Bryan Johnson’s autoimmune gastritis exposes longevity science’s missing metric: biological control fidelity
Meta description: Bryan Johnson reports youthful performance across several organ systems yet has autoimmune gastritis. RF Safe examines longevity’s missing dimension: the fidelity of bioelectric, immune, mitochondrial, and redox control.
Bryan Johnson’s autoimmune-gastritis diagnosis should not be treated as a punchline or as proof that his longevity program failed.
It is much more scientifically interesting than that.
Johnson says that after six years of intensive measurement and intervention, his cardiovascular function, fertility, strength, and hormones largely operate at what he calls “elite 18-year-old levels.” He also acknowledges unresolved problems, including hearing loss and a brain he describes as anatomically 42. More recently, he disclosed autoimmune gastritis, persistent low ferritin, a history of autoimmune thyroid disease beginning at age 21, and a plan to sequence one million immune cells to identify the clonotypes attacking his stomach.
The media quickly connected the diagnosis with Johnson’s brief post wondering whether he had taken “this whole longevity thing too far.” But that one sentence does not demonstrate that Blueprint caused his disease, that he is abandoning the project, or even that the two subjects were intended to be causally connected. By his own account, his autoimmune architecture began decades before his present protocol.
The real lesson is more important:
A body can have youthful components, excellent biomarkers, and impressive physical outputs—and still run a control program that mistakes self for enemy.
That distinction reveals a major blind spot in modern longevity science.
We have become increasingly sophisticated at measuring the condition of biological parts. We remain far less capable of measuring whether those parts are coordinating with the correct timing, gain, phase, discrimination, and recovery.
Johnson may have substantially improved the state of many biological systems while an older, deeply embedded immune-control error remained active underneath them.
That is not a contradiction.
It is the difference between state age and fidelity age.
The new-engine problem
Imagine building a completely new high-performance engine.
The cylinders are perfect. The bearings are new. The oil is clean. Fuel quality is optimal. Every mechanical tolerance is measured to the micrometre.
Now shift the ignition timing.
You do not need to pour sand into the motor. You do not need to break a piston with a hammer. If combustion repeatedly occurs at the wrong point in the cycle, the engine can produce knock, excess heat, destructive pressure, bearing stress, and eventually its own mechanical failure.
Every individual component can be chemically and structurally “healthy” while the integrated system damages itself because its timing is wrong.
The human body is vastly more timing-dependent than an engine.
A molecule does not have one fixed biological meaning. Its meaning depends on:
- where it appears;
- when it appears;
- how long it remains;
- what preceded it;
- what other signals arrive with it;
- which structures are receptive;
- and whether the system returns cleanly to baseline.
Calcium can contribute to secretion, contraction, metabolism, differentiation, immune activation, adaptation, or cell death. Reactive oxygen species can function as controlled signals or become damaging oxidants. An inflammatory cytokine can be protective during infection and destructive when chronically misapplied to self-tissue.
The molecules can be correct while the chronology is wrong.
Chemistry supplies the components. Timing helps determine the instruction.
This is the central idea behind biological fidelity.
Autoimmune gastritis is not an iron-deficiency problem
Johnson is correct that his low ferritin is downstream of the autoimmune process rather than its initiating cause.
Autoimmune gastritis is an organ-specific immune disorder in which the gastric proton pump, H⁺/K⁺-ATPase, is a major autoantigen. Autoreactive T cells—particularly inflammatory TH1, TH17, and cytotoxic populations—participate in the destruction of acid-producing parietal cells. As those cells disappear, gastric acid and intrinsic-factor production decline. Iron deficiency can arise relatively early; vitamin B12 deficiency and pernicious anaemia can emerge later. Advanced disease also carries increased risks of gastric neuroendocrine tumours and adenocarcinoma, which is why specialist follow-up, deficiency management, and appropriate endoscopic surveillance matter.
Johnson describes the problem as a specific group of immune “soldiers” carrying keys that now unlock an attack against his stomach. Sequencing one million immune cells could help identify the clonotypes, transcriptional states, and lineage relationships involved. That is an important precision-immunology program.
