Patient-Led Adaptive Experimentation (PLAE)
A Framework for Personalized Aging Interventions

Empowering Patients and Families with AI-Guided, Evidence-Informed Self-Experimentation in the Face of Incomplete Science

The Core Problem: A Poverty of Guided Experimentation

Despite trillions spent on aging-related diseases, most interventions remain one-size-fits-all, delayed by decade-long RCTs, and inaccessible until "proven." For mild cognitive impairment (MCI), sarcopenia, metabolic decline, or frailty, patients face:

• Incomplete science (small n, short duration, heterogeneous outcomes)
• Slow translation (e.g., 2025 probiotic trial: n=20, yet promising)
• No personalization (APOE4? Gut dysbiosis? Low brain creatine?)
• Zero-risk dogma → paralysis

Result: Millions wait. Decline accelerates. Hope fades.

The Proposed Solution: Patient-Led Adaptive Experimentation (PLAE)

PLAE = Structured, ethical, data-rich, AI-guided self-experimentation by informed patients and families, using off-label, GRAS (Generally Recognized as Safe), or investigational compounds in the face of incomplete evidence.

Old Model (RCT-Centric)	PLAE Model
Wait 10–15 years for Phase III	Start today with N-of-1 + aggregated insights
Physician as gatekeeper	Patient-family-AI triad as co-designers
Zero risk	Calculated, monitored, reversible risk
Static protocol	Adaptive, real-time optimization

Why PLAE? The Case in 5 Pillars

1. Scientific Necessity: Evidence Gaps Demand Action

• 2025 Nutrients study: n=20, yet p=0.003 for recall, p=0.03 for amyloid ↓
• Creatine: 16 RCTs (n=492) show SMD=0.28 in cognition — safe, cheap, available
• Lithium: Postmortem brains show 20–30% deficiency in MCI (Nature 2025)

Incomplete ≠ inaction. PLAE turns personal data into population signal.

2. Ethical Imperative: Autonomy in the Face of Decline

• Patients with MCI retain decision-making capacity (MoCA ≥19 in most)
• Families witness decline in real time — emotional cost > theoretical risk
• Right to try laws (37 US states) already allow terminal patients access

Why not early-stage decline?

PLAE respects dignity: “I’d rather risk mild GI upset than lose my memories.”

3. Technological Feasibility: AI as Co-Pilot

Modern tools enable safe, scalable self-experimentation:

Tool	Role in PLAE
Wearables (Oura, Apple Watch)	HRV, sleep, activity → early safety signals
At-home labs (SiPhox, InsideTracker)	Monthly blood: eGFR, lithium, CRP, BDNF
AI (like Grok)	Real-time protocol adjustment: “Your hs-CRP ↑ → reduce creatine 20%”
N-of-1 platforms (Quantified Self, StudyU)	Rigorous baseline → intervention → washout → replicate
Secure data lakes	Anonymized aggregation → crowd-sourced discovery

4. Risk is Not Zero — But It’s Known, Monitored, Reversible

Risk	Mitigation in PLAE
GI upset (probiotics)	Start low, titrate, stop if >3 days
Renal strain (creatine + lithium)	Monthly eGFR; stop if ↓>10%
Drug interactions	AI + pharmacist review; exclude high-risk meds
Over-optimism	Mandatory “failure log” — what didn’t work?

Risk accepted in daily life: Driving (1 in 100 crash risk), Statins (myalgia in 10%), Coffee (anxiety).
→ PLAE risk <1–5%, with escape hatches

5. Precedent Exists — We Just Need to Scale It

Example	Lesson
Ketogenic diet for epilepsy (1920s)	Parents experimented → standard of care
Lithium for mood (1949)	Discovered via patient observation
Quantified Self movement	1000s run N-of-1 trials on sleep, diet, cognition
Right to Try (2018)	Terminally ill access unapproved drugs

PLAE = Right to Try for the pre-symptomatic decline

The Word We Need: PLAE

(Patient-Led Adaptive Experimentation)

Alternatives considered:

• N-of-1 trials → too academic
• Biohacking → reckless connotation
• Personal science → vague
• Citizen science → understates medical stakes

PLAE conveys:

• Patient ownership
• Adaptive (AI-adjusted)
• Experimentation (honest about uncertainty)
• Ethical structure

How PLAE Works: A 12-Week MCI Example

Week	Action	AI Role
0	Baseline: MoCA, blood (BDNF, amyloid, lithium), stool, wearable	Generate personal risk score
1–4	Probiotic (LR5 10¹⁰ CFU)	Monitor bowel habits; adjust if bloating
5–8	Add creatine 5 g/day	Flag if HRV ↓ or weight ↑>2 kg
9–12	Add lithium orotate 300 μg	Alert if sodium ↓ or tremor
13	Washout + retest	Compare pre/post; suggest next cycle

→ Data uploaded → aggregated insights for others

The Vision: A Global PLAE Network

1. PLAE Registry (opt-in, HIPAA-compliant) → 100,000 MCI patients → faster than any RCT
2. AI "Experiment Engine" → Suggests protocols, flags risks, learns from failures
3. Clinician Oversight Light → Quarterly review, not gatekeeping
4. Regulatory Sandbox → FDA "PLAE Pathway" for GRAS compounds

Objections & Rebuttals

Objection	Rebuttal
"Too risky!"	Risk of inaction = guaranteed decline
"Not evidence-based"	PLAE creates evidence
"Placebo effect"	N-of-1 with washout controls for this
"Inequity"	Start with insured; scale via generics

Final Call

We don’t need perfect evidence.
We need PLAE.

For the 70-year-old with MoCA 23, fading recall, and a family watching in fear —
waiting is the riskiest choice.

Let patients and families, armed with a trusted HSP, RN AI, data, and courage,
lead the next leap in aging science.

PLAE: Because your brain can’t wait for 2035.