 2) visuospatial function. 1) During the mnemonic exercises, the patients will be asked to observe a screen on which different stimuli (objects) will be presented and to remember all of them. Then the subjects will be asked to recall the most of the observed objects. The number of stimuli (objects) will progressively increase in order to progressively improving the difficulty of the mnemonic training. 2) The visuospatial task will consist on the navigation in a labyrinth projected onto a screen. The subjects will have a mouse/joystick with which they will move through the labyrinth in order to find the shortest way to exit.

Inclusion Criteria:

• a diagnosis of amnesic mild cognitive impairment according to the current diagnostic criteria;
• age>60 years;
• stable acetylcholinesterase inhibitors and/or memantine at the time of enrolment for at least one month;
• subscription of the consent for personal data treatment and patients' informed consent for participation in the present study.

Exclusion Criteria:

• patients with a history of other systemic, neurologic, psychiatric diseases, head injury, cardiovascular events, and cerebrovascular alterations visible at an MRI scan;
• patients with a family history of neurodegenerative disorders;
• patients with a history of alcohol and/or psychotropic drugs abuse;
• patients who are used to do regular physical and/or cognitive activity;
• contraindication to perform MRI scan (cardiac pace-maker or other types of cardiac catheters, splinters or metallic shards, metallic prosthesis not compatible with magnetic field generated by MRI, claustrophobia, pregnancy, breastfeeding).

Blinding will be maintained for allocation concealment, which will be carried out through closed and opaque mails by a blind operator. Patients will be labelled with a numerical code to ensure anonymisationPatients, assessors (neurologists, neuropsychologists and physiotherapists who will evaluate patient inclusion in the study) and statisticians will be blind to allocation.