The present study aims at combining biochemical methods with various types of imaging techniques to identify the pathophysiology of Alzheimer's disease (AD). The main interest is to find markers associated with the very early steps in the pathology of this disease. The investigators shall thus screen for i) molecules in cerebrospinal fluid (CSF) and plasma specific for AD, and ii) brain imaging markers (e.g. MRI and PET) that correlate to detailed clinical assessments.

Biomarkers of interest would then be useful to:

1. Enable accurate detection of the disease early on. Such biomarkers need to specifically reflect the very early pathophysiology of AD and distinguish it from disorders with similar symptomatology, such as other types of dementia and major depression. The sensitivity and specificity of these biomarkers in combination with clinical assessment should be of at least 90%.
2. Enable prediction of the course of events of the disease, such as the disease rate in individual patients. Biomarkers that can predict the pattern of future symptoms will be extremely valuable.
3. Allow monitoring of early effects of new disease-modifying therapies (so-called surrogate biomarkers). Currently clinical therapeutic trials for AD require large patient groups together with long-term treatment. Both size of the groups and treatment time will be reduced with the help of surrogate biomarkers.
4. Study the pathogenesis of the disease. Biomarkers can be used to investigate in detail early alterations in AD patients. For instance, changes in the levels of certain molecules in CSF together with genetic predisposition could then be correlated to clinical signs and changes detectable by brain imaging. This can lead to identification of new therapeutic targets that could easily be monitored in future trials.

1. Baseline investigations of patients with mild cognitive deficits

   We are conducting a prospective, longitudinal study in which we consecutively include patients with mild cognitive deficits (MCI), who seek medical care at the Neuropsychiatric Clinic (Malmö, Sweden) or Unit for Cognitive Medicine (Lund, Sweden). At baseline the MCI patients undergo detailed neurological and psychiatric examination, including assessment of depressive symptoms and ADL-capacity as well as cognitive and motor tests. Patients are also genotyped for APOE. Samples of plasma, blood (for DNA and mRNA) and CSF are also collected. All patients undergo an advanced MRI scan of the brain. A subset will undergo 18F-Flutemetamol PET. We will include patients over a period of three years.

   1.2 Follow-up of MCI patients

   Thereafter, we follow patients for up to 10 years with repeated testing and clinical evaluation. During clinical follow-up we estimate how many of the patients develop any type of dementia, for instance AD. Moreover we also estimate how aggressive the progression of the disease is in those patients that develop AD with the help of repeated cognitive testing.

2. Baseline investigations of healthy elderly volunteers

   To answer the question if new biomarkers could detect early signs of AD in healthy people, we have included cognitively healthy elderly subjects. These people are recruited from a population-based study in Malmö ("Malmö Kost Cancer") where people without memory problems or cognitive difficulties, and who performs well on cognitive tests, are offered to participate. These individuals will undergo the same baseline studies that MCI patients (see above), including cognitive tests, psychiatric assessment, lumbar puncture, blood tests and MRI scan. A subset is also examined with 18F-Flutemetamol PET.

   2.1 Follow-up of elderly volunteers

   This population will also be followed-up for 10 years with repeated cognitive tests to determine which subjects develop cognitive impairment (e.g. memory problems) over this period of time.

3. Analyses of CSF and plasma/blood

   CSF and plasma/blood sampling is done at baseline, 2, 4, 6 and 10 years of follow-up. To find novel and better biomarkers to predict AD in both healthy and MCI patients, the CSF, plasma and blood will be analyzed by various biomedical techniques. We will also screen for biomarkers that can help us to predict how fast the disease will progress. We will use two different approaches, namely: a) analysis of various candidate biomarkers and b) unbiased screening using proteomics.

4. Magnetic resonance imaging (MRI) MRI is done at baseline, 2, 4 and 6 years of follow-up, using the same MRI scanner. We will evaluate the potential benefits of new MRI protocols for prediction of future AD. MRI will also be used to study the pathogenesis of AD. These new approaches include: 3D-MP RAGE, T2* GRE, DTI/DTT, ASL and MRS.

Inclusion Criteria:

Healthy elderly subjects

• No cognitive symptoms reported by patient and/or informant
• Normal performance on cognitive tests
• General cognition and functional performance preserved such that a diagnosis of MCI or dementia cannot be made by physician at the time of the baseline visit
• Between 60 and 90 years of age
• Fluent in Swedish
• Agrees to at least one lumbar puncture, and neuropsychological testing.

Mild cognitive impairment

• Cognitive symptoms reported by patient and/or informant
• Between 60 and 80 years of age
• Mini-Mental State Exam score between 24 and 30
• General cognition and functional performance sufficiently preserved such that a diagnosis of dementia cannot be made by physician at the time of the baseline visit
• Fluent in Swedish
• Agrees to at least one lumbar puncture, and neuropsychological testing.

Exclusion Criteria (for both MCI and healthy elderly):

• Any significant neurologic disease other than dementia disorders, such as Huntington's disease, normal pressure hydrocephalus, brain tumor, seizure disorder, subdural hematoma, multiple sclerosis, or history of significant head trauma followed by persistent neurologic defaults or known structural brain abnormalities.
• Major depression as described in DSM-IV.
• History of schizophrenia or other recurrent psychotic disorder
• History of alcohol or substance abuse or dependence within the past 5 years

Community sample