The traditional “amyloid hunting” era of Alzheimer’s disease (AD) drug development is yielding to a more nuanced, functional approach. A major new review by Pini et al. (2026) argues that before neurons die, they disconnect—or paradoxically, overconnect. This “Hyper-Hypo” brain pattern—where early hippocampal hyperactivity (a panicked attempt to compensate) eventually collapses into widespread silence—is the new therapeutic target.
The authors propose a “connectome-first” repurposing strategy. Instead of waiting for novel compounds, they identify existing FDA-approved drugs—specifically anticonvulsants, antidiabetics, and gut-targeted antibiotics—that can stabilize these specific network disruptions. By dampening the “electrical noise” of early AD with epilepsy drugs or fueling the “energy-starved” brain with diabetes drugs, we might preserve the functional architecture of the mind even as pathology accumulates. This represents a pivot from clearing “trash” (plaques) to maintaining “signal” (connectivity).
Context: Department of Neuroscience, University of Padova, Italy; Chiesi Farmaceutici; Lausanne University Hospital, Switzerland. Published in Ageing Research Reviews.
Impact Evaluation:
Open Source Research Paper: Targeting brain connectivity in Alzheimer’s disease with repurposed drugs
The impact score of this journal is 12.4–13.1 (Impact Factor), evaluated against a typical high-end range of 0–60+ for top general science. Therefore, this is a High impact journal, ranking Q1 in Geriatrics & Gerontology and widely cited in translational neuroscience.
Part 2: The Biohacker Analysis
Study Design Specifications
- Type: Comprehensive Systematic Review (Meta-synthesis of preclinical and clinical data).
- Subjects: Synthesizes data from Transgenic AD mice (APP/PS1, 5XFAD), Rats, and Human Clinical Trials (MCI and AD cohorts).
- Lifespan Data: Cites lifespan extension in Drosophila (Levetiracetam) and improved “brain aging” metrics in Non-Human Primates (Metformin slowed brain aging by ~6 years).
- Key Interventions: Levetiracetam (Keppra), Lamotrigine, Rifaximin, Metformin, Semaglutide, Infliximab.
Mechanistic Deep Dive
The authors abandon the simple “amyloid cascade” for a “network failure” model.
- The Excitotoxicity Brake (Levetiracetam/Lamotrigine): Early AD is defined by hyperactivity in the hippocampus (DG/CA3 regions). This “noise” blocks new memory formation. Levetiracetam binds to SV2A (Synaptic Vesicle protein 2A), reducing neurotransmitter release probability just enough to dampen this noise without sedating the patient.
- The Metabolic Fuel (GLP-1s/Metformin): AD brains are insulin-resistant (Type 3 Diabetes). GLP-1 agonists (Semaglutide) reduce neuroinflammation (shifting microglia from M1 pro-inflammatory to M2 reparative states) and improve glucose transport, essential for maintaining the high-energy demands of long-range brain connectivity.
- The Gut-Brain Firewall (Rifaximin): Systemic inflammation drives neurodegeneration. Rifaximin acts as a “firewall” in the gut, reducing the production of ammonia and endotoxins (LPS) that otherwise leak into the bloodstream and trigger brain inflammation (inflammaging).
Novelty
- The “Hyper-Hypo” Timeline: Explicitly mapping drug classes to disease stages. Anticonvulsants are useless/harmful in late-stage AD (hypo-connected) but critical in early-stage (hyper-connected) MCI.
- Connectivity as a Biomarker: Proposing fMRI functional connectivity (DMN stabilization) as a faster, cheaper surrogate endpoint for longevity trials than cognitive decline.
Critical Limitations
- Translational Gap: The “connectivity” improvements in mice (e.g., via fMRI/EEG) often fail to translate to “activities of daily living” benefits in humans (as seen in the failed HOPE4MCI and EVOKE trials referenced).
- Timing Dilemma: The “Hyper-Hypo” switch happens at different times for different patients. Administering an anticonvulsant too late (during the hypo phase) could theoretically worsen cognition by dampening already weak signals.
- Review Nature: This is a synthesis of existing data, not a new experimental validation. It relies heavily on post-hoc analyses of failed trials (e.g., finding benefits only in APOE4 non-carriers).
Part 3: Actionable Intelligence
1. Levetiracetam (Keppra) – The Neural Stabilizer
-
Translational Protocol:
-
Goal: Reduce hippocampal hyperactivity (MCI/Early Prevention).
-
HED: The effective dose in the AGB101 trial was 60 mg – 125 mg daily (Extended Release). Note: This is a “micro-dose.” Standard epilepsy dosing is 1000–3000 mg.
-
Safety (Phase I/IV): High safety margin at micro-doses. Common side effects at standard doses include irritability (“Kepprage”) and somnolence, but these are rare at <125 mg.
-
Biomarker Verification:
-
Efficacy: fMRI (reduced hippocampal activation during memory tasks) is the gold standard but inaccessible.
