Randomized control pilot 12 week feeding trial to compare the preliminary effects of ketogenic diet (versus patient choice diet) on HIV-associated neurocognitive impairment. N = 20 (n = 10/10) randomized to diet condition. Pilot data necessary to evaluate the feasibility and determine initial data for primary outcomes in order to accurately determine needed sample size for larger clinical trial.

Outcomes: 1) cognition (NIH Toolkit), 2) cardiometabolic markers (insulin glucose, insulin resistance, markers of inflammation), and 3) neural activity (as determined by functional MRI..

Specific Aims: In medically stable, older (> 50 years) HIV patients with cognitive impairment, this study will:

Overall: Test the feasibility of recruitment, retention, and adherence to a ketogenic diet versus patient choice diet (PCD) intervention.

Primary: Compare the impact of a 12-week KD versus PCD on:

1. Cognitive function (at 12 weeks).

2. Inflammation (at 12 weeks).

3. Cardiometabolic profile (at 12 weeks).

4. Persistence of cognitive effects (at 18 weeks).

   Secondary: In a subsample (n = 5; experimental group):

5. Examine the changes in neural activity and neurocognitive functioning as demonstrated via magnetic resonance imaging (MRI), after a 12 week KD diet.

6. Assess the level of agreement between NIH Neurocognitive Tests and MRI in the assessment of neurocognitive functioning.

INNOVATION: Brain metabolism regulation and reduced neural inflammation via dietary approaches have only recently been explored in the treatment/management of progressive cognitive disorders in humans. Preliminary results in Alzheimer's and Parkinson's patients suggest that the use of ketones for brain fuel improves brain metabolism and cerebral perfusion as well as reduces accumulation of harmful brain proteins20 To date, the effects of a KD in cognitively impaired older HIV patients have not been explored. In addition, this study proposes the use of MRI testing to examine the longitudinal, physiologic changes in the brain of cognitively impaired, older HIV patients consuming a KD for 12 weeks. Moreover, this study will utilize the new FDA-approved Magnetom Prisma 3T MRI system to better visualize small-scale neural brain abnormalities.

Research Design and Methods Design/Setting: Using an experimental design, a total of 20 older (> 50 years), stable HIV participants (CD4+ lymphocyte count > 350 cells/mm3 for at least two years and prescribed their current cART regimen for at least six months) with self-identified "forgetfulness" and demonstrate cognitive impairment (score </= on Telephone Interview for Cognitive Status -Modified (TICS-M)) will be recruited from the 1917 clinic (data extracted/verified using 1917 Clinic database) and will be randomly assigned to either the ketogenic diet (KD)(< CHO 50 grams/day) (n = 10) or the patient choice diet (PCD) (no dietary restrictions/changes) (n = 10) group for 12 weeks. In addition, the first five (n = 5) eligible (i.e., no metal implants, claustrophobia) participants randomized to the experimental group will also be enrolled in an additional magnetic resonance imaging (MRI) longitudinal study to generate preliminary data regarding the neurological changes associated with cognitive improvement. Previous studies in Alzheimer's and Parkinson's disease patients have reported significant cognitive improvement at 28, 45, and 90 days of ketosis.

Exclusion Criteria: Current drug or alcohol abuse (> 3 drinks/day), past medical history of mental disorders (i.e., schizophrenia, bipolar), neural injury (i.e., cerebral vascular accident or traumatic brain injury), dementia, Parkinson's, diabetes mellitus, hearing impaired, or any other condition that contraindicates participation. Due to the inconclusive evidence regarding the influence of statin drugs on cognition, statin use will not be exclusionary.

Recruitment: Fliers that provide basic study and PI contact information will be posted in the lobby/exam rooms of the 1917 Clinic. A telephone screening interview will be completed for all interested individuals. During the screening, cognitive impairment will be determined via administration of the Telephone Interview for Cognitive Assessment-Modified (TICS-M). TICS-M is a 13-item tool for assessment of cognitive function in older adults. A score < 20 indicates cognitive impairment (equivalent to a score < 25 on the validated/widely used in-person cognitive assessment, Mini Mental Status Exam [MMSE]).

