Dementia Risk and Dynamic Response to Exercise: Methodology for an Acute Exericise Clinical Trial

Background: Exercise likely has numerous, meaningful benefits for brain and cognition. However, those benefits and their causes remain imprecisely defined, especially in the context of cognitive disorders associated with aging, such as Alzheimer disease (AD). If the brain does benefit from exercise it does so primarily through exposure to brief, acute exposures to exercise over a lifetime. Methods: The Dementia Risk and Dynamic Response to Exercise (DYNAMIC) clinical trial seeks to characterize the acute exercise response in cerebral perfusion, and circulating neurotrophic factors in older adults with and without the apolipoprotein e4 genotype (APOE4), the strongest genetic predictor or sporadic, late onset AD. DYNAMIC will enroll 60 older adults into a single moderate intensity bout of exercise intervention. We will measure pre- and post-exercise cerebral blood flow using arterial spin labeling, and neurotrophic factors. We expect that APOE4 carriers will have poor CBF regulation, i.e. slower return to baseline perfusion after exercise, and will demonstrate blunted neurotrophic response to exercise, with concentrations of neurotrophic factors positively correlating with CBF regulation. If exercise-induced changes in perfusion and circulating factors can be detected, DYNAMIC will contribute to our understanding of exercise-induced brain change and potential biomarker outcomes of exercise interventions. Results: Preliminary proof-of-concept findings on 7 older adults and 9 younger adults. We have found that this experimental method can capture CBF and neurotrophic response over a time course, and best practices following exercise. Conclusions: This methodology will provide important insight into acute exercise response and potential directions for clinical trial outcomes. ClinicalTrials.gov NCT04009629


Background
The number of Americans 65 years and older will double in size over the next 40 years. (1) Aging is often associate with increased cognitive decline.(2) Alzheimer's disease (AD) is of particular concern to the health care system, with an expected two-fold increase in prevalence over the next 30 years, and high direct and indirect costs of care. (3,4) We must find effective interventions to reduce the burden of cognitive decline and AD on our society.
There is evidence that risk of age-related cognitive decline, including mild cognitive impairment and dementia, can be reduced by health behavior interventions such as exercise. (5)(6)(7)(8)(9)(10)(11)(12) Although the literature is not conclusive, (13)(14)(15)(16) there is a growing consensus that common healthy behaviors, and especially exercise, support brain health and cognitive function. (17,18) A number of potential mechanisms may link exercise with brain health. Increased brain volume,(19) regional neurogenesis,(20) circulating neurotrophic factors,(21) and cerebrovascular reserve (i.e. capacity for response to a stimulus challenge) (22) all have been implicated as mediators of exercise benefits for the brain. If the brain does benefit from exercise it does so primarily through brief, "acute" exposures to exercise over a lifetime. In general, these studies have used passive "resting" conditions when measuring CBF and blood biomarkers, rather than employing tasks or experimental challenge (e.g. exercise, . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint 5 task-based fMRI, neuropsychological test) meant to approximate the ecological stressors of daily life. Our own work and others' demonstrates the importance of measuring the response to challenge, especially to exercise. (22,(35)(36)(37)(38)(39)(40)(41)(42) For example, using transcranial Doppler (TCD), we recently described the dynamic change in a proxy measure of CBF with onset of aerobic exercise in young and older individuals.(35) Our work demonstrates that older adults have a noticeable blunting of CBF during a dynamic condition like exercise that was less appreciably different from younger adults during a static, resting, measure.
There is evidence that the Apolipoprotein epslion4 (APOE4) single nucleotide polymorphism, the strongest genetic risk factor for late-onset, sporadic AD, (43) leads to modified neurovascular coupling, a leaky blood-brain barrier, angiopathy, and disrupted nutrient transport. (44) APOE appears key to maintaining cerebrovascular integrity independent of β amyloid deposition.(45) APOE4 carriers also demonstrate altered (46-48) and generally lower resting CBF especially in regions associated with AD-related change.(29, 49-53) Because exercise has such a strong and reliable benefit for the vascular system APOE4 carriers, who are at greater vascular risk than non-carriers, (12)  is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint 6 decline. Our project will extend the prior work by directly assessing the relationship of these factors in response to an acute exercise challenge.
We set out to quantify the CBF response to exercise which has the potential to be a valuable measure of cerebrovascular health.(22) This manuscript details the clinical trial methodology and proof-of-concept preliminary data of the Dementia Risk and Dynamic Response to Exercise study (DYNAMIC -NCT04009629). The scientific premise underlying this project is that CBF and blood-based biomarkers such as VEGF, BDNF, and Insulin-like Growth Factor 1 (IGF1) are interrelated mechanisms driving chronic aerobic exercise effects on brain health and cognition. (9,20,63) Our single visit clinical trial seeks to characterize the relationship of APOE4 carrier status with CBF and blood-based biomarkers of brain health. We capture dynamic fluctuations in CBF and blood-based biomarkers in a time-sensitive manner before and after an acute bout of moderate intensity aerobic exercise. Our working hypothesis is that individuals at genetic risk for AD have poor CBF regulation and altered neurotrophic response and that this methodological approach will inform future clinical trial biomarker protocols in the future. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint 7

