Several lines of evidence suggest that the neuropathological and clinical decline leading to Alzheimer disease (AD) begins years before patients develop the full AD clinical syndrome (NINCDS-ADRDA diagnostic criteria; McKhann et al, 1984). Mild memory complaints build gradually years before patients develop dementia. The neuropathological hallmarks of AD, "preclinical" neuritic plaques (Braak \& Braak, 1991) and neurofibrillary tangles (Price \& Morris, 1999), are also present years prior to clinical diagnosis. These abnormal protein deposits correlate strongly with cognitive decline.
Preclinical amyloid deposits may begin decades prior to dementia onset. In fact, diffuse plaques in non-demented elderly persons are associated with an accelerated age-related cortical cholinergic deficit, consistent with preclinical AD (Beach et al, 1997; Arai et al, 1999). Also consistent with a prolonged preclinical disease stage is our own work showing that position emission tomography (PET) measures of cerebral glucose metabolism vary according to AD genetic risk (apolipoprotein E-4 \[APOE-4\]) and predict cerebral metabolic and cognitive decline in people with mild cognitive complaints (age-associated memory impairment \[AAMI\]; Small et al, 2000). Such observations have stimulated interest in preclinical AD markers or biomarkers of brain aging that may assist in tracking treatments of AAMI and related conditions. New PET imaging methods now make it possible to provide in vivo measures of cerebral amyloid neuritic plaques (e.g., florbetapir-PET; Clarke et al, 2011) and tau neurofibrillary tangles (e.g. FDDNP-PET; Small et al, 2006, 2009).
Despite these previous research findings, clinical trials (including those using biomarkers as response measures) and subsequent treatment recommendations have been limited to patients with the full clinical dementia syndrome or mild cognitive impairment (MCI), a condition that increases the risk for developing dementia (Petersen et al, 2001). Cholinesterase inhibitors are currently the only drugs that have FDA clearance for treatment of AD, but previous studies (e.g., Ringman et al, 2005) suggest that other interventions, such as dietary or herbal supplements, may benefit cognition, and possibly interrupt the accumulation of abnormal amyloid protein deposits in the brain. For example, curcumin (diferulomethane), a low molecular weight molecule with antioxidant and anti-inflammatory activities that is derived from dietary spice, may have both cognitive-enhancing and anti-amyloid properties (Ringman et al, 2005; Yang et al, 2005).
To address such issues, the investigators propose to build upon our group's previous longitudinal brain imaging and genetic risk studies in people with age-related memory decline. Because previous studies suggest that curcumin may improve cognitive ability and prevent the build-up of age-associated plaques and tangles in the brain, the investigators will perform a double-blind, placebo-controlled trial of curcumin to test the following hypotheses:
1. People with age-related cognitive decline (i.e., MCI, AAMI or normal aging),, who receive curcumin 90 mg twice each day will show less evidence of cognitive decline (as measured with neuropsychological assessments) than those receiving placebo after 18 months.
2. People with age-related cognitive decline who receive an oral dose of curcumin 90 mg twice each day will show less build-up of plaques and tangles (as measured with FDDNP-PET imaging) than those receiving placebo after 18 months.
3. People with age-related cognitive decline, who receive curcumin 90 mg twice each day, will show decreased measures of inflammation in the blood compared with those receiving placebo after 18 months.
4. Cognitive change, FDDNP-PET measures, and treatment response will vary according to genotypes found to influence age at dementia onset (e.g., apolipoprotein E \[APOE\] TOMM40).
Because curcumin may alter inflammatory markers in the blood, the investigators will draw blood samples at baseline and at 18 months and freeze them for later analyses.
To test theses hypotheses, subjects with age-related cognitive decline will be enrolled (Crook et al, 1986; Petersen et al, 2001). Subjects will be randomized, using a double-blind design, to one of two treatment groups: curcumin (three 30 mg capsules twice each day) or placebo, and followed for 18 months. FDDNP-PET scanning will be performed at baseline and at 18 months. Magnetic resonance imaging (MRI) scans also will be performed for co-registration of PET and assistance in identifying regions of interest. Neuropsychological assessments will be performed at baseline, 6 months, 12 months and at the conclusion of the clinical trial (18 months). Blood will be drawn at baseline to perform genotyping.