Randomization: Participants will be randomly assigned, using permutated block randomization, to either the KD or PCD group with equal size. The CCTS BERD methodologists will develop the permuted block randomization algorithm. Because this is a diet intervention, it is not possible for participants/study personnel to be blinded to group assignment. In order to minimize bias/ensure allocation concealment, the randomization schedule will include sequentially-numbered, opaque sealed envelopes and will be kept by Dr. Morrison.

Data Collection:

Demographics: Age, education, gender, marital/partner status, employment status/disability eligibility, SES (Medicaid, food stamps, p Public housing eligibility), years since HIV diagnosis (baseline assessment only).

Anthropometrics: Height, weight, BMI, waist and hip circumference assessed at baseline and week 1. Height (nearest 0.1 cm), waist circumference (nearest 0.1 cm), and hip circumference (nearest 0.1 cm) measured with Gulick tape measure. Weight (nearest .01 kg) assessed using a Tanita body composition analyzer BC418, Tanita Corp of America, Arlington Heights, IL. BMI will be calculated. Waist circumference will be measured at the umbilicus at end of inspiration. Hip measures will be evaluated at widest hip portion.

Cognition: Cognition will be determined via the well-validated NIH Cognition Toolbox Battery. The NIH Toolbox is a comprehensive computerized battery that assesses each major cognitive domain in ~30 minutes. This technology has been in design/beta testing for nearly a decade by world-renowned neuropsychologists and computer programmers. Scoring includes the following corrected (age, education, gender, race) scores: fluid cognition composite, crystalized cognition composite, overall cognitive function composite, and scores for each individual domain (i.e., executive function, attention, episodic memory, language, processing speed, and working memory). Impairment scores can be derived from each of the aforementioned scores.

Neuroimaging: All neuroimaging screening, data collection, and data analyses will be conducted by Dr. Jarred Younger. Approved participants will attend a baseline imaging session. All imaging will be performed on the UAB Civitan International Research Center's 3-Telsa Siemens Prisma scanner and 64-channel head/neck receiving coil. Participants will first complete a high-resolution structural scan for group registration. Participants will then complete the T2*-weighted blood oxygen level dependent (BOLD) functional scan. The functional scan will use a "resting-state" protocol in which the participants are instructed to close their eyes and rest. The 12-minute run will collect functional volumes every 2 seconds. Participants will be invited back to the scanner after completing the intervention, at which time they will repeat the entire scan sequence. Structural and functional data will be examined longitudinally to determine changes occurring as a result of the intervention. Changes in gray-matter structure will be examined using a high-dimensional warping (HDW) tensor-based morphometry approach. Functional data will be analyzed using a Fractional Amplitude of Low Frequency Fluctuations (FALFF) approach. The relative activity pattern will be contrasted in the pre-intervention and post-intervention scans to determine areas showing greater or lesser neural activity after the intervention diet. To help mitigate the loss of statistical power due to the low sample size in this preliminary analysis, all statistics will be run within-person with a voxel-wise statistical threshold of p < 0.005.

Oral Glucose Tolerance Test (OGTT): Screening consists of a glucose load of 75 g at baseline (time 0). At ~7 am, after a 12-h fast, a flexible intravenous catheter will be placed in an antecubital space. Blood samples will be collected at times 0, 10, 20, 30, 60, 90, and 120 minutes. Sera will be stored at -85oC until analyzed for glucose, insulin, and C-Peptide. Whole-body insulin sensitivity and β cell responsiveness will be derived via OGTT at baseline/week 12.

Analysis of Glucose, C-peptide, and Insulin: Concentrations of glucose, insulin, and C-peptide will be analyzed in the UAB Diabetes Research Center (Core Director: B. Gower, Primary Mentor). Glucose will be measured in 3 µL sera using the Glucose oxidase method on a Stanbio Sirrus analyzer (Stanbio Laboratory, Boerne, TX). Insulin will be assayed in 50-µL aliquots using immunofluorescence on a TOSOH AIA-II analyzer (TOSOH Corp, S. San Francisco, CA). C-peptide will be assayed in 20-µL aliquots using the TOSOH analyzer.