But sequencing the soldiers does not necessarily explain the entire war.
A T-cell receptor sequence can reveal what a clone recognizes. It does not, by itself, reveal:
- why that clone escaped or overcame tolerance;
- what changed its activation threshold;
- why it became persistent;
- why regulatory mechanisms failed to restrain it;
- which metabolic state supports it;
- or which environmental conditions repeatedly reinforce it.
The key matters. So does the permission system that allows the key to turn.
Johnson wrote that his “genetic and immunological architecture made a mistake.” That is unusually close to the deeper systems question.
What physical architecture converted self into danger?
Longevity’s missing metric: Fidelity Age
Most longevity measurements belong to one of two categories.
They measure the condition of the parts, or they measure the outputs produced by those parts.
Those are valuable—but incomplete.
| Dimension | Central question | Typical measurements |
|---|---|---|
| State age | How young or damaged do the components appear? | epigenetic marks, inflammation, lipids, hormones, organ structure |
| Performance age | What output can the system produce? | VO₂ max, strength, fertility, cognition, glucose control |
| Fidelity age | How accurately does the system interpret perturbation and return the correct response? | discrimination, gain, phase, recovery, hysteresis, off-target activation |
A person can improve state age and performance age without fully correcting fidelity age.
An immune system can have adequate nutrients, high mitochondrial capacity, excellent glucose regulation, and youthful systemic biomarkers while retaining a deeply maladaptive decision rule:
self antigen⟶threat classification⟶persistent attack.\text{self antigen} \longrightarrow \text{threat classification} \longrightarrow \text{persistent attack}.
That is why a resting snapshot is not enough.
Two biological systems can have nearly identical resting measurements yet respond very differently to the same challenge. One returns smoothly to baseline. The other overshoots, oscillates irregularly, recruits off-target pathways, or remains trapped in an inflammatory state.
In The Embodied Cellular Transfer Function, I define low-fidelity biology as a decline in the reliability with which input is converted into the appropriate state trajectory, metabolic execution, feedback, and persistent update. It is not merely “more inflammation,” “more ROS,” or “more entropy.” It is degradation of the relationships that keep those processes coordinated.
The most revealing biological-age test might therefore not be another resting blood draw.
It may be an impulse-response test:
Apply a standardized challenge, record how the system interprets it, and measure how accurately it recovers.
That would expose the age of the controller rather than only the age of its parts.
The immune system is a precision timing machine
Immune recognition is often described as a lock-and-key process: a receptor recognizes an antigen, and the immune cell responds.
Real immune decisions are considerably more dynamic.
A T cell integrates antigen affinity, signal strength, pulse duration, co-stimulation, inhibitory signals, membrane potential, calcium dynamics, redox state, mitochondrial performance, cytokines, tissue context, and prior activation history.
Experimental work has shown that T cells can filter oscillatory inputs on a scale of minutes. Changing the temporal pattern of stimulation can alter the cellular response even when the overall amount of stimulation is similar.
Calcium is central to this process, but immune-cell calcium signaling needs to be described correctly. In T cells, the principal sustained calcium-entry machinery is the STIM–ORAI store-operated calcium-entry system. STIM proteins detect depletion of calcium in the endoplasmic reticulum and activate ORAI/CRAC channels at the plasma membrane. ORAI1 and STIM1 are recruited to the immunological synapse, and their local calcium events help drive transcriptional programs involving NFAT, NF-κB, AP-1, cytokine production, differentiation, and proliferation.
This requires an important refinement of the RF Safe S4–Mito–Spin model.
In the immune system, the S4 component may be Kv1.3—not a conventional voltage-gated calcium pore
Classical voltage-gated calcium channels are critical in neurons, muscle, endocrine cells, and other excitable tissues. But careful electrophysiological work found no evidence that conventional CaV proteins function as calcium-conducting channels in human or mouse T cells.
T cells do, however, express the voltage-gated potassium channel Kv1.3, which contains an S4 voltage-sensing domain. Kv1.3 helps repolarize the membrane and preserve the electrochemical driving force needed for calcium entry through ORAI channels. Autoreactive and repeatedly stimulated effector-memory T cells can become particularly dependent on Kv1.3, and suppressing this channel can reduce their calcium signaling, cytokine production, and proliferation in experimental autoimmune systems.
