This protocol describes a randomized, placebo-controlled clinical trial to test the Microbiome-Metabolic-Inflammation Triad hypothesis in Alzheimer's disease (AD)kowalski2019 2019, Gut-brain axis in Alzheimervancassel2021 2021, Targeting the gut-brain axis: therapeutic strategies for Alzheimer. The hypothesis proposes that combined gut microbiome dysbiosis, metabolic dysfunction, and chronic neuroinflammation interact synergistically to drive AD progressioncatana2022 2022, Gut microbiota alterations in Alzheimervogt2018 2018, Gut microbiome alterations in Alzheimer, and that a combined intervention targeting all three pathways will demonstrate superior efficacy compared to single-target approachespistollato2020 2020, Role of gut microbiota and nutrients in amyloid formation and neurotransmission.
Primary Hypothesis: Combined intervention with GLP-1 agonist (liraglutide) plus multi-strain probiotic will show greater efficacy in improving cerebrospinal fluid (CSF) Alzheimer's biomarkers and cognitive outcomes compared to placebo in AD patients with metabolic syndromeMISSING:bassil2020gallagher2019 2019, Liraglutide crosses the blood-brain barrier in humans.
Mechanistic Hypothesis: The triad of microbiome dysbiosis, metabolic dysfunction, and neuroinflammation operates through interconnected pathwayschen2023 2023, Short-chain fatty acids and brain function in Alzheimerschroeder2020 2020, The gut-brain axis and Alzheimer where:
| Criterion | Requirement |
|---|---|
| Age | 60-85 years |
| Diagnosis | MCI due to AD or mild AD dementia (NIA-AA criteria)apa2018 2018, NIA-AA research framework: toward a biological definition of Alzheimer |
| Cognitive | MMSE score 18-26mcguinness2015 2015, MMSE as a screening tool for Alzheimer |
| Metabolic | Metabolic syndrome (≥3 criteria: waist circumference, triglycerides, HDL, blood pressure, fasting glucose)agorastos2021 2021, Metabolic syndrome and Alzheimer |
| BMI | > 28 kg/m² |
| Stable medications | Cholinesterase inhibitors or memantine allowed if stable ≥ 3 months |
| Criterion | Rationale |
|---|---|
| Active infection | Inflammation confounder |
| Autoimmune disease | Immune modulation |
| Antibiotic use < 3 months | Microbiome disruption |
| Probiotic/prebiotic use < 3 months | Baseline contamination |
| Type 1 diabetes | Metabolic confounder |
| Severe renal/hepatic disease | Safety |
| MRI contraindications | Safety |
| Active psychiatric disorder | Confounding |
Component 1: GLP-1 Agonist (Liraglutide)
Component 2: Multi-Strain Probiotic
Method: Shotgun metagenomic sequencing
| Parameter | Specification |
|---|---|
| Platform | Illumina NovaSeq 6000 |
| Depth | 10 Gb per sample |
| Analysis | Species-level abundance, functional gene families (MetaCyc), virulence factors |
Timepoints: Baseline, Week 12, Week 24
Key Outcomes:
Method: LC-MS/MS untargeted metabolomics
| Parameter | Specification |
|---|---|
| Platform | Q-TOF MS |
| Coverage | 2000+ metabolites |
| Focus | Short-chain fatty acids, bile acids, amino acids, lipids |
Timepoints: Baseline, Week 12, Week 24
Target Analytes:
Method: Multiplex immunoassay (Luminex)
| Parameter | Specification |
|---|---|
| Platform | Bio-Plex Pro Human Cytokine Panel |
| Volume | 1 mL CSF per draw |
| Storage | -80°C, protease inhibitors |
Target Cytokines:
Timepoints: Baseline, Week 24
| Outcome | Method | Timepoint | Expected Change |
|---|---|---|---|
| CSF Aβ42 | ELISA | Week 24 | Increase ≥ 20% vs placebo |
| CSF Total Tau | ELISA | Week 24 | Decrease ≥ 15% vs placebo |
| CSF Phospho-tau | ELISA | Week 24 | Decrease ≥ 20% vs placebo |
| ADAS-Cog13 | Cognitive testingadascog 1984, A new rating scale for Alzheimer | Week 24 | Improvement ≥ 3 points vs placebo |
| MMSE | Cognitive testingmcguinness2015 2015, MMSE as a screening tool for Alzheimer | Week 24 | Improvement ≥ 2 points vs placebo |
| Outcome | Method |
|---|---|
| CSF IL-6 | Luminex |
| CSF TNF-α | Luminex |
| Plasma SCFAs | LC-MS/MS |
| Microbiome diversity | Metagenomics |
| APOE genotype | PCR |
| HOMA-IR | Fasting glucose/insulin |
| BMI | Clinical measurement |
Assumptions:
Calculation:
Adjusted for dropout: 90 participants (45 per arm)
Intent-to-Treat (ITT) Population: All randomized participantsitt2010 2010, Intent-to-treat analysis in clinical trials
Statistical Methods:
Primary: Mixed-effects model for repeated measures (MMRM)
Sensitivity: Per-protocol analysis (participants with ≥80% adherence)
Multiple comparison adjustment: Bonferroni for primary outcomes
| Category | Assessment |
|---|---|
| GI symptoms | Daily diary, weekly assessment |
| Hypoglycemia | Fingerstick glucose, symptom diary |
| Injection site reactions | Visual inspection |
| Serious adverse events | Continuous monitoring |
10% severe GI adverse events: Pause enrollment, review
| Milestone | Timepoint |
|---|---|
| Protocol finalization | Month 0 |
| IRB approval | Month 1 |
| Participant recruitment | Months 2-8 |
| Intervention period | Months 3-9 |
| Follow-up assessments | Month 9 |
| Data lock | Month 10 |
| Primary analysis | Month 11 |
| Publication | Month 14 |
| Category | Cost (USD) |
|---|---|
| Personnel (PI, coordinators) | 350,000 |
| Laboratory (omics) | 150,000 |
| Study drug/placebo | 100,000 |
| Imaging (subset) | 50,000 |
| Administrative | 50,000 |
| Total | 700,000 |