Lipids: Total, high density cholesterol (HDL) and triglycerides will be measured using SIRRUS analyzer (Stanbio Laboratory, Boerne, TX); low density lipoprotein will be calculated using the method of Friedewald.

Inflammation Markers: Assessed by immunoassay in fasting morning sera before and after the intervention. High-sensitivity C-reactive protein (CRP) will be assessed by turbidometric methods by using a SIRRUS analyzer (Stanbio Laboratory), with reagents obtained from Pointe Scientific, and TNF-a and IFNy, IL-1B, IL-6, IL-8, and IL-12p70 by using electrochemiluminescence (Meso Scale Discovery).

Urine Ketones: Presence of acetoacetic acid (ketones) assessed via Precision XTRA Ketone Monitoring System. All participants will monitor capillary glucose daily for first two weeks and then weekly. Results will be documented on log sheet and returned to study staff. Participants will be contacted via telephone on weeks 4 and 8 to remind of ketone log submission.

Procedures: Eligible participants will be randomized to either the experimental (n = 10) or control group (n = 10) and scheduled for baseline data collection. Laboratory, demographic, and anthropometric assessments will be completed UAB's Clinical Research Unit (CRU). Baseline data will include: demographic questionnaire, oral glucose tolerance test (OGTT), inflammatory assays and lipid panel as well as an anthropometric evaluation. In addition, participants will be provided training related to the assessment of urine ketones, including a visual aid/handout that displays the color rubric and corresponding numeric value. Even though all snacks and meals for the KD group will be provided, experimental group participants will receive nutritional counseling related to the KD and provided with handouts that highlight important components of a KD so that participants can remain adherent to the intervention even if eating in a restaurant or in other social settings. After completion of the aforementioned items, participants will be escorted to the UAB Hospital cafeteria and provided a voucher for breakfast prior to completion of cognition assessments. After breakfast completion, participants will return to the CRU for completion of the cognitive battery in a private conference room. Completion of the questionnaires will be completed on a laptop computer that is secured by PGP Whole Disk Encryption® software to ensure participant confidentiality/privacy. After completion of the cognitive battery, individuals randomized to the experimental group and willing/eligible to undergo MRI testing, will be escorted to the UAB Highlands radiology department. Delivery of meals/snacks (week 1) will begin the week following baseline data collection. Bi-weekly phone conferences will be conducted with all participants to maintain participant contact as well as to assess for intervention adherence, challenges, and answer questions. In addition, all participants will complete a urine ketone assessment on weeks 4 and 8 and report those findings during conference calls. Note: Participants prescribed medications known to cause abnormal urine color (i.e., azo dyes, nitrofurantoin, and riboflavin) or affect the results (levodopa, hydroxyquinoline , methyldopa/captopril) or who report difficulty distinguishing differences between colors will return to UAB's CRU for serum ketone analysis at weeks 4 and 8. In addition, a survey will be administered to the KD group to determine what participants liked and did not like about the diet (week 12). All participants will be compensated for their time as follows: 1) Baseline and post intervention (wk 12) laboratory and cognitive assessments, $50/visit (N = 20); 2) MRI completion (n = 5), $50/scan; 3) Ketone assessment (wks 4 and 8, n = 20), $10 each; and 4) Six-week post intervention cognitive assessment (wk 18), (N = 20), $25.

Dietary Intervention:

KD: Meals/snacks will be provided/prepared by the UAB Clinical Research Unit's Bionutrition Department, a unit that employs two Dietitians who are all registered through the Academy of Nutrition and Dietetics and licensed in their perspective states. Food will be delivered weekly via a courier in appropriate food transportation coolers. The UAB Bionutrition Department is has a wealth of experience in research feeding study implementation including studies requiring specialized meal plans. Meals will be matched for energy content to maintain current energy balance and consist of < 50 grams CHO/day. Daily CHO intake will primarily be derived from fresh vegetables.

PCD group: Participants in this group will be instructed to maintain their normal dietary patterns. To ensure control group participants are not electively consuming a ketogenic diet (< 50 g CHO/day) during the 12 week intervention, participants will complete an at-home urine ketone screen and report results to project coordinator or PI on weeks 4 and 8. If a ketogenic diet is electively being consumed by a member of PCD group, the participant will be allowed to complete the study; however, the participant's data won't be included in analyses.