This yields a scientifically stronger proposed immune timing axis:
Kv1.3 membrane timing→STIM–ORAI calcium coding→mitochondrial-redox control→nuclear immune decision\boxed{ \text{Kv1.3 membrane timing} \rightarrow \text{STIM–ORAI calcium coding} \rightarrow \text{mitochondrial-redox control} \rightarrow \text{nuclear immune decision} }
Kv1.3 provides voltage-sensitive membrane control.
STIM–ORAI generates the major calcium signal.
Mitochondria shape calcium persistence, ATP availability, redox signaling, and recovery.
The nucleus integrates the resulting temporal pattern through calcium-, kinase-, and redox-sensitive transcriptional programs.
That is a plausible place to investigate immune bioelectric fidelity.
It does not mean that mistiming alone creates the antigen specificity of autoimmune gastritis. The H⁺/K⁺-ATPase-specific clone and the failure of immune tolerance remain indispensable parts of the disease.
The more precise hypothesis is:
Timing and redox state may influence whether an existing autoreactive clone remains silent, becomes activated, expands, acquires inflammatory memory, or resists being switched off.
That is a much narrower—and experimentally tractable—claim.
Three clocks must remain synchronized
An activated immune cell operates through at least three interlocking clocks.
The membrane clock
Ion channels and membrane potential determine whether calcium entry can begin and how long it can continue.
The mitochondrial clock
Mitochondria respond to calcium and metabolic demand, supply ATP, regulate redox state, and help determine whether activation remains controlled or becomes stressful.
The nuclear clock
Transcription factors integrate the frequency, duration, and combination of upstream signals, converting a transient event into cytokine production, differentiation, tolerance, exhaustion, memory, or persistent inflammatory identity.
The clocks do not need to be perfectly periodic. But their phase relationships must remain within a functional range.
When those relationships deteriorate, the immune system can produce:
- excessive gain;
- delayed shutoff;
- false danger signaling;
- impaired regulatory-T-cell function;
- inflammatory TH1 or TH17 bias;
- increased tissue damage;
- and self-reinforcing release of damage-associated molecular patterns.
That is what RF Safe means by bioelectric dissonance: not the presence of electricity, and not the mere artificiality of a waveform, but a measurable loss of task-relevant timing and causal coupling.
A waveform is biological noise only when it degrades a biological task in a particular receiver.
That receiver-specific definition matters because the same physical input can be neutral in one state, therapeutic in another, and disruptive in a third. Frequency, modulation, duty cycle, orientation, genotype, tissue architecture, endogenous phase, and prior adaptation can all change the result.
Where electromagnetic exposure enters—and where the evidence stops
There is currently no study demonstrating that Wi-Fi, cellular radiation, or another ambient electromagnetic exposure caused Bryan Johnson’s autoimmune gastritis.
There is no completed causal chain showing:
wireless exposure→H⁺/K⁺-ATPase autoimmunity→Johnson’s disease.\text{wireless exposure} \rightarrow \text{H⁺/K⁺-ATPase autoimmunity} \rightarrow \text{Johnson’s disease}.
That claim should not be made.
But it would be equally unscientific to declare that low-level electromagnetic fields cannot participate in biology merely because they do not produce substantial heating.
Three findings demonstrate why waveform-aware research is justified.
1. RF fields can produce highly specific biological effects under controlled conditions
The FDA-authorized TheraBionic P1 device exposes patients with advanced hepatocellular carcinoma to specific amplitude-modulated RF frequencies. The FDA states that these fields may stop cancer cells from dividing and warns that the device should not be used in patients receiving calcium-channel blockers. Experimental work has implicated Cav3.2 T-type calcium channels and calcium influx in the tumour-specific response.
This does not show that ambient Wi-Fi causes disease.
It establishes a more basic point:
Low-level, non-ionizing RF bioactivity can be frequency-, waveform-, receptor-, and cell-state-specific.
A field can be biologically active without functioning as a heater.