-
Surrogate: Cognitive testing focusing on Pattern Separation (differentiating similar memories).
-
Population Applicability:
-
Contraindication: History of depression or mood disorders (risk of suicidal ideation, even at low doses). Kidney impairment (requires dose adjustment).
2. Rifaximin (Xifaxan) – The Gut Shield
-
Translational Protocol:
-
Goal: Reduce “leaky gut” derived neuroinflammation.
-
HED: 550 mg BID (twice daily). This is the standard hepatic encephalopathy dose.
-
Safety: Non-absorbable (<0.4% systemic absorption). extremely safe profile. No significant CYP450 interactions.
-
Feasibility & ROI:
-
Cost: Prohibitively high in the US (~$2,000+/month) without insurance approval for IBS/Liver disease. Sourcing via generic international markets is common but carries purity risks.
-
Biomarker Verification:
-
Efficacy: Reduction in Serum Neurofilament Light (NfL) and hsCRP.
-
Gut: Increase in Firmicutes abundance in stool analysis.
3. Semaglutide (Rybelsus/Wegovy) – The Metabolic Shield
-
Translational Protocol:
-
Goal: Neuroprotection via anti-inflammation and insulin sensitization.
-
HED: 14 mg oral daily or 2.4 mg subcutaneous weekly.
-
Safety: FDA Black Box warning for Thyroid C-cell tumors (rodents). Risk of pancreatitis and significant muscle mass loss (sarcopenia) if protein intake isn’t managed.
-
Interactions: Delays gastric emptying—can alter absorption of oral longevity supplements.
Part 4: The Strategic FAQ
Q1: If early AD is “hyperactive,” does that mean caffeine or modafinil accelerates brain aging?
A: [Confidence: Medium] Potentially. The “hyperactivity” discussed here is aberrant, non-functional firing (noise), not functional arousal. However, stimulants that increase glutamate release in an already hyperexcitable hippocampus (MCI stage) could theoretically exacerbate excitotoxicity.
Q2: Can I stack Rapamycin and GLP-1 agonists (Semaglutide)?
A: [Confidence: High] Proceed with Caution. There is a mechanistic conflict. Rapamycin is an mTOR inhibitor. GLP-1 agonists have been shown to stimulate mTOR signaling in specific hypothalamic nuclei (VMH) to regulate appetite and metabolism. While they may work synergistically on glucose, their opposing effects on mTORC1 require careful dosing schedules (e.g., cycling).
Q3: Is Levetiracetam safe for healthy biohackers seeking “preventative” maintenance?
A: [Confidence: Low] No data supports this. Levetiracetam dampens synaptic vesicle release (SV2A). In a healthy brain with optimal excitation/inhibition balance, this could blunt cognitive sharpness, reaction time, and mood. It is a “fix” for a broken circuit, not a “boost” for a healthy one.
Q4: Does Metformin actually cross the Blood-Brain Barrier (BBB)?
A: [Confidence: High] Yes, but slowly. It accumulates in the CSF. However, its primary neuroprotective benefit might be peripheral—improving systemic insulin sensitivity and reducing vascular inflammation—rather than direct neuronal action.
Q5: Why recommend Rifaximin over a probiotic?
A: [Confidence: High] Precision. Probiotics add bacteria to a chaotic ecosystem. Rifaximin acts as a “weeding” agent, specifically reducing ammonia-producing and pro-inflammatory bacteria without nuking the entire microbiome (unlike broad-spectrum antibiotics), thereby lowering the toxic load on the brain.
Q6: Did the EVOKE trial failure kill the GLP-1 for Alzheimer’s hypothesis?
A: [Confidence: Medium] No. The text notes that while clinical outcomes were missed, biomarkers moved in the right direction. This suggests the timing was likely too late (damage already done) or the duration was too short to reverse decades of pathology.
Q7: Is Infliximab (anti-TNF) a viable longevity strategy?
A: [Confidence: High] No. The immunosuppression risks (tuberculosis reactivation, serious infection) vastly outweigh the potential preventative benefits for a healthy individual. It is a “break glass in case of emergency” drug, not a prophylactic.
Q8: How do I test if I have “Hippocampal Hyperactivity”?
A: [Confidence: High] You generally can’t without fMRI. However, early subtle signs include “pattern separation” deficits—e.g., remembering where you parked your car today vs. yesterday, or distinguishing between very similar conversations.
Q9: Does Levetiracetam cause “brain fog”?
A: [Confidence: High] Yes, it is a common side effect (>10% at standard doses). This is why the AGB101 trials focused on low-dose formulations (60-220mg), attempting to find a window that stabilizes the hippocampus without sedating the cortex.
Q10: What is the “Third Way” mentioned in the text?
A: [Confidence: High] The authors propose moving away from (1) Amyloid clearing and (2) Tau clearing, to (3) Connectome Stabilization. Treat the network dysfunction directly to maintain cognition, regardless of the protein aggregates present.