Summary of design
DYNAMIC is a single site, non-randomized, prospectively enrolling trial testing APOE4related response differences to a single, 15-minute bout of moderate intensity aerobic exercise.
The study plans to enroll 60 older adults (>=65 years), approximately balanced for E4 carriage, and up to 20 younger individuals to serve as normative cohort. We hypothesize that APOE4 carriers will have poor CBF regulation, i.e. slower return to baseline perfusion (reduced area under the curve) after exercise, and will demonstrate blunted neurotrophic response to exercise, with concentrations of neurotrophic factors positively correlating with CBF regulation. We will also explore the relationship of the CBF and neurotrophic responses to cognitive performance.

Outcomes
As a registered clinical trial, we have identified a single primary outcome and several secondary outcomes of interest. Our primary outcome of interest is global CBF area under the curve (AUC) of the cumulative cerebral blood flow before and after our exercise intervention.

Recruitment and Eligibility
We significantly reduce participant burden by heavily leveraging the infrastructure of the Clinical Cohort 95% already have APOE genotype determined from prior study activities and have consented to share that information with collaborating studies such as DYNAMIC.
Genotype is not disclosed to the participant. Study staff with direct participant contact remain blinded to genotype. Because APOE4 does not have equal penetrance in the Clinical Cohort, the KU ADC continually monitors enrollment rates for DYNAMIC based on sex and E4 carriage and an unblinded study team member not involved in recruitment, consent, or study visit execution reviews and provides enriched contact lists to blinded staff to support balanced participation. Participants who have not previously had their E4 characterized consent to have genotyping performed for the purpose of the study. We exclude individuals with an E2 allele. We make efforts to preferentially match APOE4 genotype groups based on sex.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint Procedures Participants attend a single study visit. Because our focus is on assessment of dynamic time-and intervention-related changes in our outcomes, precise study timing and short transitions between study activities is critical. Our procedures have been planned to minimize waiting and transition time between study events.
Participants first change into provided magnetic resonance imaging (MRI) compatible clothes (scrubs) and remove all MRI-incompatible dental appliances, jewelry, etc. The exercise bicycle ergometer (Corival, Lode B.V., www.lode.nl) is adjusted so that the knee achieves near but not complete extension. Participants practice pedaling between 60-70 rpm until they report feeling comfortable with the movement and equipment. Imaging data are collected with a 3 Tesla whole-body scanner (Siemens Skyra, Erlangen, Germany) fitted with a 20-channel head and neck receiver coil. The MRI session is split into two parts: pre-exercise, and post-exercise. Each portion of the session begins with automated scout image acquisition and shimming procedures to optimize field homogeneity.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020. Participants return to the testing room. An optical heart rate sensor (OH1 Polar Electro, Inc. polar.com) is secured to the forearm with self-adhering wrap. Blood pressure is taken again.
Then, a flexible intravenous catheter is placed and 10mL of blood is collected in vacutainer tubes containing EDTA as an anti-coagulant. If the genotype is not available, an additional 3mL of blood is collected in a single vacutainer tube containing acid citrate dextrose and stored for future genotyping.
Participants then remount the cycle ergometer and begin a 5-minute warm up. During the initial 5 minutes, study staff gradually increase resistance with a goal of achieving the target heart rate of 45-55% of HR reserve in minutes 4 and 5. Participants begin pedaling at 60-70 rpm and a resistance based on an age-dependent decision algorithm (Figure 1). The 15-minute aerobic exercise bout begins immediately following the warm-up. Study staff check heart rate . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint every 1-minute and adjust cycle resistance to maintain the heart rate in the target zone. After 15 minutes of exercise resistance is reduced to 10W and participants pedal at a self-selected cadence for 3 minutes to cool down and drink 100mL of water to reduce potential perspirationrelated changes in blood volume. Post-exercise vitals are taken immediately at the beginning of the cool down period. An additional 10mL of blood is collected during cool down. The heart rate monitor is removed, and the participant is quickly escorted back to the MRI room.
Once back in the MRI room, the same preparatory procedures for MRI are repeated and 4 consecutive pCASL sequences are acquired, yielding 23 minutes and 12 seconds of postexercise pCASL data. Finally, the participant is escorted back to the testing room where vitals and 10mL of blood are taken one more time, and neuropsychological testing is repeated.
Participants are compensated $100 upon completion of the visit.