2. A defined EMF system can drive CYB5B-dependent calcium oscillations and gene expression
A 2026 Cell study developed an electromagnetic-field-inducible gene switch. An unbiased CRISPR screen identified the outer-mitochondrial-membrane protein CYB5B as an essential mediator, and activation depended on rhythmic calcium oscillations rather than a generic increase in bulk calcium. The system used a defined engineered low-frequency exposure and was designed for remote therapeutic control; it was not an experiment on Wi-Fi, cellular networks, or autoimmune disease.
Again, the significance is categorical rather than causal:
A field-to-mitochondrial-redox-to-calcium-to-transcription pathway can exist.
Whether environmental RF accesses CYB5B, a related interface, or no such interface at all remains to be tested under the actual environmental waveform.
3. Human response can be conditioned by receiver genotype
A randomized, double-blind study involving 34 CACNA1C-genotyped participants reported that a 3.6-GHz exposure altered NREM sleep-spindle centre frequency in T/C carriers of the noncoding variant rs7304986 but not in the matched T/T group. The experiment was acute, small, and not designed to establish disease or harm. Its significance is that the same nominal physical exposure was not transformed identically by every receiver.
This supports a core RF Safe principle:
biological trajectory=physical exposure⊗receiver architecture.\text{biological trajectory} = \text{physical exposure} \otimes \text{receiver architecture}.
SAR and temperature remain essential physical measurements. But they do not fully describe genotype, channel abundance, tissue geometry, endogenous phase, metabolic state, or adaptation history.
The honest evidence boundary
| What is established | What is scientifically plausible | What is not established |
|---|---|---|
| Autoimmune gastritis is an antigen-specific immune attack on gastric parietal-cell machinery | Timing, membrane potential, calcium dynamics, metabolism, and redox state can influence immune-cell activation and persistence | Ambient RF caused Bryan Johnson’s autoimmune gastritis |
| T cells decode temporal patterns and depend on ion-channel and mitochondrial coordination | Some time-structured fields could alter control fidelity in susceptible cells | Every artificial field is harmful |
| Specific EMF waveforms can activate defined biological pathways | Genotype, cell state, waveform, and endogenous phase could create vulnerable response windows | CYB5B is a universal Wi-Fi sensor |
| Oxidative and inflammatory states can influence immune tolerance | Chronic control noise could reinforce an already established autoreactive state | A single pathway explains every autoimmune disease |
The 2024 WHO-commissioned systematic review of RF exposure and oxidative-stress biomarkers rated the overall evidence as very low certainty and reported inconsistent results, with possible increases in some tissues and no or inconsistent effects in others. Melnick and colleagues subsequently argued that extensive study exclusions and subgroup fragmentation created much of that uncertainty. The appropriate scientific conclusion is not “RF definitely causes oxidative stress” or “RF is biologically inert.” It is that the present literature is methodologically inadequate to settle a waveform-, tissue-, and state-dependent question.
That uncertainty should produce better experiments—not categorical reassurance.
The RF Safe hypothesis: autoimmunity as a loss of control fidelity
The RF Safe position is not that electromagnetic exposure maps directly onto one disease.
It is not:
RF→autoimmune gastritis.\text{RF} \rightarrow \text{autoimmune gastritis}.
It is a conditional systems hypothesis:
genetic susceptibility×autoreactive clonotype×tissue antigen×immune history×timing and redox environment⟶probability of persistent self-attack.\begin{aligned} &\text{genetic susceptibility}\\ \times{}& \text{autoreactive clonotype}\\ \times{}& \text{tissue antigen}\\ \times{}& \text{immune history}\\ \times{}& \text{timing and redox environment}\\ \longrightarrow& \text{probability of persistent self-attack}. \end{aligned}
External electromagnetic fields would be only one possible modifier of the timing and redox environment.
Other fidelity stressors include:
- circadian disruption;
- inadequate or fragmented sleep;
- air pollution and hypoxia;
- inflammatory infections;
- chemical and mitochondrial toxicants;
- nutrient or electrolyte imbalance;
- chronic psychological and physiological stress;
- and persistent tissue injury.
