Cognitive training has been shown to be successful in patients with amnestic mild cognitive impairment (aMCI), a group at high-risk for Alzheimer's disease (AD). Moreover, in a randomized controlled trial, the investigators recently found that aMCI patients receiving a cognitive intervention showed stable hypometabolism in FDG-PET, whereas patients in an active control group showed pronounced hypometabolism on follow-up scans in regions typically affected in AD.

Previous studies indicate that not all patients respond equally well to a cognitive intervention. Identifying factors that predict response to treatment could help selecting patients for a targeted intervention. A potentially important predictor is cognitive reserve defined as premorbid cognitive performance. The hypothesis is that different levels of cognitive reserve (high cognitive reserve vs. low cognitive reserve) have different neurostructural and neurofunctional correlates and influence treatment response in a different way. Moreover, the impact of white matter lesions on treatment effects will be investigated.

The investigator will perform a complex cognitive training program. Forty patients with aMCI (20 with high cognitive reserve, 20 with low cognitive reserve) will be recruited in this study. Since the patients are recruited consecutively, an estimated overall number of 80 will be included and receive the training of whom about 40 will meet the inclusion criteria for our cognitive-reserve-study (high or low cognitive reserve). Data of the whole group will be used to analyze the potential impact of white matter lesions on response to the intervention. Cognitive effects of the intervention will be evaluated by neuropsychological testing. Neurofunctional and neurostructural changes depending on cognitive reserve will be measured using resting state fMRI and diffusion tensor imaging (DTI).

For comparison, a group of 30 aMCI patients will be recruited as an active control group receiving study investigations (neuropsychological testing as well as MRIs), and exercises for self-study at home, not the complex cognitive intervention.

Inclusion Criteria:

• male or female patients, fulfilling Petersen´s criteria of amnestic mild cognitive impairment (aMCI)
• No evidence for other psychiatric axis I disorders according to DSM-IV criteria.
• No evidence for neurological disorders (e.g. stroke)
• No uncontrolled arterial hypertension or diabetes mellitus
• No history of drug / alcohol abuse
• The patient is able to provide written informed consent to participate in the study.
• for the cognitive-reserve-study patients, a high (low) cognitive reserve is defined as a verbal IQ score of >/= 120 (\

Consecutive aMCI patients of an University-based memory clinic