Blood Processing and Assessment
We optimized sample collection and processing procedures for accurate measurement of plasma neurotrophins in 5 samples independent of this study. As demonstrated in our proofof-concept results, when platelets remain in a blood sample, a freeze thaw can greatly increase the concentration of such factors and may not accurately reflect levels that were circulating in the plasma at the time of acute exercise. Consequently, our protocol emphasizes immediate processing. Our optimized protocol is as follows: plasma is generated immediately upon collection by centrifugation by processing at 1500 RCF (g) (2800 RPM) at 4°C for ten minutes.
Platelet-rich plasma is then centrifuged in four, 1.5mL aliquots at 1700 x g (4500 RPM) (4C) for 15 minutes. The resulting platelet-poor plasma is separated from the pellet and snap frozen in liquid nitrogen until stored at -80 C at the end of the visit.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Imaging Analysis
Planned pCASL data analyses include using the USC Laboratory of Functional MRI Technology CBF Preprocess and Quantify packages, (75,76) and Statistical Parametric Mapping CAT12 package. (77) We motion correct labeled and control pCASL images separately, calculate the CBF timeseries via simple subtraction, and coregister with the T 1weighted anatomical image. Mean CBF in the entre gray matter mask is extracted from each label-control subtraction image in the timeseries in units of mL*100g tissue -1 *min -1 . Imaging analysis may change as the field improves methods.

Cognitive Test Assessment
Cognitive assessments scores are calculated automatically by the NIH Toolbox software (https://www.healthmeasures.net/). We plan to use change in T-Score for each domain which is age, education, gender, and ethno-racial identification corrected and provides a score based on a normative mean of 50 with a standard deviation of 10.

Data Collection
All data are collected and organized in a custom designed REDCap(78) database.
Project access is role based. APOE4 genotypes are kept in a separate database and the linking list is kept by a designated, unblinded investigator.

Sample Size
To our knowledge, there are currently no peer-reviewed reports of genotype-based CBF differences in response to acute exercise. However, perfusion measures in genetic risk for AD (APOE4) have been performed previously and can form the basis of a reasonable power analysis. Two prior cross-sectional estimates of the relationship of perfusion and E4 carriage . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint 13 have delivered similar effect sizes (d=1.0).(48, 49) Given this effect size, we expect to be able to discern differences based on APOE4 in a sample of 60 older adults.

Safety
Adverse Events. Adverse events are defined as any untoward medical occurrence in study participants, which does not necessarily have a causal relationship with the study treatment.

Response to SARS-CoV-2
Data collection began prior to the SARS-CoV-2 novel coronavirus pandemic, was paused between March 11 and June 1, 2020, and resumed with additional safety procedures.
The blood processing centrifuge and staff member were moved to a separate, nearby testing room to allow for physical distancing. All participants are screened via telephone a maximum of 72 hours before the day of visit, and are screened again (including temperature) outside the imaging facility on the day of the visit. All staff wear level 1 surgical masks, gloves, and face . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020.