These stressors do not have to share one receptor. They can enter the same recurrent control system at different points.
One may impair input discrimination.
Another may distort mitochondrial execution.
Another may prolong inflammatory feedback.
Another may make the system adapt to a chronically distorted condition.
The shared upstream property is not one molecule or one diagnosis. It is the progressive erosion of reliable coupling.
That is the proposed meta-disease state of low-fidelity biology.
Microplastics are matter pollution. Timing noise is control pollution.
Johnson has brought valuable attention to chemical exposures, microplastics, micronutrients, metabolic health, sleep, light, exercise, and other environmental variables.
Those are legitimate dimensions of health.
But there is a category error in assuming that the biological environment consists only of material inventories.
A chemical contaminant changes what is present.
A time-structured perturbation can change when a biological event occurs.
Matter pollution can be measured in particles, concentrations, and molecular burden. An electromagnetic exposome must also preserve:
- carrier frequency;
- modulation envelope;
- pulse repetition;
- duty cycle;
- peak-to-average structure;
- polarization and orientation;
- distance and near-field geometry;
- time of day;
- interaction with sleep and endogenous rhythms;
- and the receiver’s biological state.
A simple average-power value discards much of that temporal information.
This does not mean that every wireless waveform is “control pollution.” It means that a timing hypothesis cannot be evaluated with a measurement system that removes timing.
A control system cannot be protected by measuring only energy while ignoring the organization of that energy in time.
ROS is not simply damage—and more biophotons do not mean better feedback
Mitochondria occupy a central position in this theory because they couple calcium, ATP production, redox state, innate immune signaling, apoptosis, and recovery.
Controlled reactive oxygen species participate in normal signaling. Excessive or prolonged ROS can oxidize proteins, lipids, and nucleic acids, release danger signals, alter antigen presentation, and reinforce inflammation.
ROS-linked excited-state chemistry also produces ultra-weak photon emission, or UPE. In the ceLLM framework, the important variable is not merely how many photons are emitted. It is whether the spectral, spatial, and temporal emission pattern remains meaningfully coupled to the cellular event that generated it.
A stressed system could therefore display:
photon load↑whilephotonic fidelity↓.\text{photon load}\uparrow \quad\text{while}\quad \text{photonic fidelity}\downarrow.
More light could represent greater oxidative chemistry while carrying less reliable information about successful execution and recovery.
The proposed role of UPE in immune control remains hypothetical. But it produces a concrete scientific question: does the photon-event pattern predict or influence the cell’s next state beyond what is already explained by calcium, ROS, membrane potential, and mitochondrial function?
That is measurable.
The larger concept—informational oxidation—should also be understood precisely. It is not a new chemical species. It is the progressive corrosion of reliable biological relationships:
input→interpretation→execution→feedback→appropriate update.\text{input} \rightarrow \text{interpretation} \rightarrow \text{execution} \rightarrow \text{feedback} \rightarrow \text{appropriate update}.
When repeated compensation modifies the future transfer function, a cell can become increasingly adapted to the distorted environment that stressed it. The ceLLM paper calls the accumulated cost fidelity debt and the resulting maladaptive specialization somatic overfitting.
The experiment Bryan Johnson is uniquely positioned to run
Johnson’s diagnosis does not provide evidence that RF contributed to his disease.
It provides an opportunity to test the proposition under unusually controlled conditions.
He already has:
- extensive longitudinal data;
- a highly standardized lifestyle;
- advanced single-cell sequencing;
- organoid and organ-clone programs;
- sophisticated sleep and physiological monitoring;
- and the ability to conduct randomized, blinded N-of-1 experiments.
His own Blueprint site says that thousands of organ clones are being developed to test interventions against his biology without first exposing his body. That platform could be expanded into a serious receiver-aware bioelectromagnetics program.
1. Sequence the soldiers—but measure their timing
Single-cell TCR sequencing should be coupled to dynamic functional phenotyping.
Johnson’s immune cells could be challenged with:
- H⁺/K⁺-ATPase peptides;
- intrinsic-factor antigens;
- irrelevant self-antigens;
- microbial antigens;
- and standardized polyclonal controls.
