Aims:

A. To identify early functional and vascular brain imaging changes in individuals at increased risk of developing Alzheimer's disease (as measured by the presence of risk genes in their DNA, in particular APOE)

B. To understand how any such difference might relate to potential cardiovascular and inflammatory processes, such as disrupted cholesterol transport and elevated pro-inflammatory cytokine production

Hypotheses:

Functional MRI: The APOE epsilon4 group will show increased activation for scenes, but not other conditions, in PCC on all experimental tasks. This difference will be greatest in conditions where there is increased cognitive demand (e.g., longer delays between repeats of stimuli). During learning to discriminate scenes, the difference between carriers and non-carriers will lessen as cognitive 'effort' reduces (and the need for compensatory mechanisms diminishes). In terms of allele risk combination, we predict that, during spatial tasks, activity will show the following pattern: 'epsilon4, epsilon4' greater than 'epsilon3, epsilon4' greater than 'epsilon3, epsilon3' = 'epsilon2, epsilon4' greater than 'epsilon2, epsilon3'.

Vascular Imaging: We predict increased baseline CBF in PCC, elevated 1H MRS metabolite levels, and an increased task-induced increase in CBF. Based on findings in MCI, we also hypothesize a reduced vascular reactivity to carbon dioxide in PCC.

Inflammation and Immunity: Studies in AD predict that, if there are carrier/non-carrier differences, carriers will show higher levels of cholesterol, high sensitivity CRP and IL-6.

Sample:

Based on power calculations we wish to test male (n=125) and female (n=125) participants in their early twenties, who are MRI-compatible. We will use ALSPAC's whole genome data to sort these individuals into five separate APOE groups (n=50 in each group) based on genotype status (e.g., 'epsilon4, epsilon4', 'epsilon3, epsilon4','epsilon3, epsilon3', 'epsilon2, epsilon4' and 'epsilon2, epsilon3', ordered by risk of developing Alzheimer's disease from high to low). This will support two different analyses: (a) comparisons between carriers (n=100, 'epsilon4, epsilon4', 'epsilon3, epsilon4') and 'non-carriers (n=100,'epsilon3, epsilon3', 'epsilon2, epsilon3'), consistent with our earlier study, and (b) investigation of differing brain responses based on allele combination risk profile, in particular whether epsilon2, even in combination with epsilon4, is protective (n in each group = 50). The size of this sample will also allow us to look at our Alzheimer's disease risk genes (e.g., clusterin, ABCA7) in which functional neuroimaging hyper-activity has been documented, as well as undertake polygenic analyses using weighted multiple gene analysis.

Participants will be allocated to blinded research groups prior to visiting the Cardiff University Brain Research Imaging Centre (CUBRIC) on two separate occasions (lasting 3 hours). These sessions would involve imaging (both times), blood and urine sample collection (for additional genotyping (if required), measurement of cholesteol, insulin, high-senstivity CRP, prostaglandins and other inflammatory markers), an exercise challenge and some behavioural testing (see below).

Our participants will also complete validated questionnaires measuring physical activity, general health, and diet and eating styles. We will also measure BMI and blood pressure, and they will undertake an exercise challenge to measure overall fitness. This will be complemented by data from the cohort already collected by ALSPAC researchers. For example, as part of measuring potential cardiovascular health differences, we wish measures of height, weight, fat and lean mass, BMI, fitness, physical activity, resting BP and pulse, BP after exercise and various dietary and food questionnaires. We also require information about potential immune illnesses, such as asthma, eczema and diabetes and have asked for the child-based questionnaire 'Wellbeing of my teenage son/daughter (TB)' allowing us to obtain information on this. We have requested two IQ measures so we can co-vary out any IQ differences across groups, as well as measures of depression (MFQ), mental health (CIS-R).

Outcome variables:

The primary independent variable is regional task-related activity in fMRI. Predictor is risk gene status (known APOE, other genes less clear), with biological measures (e.g., cardiovascular health, cholesterol, cytokines) as potential mediators of the relationship between genotype and fMRI activity.