The 'Western diet' characterized by high fat feeding (HFF) is a risk factor for Alzheimer's disease (AD); however, little is known about mechanisms. Paradoxically, individuals who carry the AD risk gene APOE E4 cognitively improve after HFF and show different metabolic responses to fat intake. These responses include a more pronounced drop in systolic blood pressure and a decrease in cerebrospinal fluid glucose and lactate levels. This proposal will further examine the brain metabolic response to HFF using non-invasive imaging. Cerebral blood flow (CBF) as assessed by functional MRI is increasingly utilized to investigate brain function, given that CBF is tightly coupled to brain metabolism. Acute changes in CBF have been described in response to carbohydrate ingestion in regions important for learning and memory; however, no one has yet assessed CBF response to lipid. Here investigators propose to characterize CBF changes to a high fat drink in older adults. This study will include a time course assessment as well as analyses of global and regional change in CBF. These data would serve as feasibility pilot data for an R01 application which would include cognitive and peripheral metabolic data. These types of studies will provide a clearer picture of acute brain metabolic response to HFF in older adults, including the effects of APOE.

Overview: High fat feeding (HFF) is a risk factor for AD, but this risk factor is modulated by APOE genotype. Acute meal studies including lipid tolerance tests are increasingly being used to understand mechanisms by which diet affects disease processes. Investigators have consistently shown that E4 carriers respond differently to HFF in acute meal challenges; this study will further explore this phenomenon in older adults by assessing cerebral blood flow (CBF), given its tight coupling to brain metabolism. Here investigators propose to use arterial spin labeling (ASL) MRI as this non-invasive imaging technique can detect small changes in global and regional CBF, and acute ∆CBF in response to nutrients have been demonstrated in humans. With this project investigators will determine the most optimal time point for detecting post-lipid ∆CBF, as well as determine which regions of the brain respond to HFF. To our knowledge, this study would be the first to assess ∆CBF using ASL; these pilot data will then assist with future grant applications in this area.

Specific aims: 1. To determine which time point (1, 2, or 3 hours) produces the most robust ∆CBF after ingestion of a lipid drink in 30 older adults. 2. To determine which regions of the brain respond to HFF.

Methods: Participants: All study procedures will be submitted to the UW Institutional Review Board for approval. Investigators maintain a database of ~40-50 participants from the Meal and Memory study who have already agreed to be contacted for future studies; investigators will recruit 30 participants. Inclusion criteria are as follows: age 55 and older, able to read and understand English, able to give informed consent, and able to undergo MRI and ingest dairy products. All interested participants will undergo telephone screening to ensure they meet inclusion criteria. If they meet criteria, they will be mailed a reminder letter which will instruct them to fast the night before. They will also receive a copy of the consent form to look over, and a medical history form to fill out, including an MRI safety checklist which asks them about prior surgeries and whether they have any metal in their body.

Study visit: Participants will fast the night prior (12 hours prior to the visit time); medications and water are ok. They will arrive at the South Lake Union facility between 8 and 10 am. They will undergo informed consent in a quiet room from the study team (either the PI or the study coordinator) and given plenty of opportunities to ask questions. If they choose not to enroll in the study, they will be discharged from the unit and parking or transportation will be reimbursed. As a part of the consent process, investigators will ask participants to give optional written permission on the consent form to release their APOE genotype from the Meal and Memory Study, or other Alzheimer's research study they may have participated in. Investigators will not disclose this genotype to anyone outside the study team including the participant.

If the participant enrolls in the study and signs a consent form and a research HIPAA form, then the protocol will begin. Vitals will be assessed by the study team, including blood pressure, heart rate, and height/weight and BMI calculation. Participant will be escorted to the MRI machine by the study team, who will also bring down the lipid drink in a cooler. An MRI technician experienced with research MRI procedures will conduct the MRI portion. First a 15 minute ASL protocol will be run, then the participant will be given the heavy cream (100 mls, or just under half a cup, which contains 370 calories, 40.4 grams of total fat, 23.6 grams of saturated fat). Investigators will encourage the participant to drink the product over 5 minutes. Participants will then undergo 3 more 15-minute scans at 1, 2, and 3 hours post drink. It is possible that some of the acquisition times can be slightly longer than 15 minutes. After the last MRI image acquisition, lunch will be offered (up to $15 at a nearby restaurant), participants will receive a gift card, and parking or transportation will be reimbursed, and be discharged.

Alternative site: UWMC will run the last 2-4 participants to ensure the MRI protocols and sequences are similar, in order to prepare for Lipid MRI 2.0. Participants will check in at the DISC lab and vitals will be done there including height and weight for BMI calculation.

MRI protocol: The MRI is non-invasive with no need for contrast. In addition, investigators will not be performing a clinical MRI and these findings cannot be used to diagnose any brain condition. For the ASL protocol, we will use pseudo-continuous arterial spin labeling to measure CBF in ml/100g/min as a marker of perfusion for improved signal quality. In this approach, magnetically labeled arterial blood water serves as the endogenous contrast. Investigators will use a sequence (5.5 minutes) with long label duration = 1.8 s, long post-labeling delay = 2 s, with labeling offset = 25-30 mm, slices = 20, spatial resolution = 3.5×3.5×5 mm3, field of view = 240×240×100 mm3, SENSE-factor = 2, TR/TE= 5000/18 ms. Investigators will apply dual adiabatic background suppression pulses to minimize gray and white matter tissue contamination at TI = 2.05 and 3.25 s. Finally, investigators will acquire an equilibrium magnetization scan (M0, 1 minute), identical to the above scan, but with TR = 10,000 ms and no labeling or background suppression. For the last 8 subjects investigators will add a T2 multi-echo sequence (approximately 10 minutes) to detect changes in the appearance of the hypothalamus from before to after lipid ingestion.

Analysis: Analysis will be done under the supervision of Dr. Rane and the UW Diagnostic Imaging Sciences Center (DISC). Investigators will first apply motion correction to the arterial spin labeling images using FSL-MCFLIRT and register them to the M0 image. Then, investigators will perform a pair-wise subtraction between the control and null images and apply a two-compartment model to quantify CBF. Investigators will co-register the resulting CBF map to the T1 scan followed by a transformation to MNI space. The temporal regions such as the entorhinal cortex, temporal lobe, hippocampus as well as the posterior cingulate and lateral parietal lobules will be identified using standard Harvard-Oxford Cortical and Subcortical atlas. CBF values will be compared in these regions between the 3 groups. The region of interest approach can dilute effects that are smaller than the size of the region of interest. Therefore, we will also perform permutations testing in FSL to evaluate voxel-wise CBF differences between group. In order to ensure that the voxel-wise outcomes are not mere chance, investigators will impose strict family-wise error correction for multiple comparisons.

Statistics: From these images investigators will obtain measures on global blood flow as well as regional blood flow in pre-defined regions. Investigators will utilize standard statistical models to compare the 1, 2, and 3 hour ASL to the baseline ASL to ascertain what time point produces the most robust change result. If analyses of the first few subjects necessitate, time points can shift earlier or later, depending on what time point shows the maximum change in CBF-the primary outcome.

Inclusion Criteria:

• Previously completed the Meal and Memory study
• Age 55 and older
• Able to read and understand English
• Able to give informed consent
• Able to undergo MRI and ingest dairy products

Exclusion Criteria:

• Did not screen in the Meal and Memory study
• Allergic to or not able to ingest study ingredient (dairy)
• Metal in body or otherwise unable to lay still for an MRI

Through study completion, an average of 1 day

This study was extremely low risk and we did not expect any serious adverse events or deaths.