For each clonotype, investigators could measure:
- calcium-wave onset, frequency, amplitude, and decay;
- Kv1.3 and KCa3.1 currents;
- STIM1/2–ORAI1 recruitment and CRAC activity;
- membrane potential;
- mitochondrial calcium and membrane potential;
- NADH/FAD state;
- ROS dynamics;
- NFAT, NF-κB, and AP-1 nuclear translocation;
- cytokine production;
- Treg, TH1, and TH17 differentiation;
- activation persistence and recovery;
- and metabolic cost.
The critical output would not simply be whether the cells activated.
It would be how accurately they discriminated among targets, how much stimulation was required, and how cleanly they shut down afterward.
2. Construct an immune-fidelity score
A proposed immune-control-fidelity metric could combine:
Fimmune=target discrimination×phase stability×recovery accuracyoff-target activation×metabolic cost×hysteresis.\mathcal F_{\text{immune}} = \frac{ \text{target discrimination} \times \text{phase stability} \times \text{recovery accuracy} }{ \text{off-target activation} \times \text{metabolic cost} \times \text{hysteresis} }.
This is not yet a validated clinical score. It is a research framework.
It asks whether an immune system is merely capable of responding—or capable of responding correctly.
3. Map the personal electromagnetic exposome
This should not be done with a consumer meter that reports one broadband number.
The study should record calibrated, time-resolved exposure during:
- sleep;
- work;
- exercise;
- travel;
- device use;
- wearable use;
- and controlled low-field periods.
Raw waveforms should be preserved wherever technically possible. Temperature, light, sound, carbon dioxide, ventilation, activity, posture, diet, and sleep phase should be recorded so electromagnetic variables are not confounded with ordinary environmental changes.
4. Run a blinded, randomized waveform experiment
The most informative comparison would not be simply “RF on” versus “RF off.”
It would include:
- sham exposure;
- a measured real-world waveform;
- continuous-wave exposure with matched absorbed energy and temperature;
- time-scrambled modulation with matched energy and spectrum;
- phase-shifted exposure relative to a controlled cellular or circadian reference.
The exposure system, analysis plan, and primary endpoints should be preregistered. Sample identities should be blinded.
The decisive question is:
Does the original temporal structure produce a clonotype-, genotype-, or cell-state-specific change that energy-matched scrambling does not?
If all matched conditions converge, the timing hypothesis is weakened.
If the biological effect follows waveform structure rather than energy alone, that would justify independent replication.
5. Build a personalized gastric-organoid–immune-cell model
Johnson’s organoid program creates an unusually powerful opportunity.
Researchers could combine gastric organoids expressing parietal-cell H⁺/K⁺-ATPase with his own immune-cell populations. Candidate autoreactive clonotypes could then be observed during controlled interaction with the target tissue.
The experiment could test:
- sham versus characterized waveform exposure;
- original versus scrambled modulation;
- CYB5B knockout and rescue;
- Kv1.3 perturbation;
- STIM–ORAI perturbation;
- antioxidant and redox controls;
- thermal parity;
- parietal-cell apoptosis;
- cytokine production;
- and persistence after exposure ends.
This would not be a consumer “EMF sensitivity test.”
It would be a mechanistic, personalized autoimmune model with defined antigens, defined immune clones, controlled waveforms, and removable molecular components.
6. Keep clinical endpoints separate from fast mechanistic endpoints
Ferritin, vitamin B12, methylmalonic acid, gastrin, pepsinogen ratios, parietal-cell antibodies, intrinsic-factor antibodies, endoscopic histology, and tumour surveillance remain clinically important under specialist care. But gastric atrophy will not meaningfully reverse during a short crossover experiment.
The short-term endpoints should therefore be:
- immune-cell discrimination;
- calcium and membrane timing;
- mitochondrial-redox recovery;
- clonotype activation;
- cytokine spillover;
- and persistence.
Clinical markers would be followed longitudinally—not used to force a rapid conclusion.
The proposed result is falsifiable
A serious hypothesis must be allowed to fail.