Proof-of-Concept Results
To date, we have enrolled 16 participants into the study. Demographics are provided in Table 1. To demonstrate proof-of-concept, we provide preliminary analyses and comparisons of individuals above and below 65 years of age. E4 carriage has not been unmasked to date and is not included in this report. There were more women in the older adult group. Figure 2 depicts the study flow and average time for each study event. Figure 3 demonstrates our ability to capture dynamic blood flow changes post-exercise. The area under the CBF curve ( Figure 3A) is greater in the younger individuals. Additionally, even when we normalize CBF to the average of the second pre-exercise pCASL sequence (a stabilized resting CBF), the area under the curve of younger adults remains greater. Notably, these dynamic changes do not appear to be heavily driven by blood pressure which is plotted underneath the CBF curves.
We also provide evidence for our decision to process blood collection immediately. In plasma processed with one centrifugation, mean plasma levels of BDNF were 17,506 ± 4031 pg/mL. Duplicate samples that underwent an immediate second centrifugation to remove the platelet pellet, followed by an immediate snap-freeze, were measured as having mean levels to 29 ±10.6 pg/mL. A delay of 15 minutes prior to each centrifugation step, which may allow increased time for platelet release when activated by the shear stress of centrifugation, increased levels to 245 ±108 pg/mL. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Adverse Events
To date there has been one adverse event, nausea, upon IV placement. Symptoms resolved with a light snack and rest.

Discussion
We have designed and implemented a single visit clinical trial to test the effect of APOE4 carriage on brain blood flow response to an acute, 15-minute, moderate intensity exercise challenge. We have also refined optimal blood collection and processing procedures to characterize circulating, bio-available neurotrophic responses to exercise. We expect that our strict and time sensitive protocol will allow us to identify CBF and blood-based neurotrophic responses that are obscured during typical resting conditions. We will also explore whether CBF or neurotrophic responses are related to performance changes on neuropsychological tests. To be clear, we do not expect measurable vascularization, neurogenesis, or other benefits immediately following exercise. Nor do we suggest that any neurotrophic increases are causal of ad hoc cognitive change or CBF response. Rather, we seek to index the transient changes and relationships that are hypothesized to mediate these benefits with chronic exposure to exercise.
Our proof-of-concept results suggest that we can capture a dynamic CBF recovery response following exercise that can differentiate groups. We found that older adults have lower overall blood flow and slower return to resting blood flow as measured by area under the CBF curve. Our blood processing results provide a clear case for immediate post-processing to identify the bio-available circulating neurotrophic factors following exercise. Optimization of preand post-processing techniques revealed higher BDNF levels in plasma collected without removal of platelets compared to plasma where platelets had been removed. In addition, sitting time after the centrifugation step to remove platelets, which may result in platelet activation, also affected neurotrophic factor levels.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint Despite the many strengths and carefully constructed protocol, there are notable limitations. First, our pre-post exercise measures are proxies of CBF during exercise as we are not imaging during exercise. Some protocols for MRI during exercise are beginning to emerge, though concerns about motion artifact remain. CBF can also be measured using contrastenhanced MRI, TCD, or positron emission tomography (PET). TCD has the advantages of temporal resolution and ease of use during exercise, but can only index blood velocity. Xenon PET imaging can produce whole brain CBF but requires a radioactive isotope. pCASL provides both the whole brain spatial resolution, and potentially improved temporal resolution compared to PET, using only magnetically labeled arterial blood water, an endogenous tracer that is highly reproducible.(70) Additionally, optimal imaging parameters to capture CBF response to acute exercise using have been investigated previously.(79) Pseudo-continuous ASL sequences have improved signal-to-noise ratio while maintaining high labeling efficiency. (80) Common to all ASL sequences are pairs of images with and without labeling, that allow for CBF quantification.
Typically, the subtraction of these pairs is averaged over a sequence to calculate CBF.
However, the pairs viewed as a timeseries, may also capture information about transient CBF changes, for example following exercise, as we have done here. We believe this timecourse will provide important additional information about dynamic responsiveness, or cerebrovascular reserve, of the system. Despite these limitations they DYNAMIC study offers a blueprint for a unique and innovative methodology to capture acute exercise response in individuals at risk for Alzheimer's disease.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint 17 Aerobic exercise is among the most important and cost effective tools available for chronic disease management.(83) Our work along with others' strongly suggests that moderate-intensity aerobic exercise benefits brain health and cognition. (10,11,(84)(85)(86)(87) However, the field continue to struggle with adequate methods for capturing mechanistic drivers of exercise benefits on the brain. New protocols such as DYNAMIC are necessary to move forward our mechanistic explorations of exercise effects on brain health and cognition.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020.

Consent for publication Not applicable.
Availability of data and materials Data may be obtained by contacting the corresponding author. Data will not be released prior to the closing of the trial. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted August 25, 2020.  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2020. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2020. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2020. . https://doi.org/10.1101/2020.08.22.20179564 doi: medRxiv preprint