Intervention Fidelity: Intervention fidelity will be assessed using strategies consistent across the literature. 31-34 35 A comprehensive manual will be developed detailing recruitment, retention, assessment, and standardized intervention protocols. Standardized checklists will be utilized to review procedures. When a checklist item is not met, reasons will be explored/re-training will occur. Each participant will be contacted by telephone biweekly to reinforce intervention concepts (based on group assignment), appointment reminders, answer questions, and address areas of concern. Intervention adherence will be assessed at weeks 4 and 8 via urine ketone analysis unless contraindicated which, in case, serum ketones will be analyzed.

Statistical Analysis:

Demographic, Anthropometrics and Baseline Characteristics: Demographic data (i.e., age, education, gender, marital/partner status, employment status, and years since HIV diagnosis; anthropometrics (i.e., height, weight, BMI, waist and hip circumference); and baseline measures including cognition (cognitive scores and impairment score), inflammatory markers (CRP, TNF-a, IFNy, IL-1B, IL-6, IL-8, and IL-12p70), cardiometabolic factors (OGTT, glucose, C-peptide, and insulin) will be reported for each subject. Comparisons between the diet groups will be conducted to assess the degree to which comparability of randomization was achieved.

Primary analysis: All the primary outcomes (impairment score, inflammatory markers and cardiometabolic factors) will be examined by an analysis of covariance (ANCOVA) including the change from baseline to completion of diet treatment as dependent variable, study group as independent variable, and baseline measures as a covariate, to evaluate the effects of a 12 week KD versus PCD. The lasting cognitive effects of a 12 week KD versus PCD six weeks post intervention completion will be examine by an ANCOVA including the change of impairment score form 12 week to 18 week as dependent variable, study group as independent variable, and impairment score at 12 week as a covariate.

Secondary analysis: The changes in neural activity and neurocognitive functioning as demonstrated via MRI in the subsample of KD group will be examined by a paired t test. Spearman R correlation will be conducted to assess level of agreement between NIH Neurocognitive Tests and MRI in the assessment of neurocognitive function in medically stable HIV patients aged > 50 years.

Next steps: An important reason for conducting this pilot study is to evaluate the feasibility and determine initial data for the primary outcomes, in order to perform a sample size calculation for a larger trial. The mean and the standard error of the difference between the changes of KD group and PCD group will be used for the sample size calculation to reach an 80% power. In addition, the study protocol, data collection questionnaires, randomization procedure, recruitment and consent, and acceptability of intervention will be evaluated. Findings and the tested/improved protocol of this study will support an R21/R01 mechanism (i.e., PAR-15-280/PAR-15-282: Multidisciplinary Studies of HIV and Aging) that will include and expand upon following specific aim and hypotheses:

Specific aim: Compare the cognitive, cardiometabolic, and neural structural effects of a KD versus PCD in medically stable, older (> 50 years) HIV patients with cognitive impairment.

It is postulated that, in comparison to the control group, older (aged > 50 years) HIV positive, cognitively impaired individuals randomized to the KD will demonstrate the following:

Hypothesis 1: Improved cognition. Hypothesis 2: Decreased insulin, glucose, and systemic inflammation as well as improved insulin sensitivity.

Hypothesis 3: Heightened neural activity as observed by MRI.

Inclusion Criteria:

• Stable HIV participants (CD4+ lymphocyte count > 350 cells/mm3) for at least two years
• prescribed their current cART regimen for at least six months) with mild to moderate cognitive impairment

Exclusion Criteria:

• Drug/alcohol abuse
• History of Parkinsons
• Alzheimers, stroke or traumatic brain injury
• diabetes mellitus
• renal insufficiency

Three participants dropped out of the patient choice diet due to baseline screen failure.

Adverse events data were collected across the 18 week study.

One participant was admitted to hospital for asthma exacerbation.

Small sample size (pilot study), limited ability to monitor intervention fidelity, no reassessment until 12 weeks thus it is unclear as to when cognitive benefits emerged in persons randomized to the KD group.