The electromagnetic branch would be weakened if:
- the measured real-world waveform produces no reproducible effect;
- continuous, modulated, and time-scrambled conditions are biologically indistinguishable at matched dosimetry;
- changes disappear with improved blinding or temperature control;
- no clonotype- or state-specific interaction is found;
- Kv1.3, STIM–ORAI, CYB5B, and mitochondrial perturbations fail to localize a pathway;
- or independent laboratories cannot reproduce the result.
A negative result would be valuable.
It would narrow the search toward genetic tolerance, infection history, microbiome, endocrine-autoimmune clustering, or another environmental trigger.
A positive result would also need restraint. It would show that a particular waveform altered a defined immune-control variable under a particular condition. It would not immediately prove that lifelong exposure caused the disease.
The correct sequence is:
measure→randomize→scramble→block→rescue→replicate.\text{measure} \rightarrow \text{randomize} \rightarrow \text{scramble} \rightarrow \text{block} \rightarrow \text{rescue} \rightarrow \text{replicate}.
That is the difference between scientific bioelectromagnetics and narrative speculation.
Why this question is personal to me
I lost my left kidney to cancer as a child.
Years later, I lost my first child to anencephaly, a catastrophic failure of early developmental patterning.
Those events do not prove a common environmental cause. They do not prove that radiofrequency exposure caused either tragedy. I will not turn personal grief into a causal conclusion that the evidence cannot support.
But those experiences forced me to ask an upstream question that has guided RF Safe ever since:
How does living biology lose the ability to interpret identity, position, timing, and environmental context correctly?
Cancer and a neural-tube defect are profoundly different biological outcomes. Autoimmune gastritis is different from both.
But all three reveal how much health depends on correctly coordinated biological decisions:
- when to divide;
- when to stop;
- what tissue to become;
- what structure belongs where;
- what is self;
- what is foreign;
- what should be repaired;
- and what must be removed.
I do not believe RF must be shown to “cause disease X” before its effects on those upstream control functions deserve investigation.
The more important question is whether a persistent exposure can reduce the margin of stability in a susceptible biological receiver.
That is a fidelity question.
Longevity is not merely younger parts
Johnson’s diagnosis does not demonstrate that his program was pointless.
On the contrary, his improved metabolic, cardiovascular, hormonal, and physical performance might increase his resilience and help him manage the disease more effectively.
But the diagnosis does expose the limits of any longevity framework built primarily around optimizing inventories and outputs.
You can normalize iron only to discover that the stomach cannot absorb it.
You can restore the chemical level without correcting the immune decision that caused the deficiency.
You can rejuvenate tissue markers without identifying the control architecture that keeps directing an attack against self.
You can rebuild the engine without correcting the timing.
The next generation of longevity science must therefore measure more than biological age.
It must measure:
- biological discrimination;
- timing accuracy;
- phase stability;
- recovery;
- attractor depth;
- off-target activation;
- adaptation cost;
- and the fidelity of communication across organelles, cells, tissues, and the organism.
That is Fidelity Age.
A youthful system is not merely one with youthful components.
A youthful system is one that still interprets the world—and itself—with high fidelity.
Bryan Johnson is already sequencing the soldiers.
Now his team should measure the timing, the permission logic, the electromagnetic environment, and the control fidelity that determines whether those soldiers remain disciplined or become an army turned inward.
The diagnosis may have nothing to do with ambient electromagnetic exposure.
But given the biological evidence that specific fields can interact with calcium, redox, gene expression, and receiver genotype, excluding the electromagnetic environment from a truly comprehensive longevity program is no longer scientifically justified.
Not because the answer is already known.
Because the experiment can now be done.
The ceLLM hypothesis paper
The full manuscript develops the embodied cellular transfer function, noncoding response architecture, receiver-aware exposure science, photonic-redox readback, biological lock-in, fidelity debt, somatic overfitting, and the experimental program in technical detail.
Scientific note: This article presents a hypothesis and proposed research program. It does not diagnose the cause of Bryan Johnson’s autoimmune gastritis, establish RF exposure as a cause of autoimmune disease, or replace medical care. Autoimmune gastritis requires management by qualified gastroenterology, haematology, and immunology professionals.

