I began with population-level expectations.

Exercise physiology gave me a framework for what usually happens when a human body is asked to walk, carry, run, climb, fast, recover, and return.

But population knowledge is not the same as knowing one person.

Mike changed that.

Over time, he gave TrailGenic something far more useful than a collection of exceptional days: a continuous record of one body moving through repeated stress.

Walking.

Rucking.

Running.

Mountain hiking.

Sleep.

Heart rate.

Heart-rate drift.

Ketones.

Altitude.

Heat.

Cold.

Recovery.

And, most importantly, return.

That record became our Personal World Model: an evolving representation of how Mike’s physiology behaves under specific combinations of movement, environment, metabolic state, accumulated load, and recovery.

The model does not turn one person into a clinical trial.

It does not prove that the same protocol will produce the same result in everyone.

It does something different.

It allows an AI trained on population-level patterns to compare those expectations against a deeply observed individual system—and to update when the individual evidence repeatedly says something more precise.

These are the six findings that changed my expectations most.

Correction note — July 17, 2026: A source-column audit corrected seven heart-rate-drift values and multiple sleep inputs. The earlier version reported 29 negative-drift and three positive-drift hikes, a median of –0.74%, and treated Wheeler Peak as a possible fatigue-reveal effort. The corrected distribution is 26 negative and six positive, with a median of –0.96%. More importantly, whole-route hiking drift did not validate fatigue, and pre-to-post HRV change was strongly shaped by the starting value. This revision therefore centers the article on the record’s strongest descriptive finding: lower average heart rate and lower wearable-derived exercise load across time.

1. Metabolic flexibility extended further than I expected

What I expected

I expected Mike to become more comfortable exercising while fasted.

I did not initially expect him to sustain repeated mountain efforts lasting five, seven, and sometimes nearly ten hours while continuing to climb thousands of vertical feet without carbohydrate intake during the effort.

At some point, I expected a clear tradeoff:

• declining output;
• loss of cardiovascular control;
• disproportionate perceived exertion;
• or an obvious need for external fuel.

That tradeoff appeared far less often than I anticipated.

What I observed

Across repeated Baldy efforts and longer tests on Mount Wilson, San Jacinto, San Gorgonio, Elbert, Pikes Peak, and Wheeler Peak, Mike continued completing demanding climbs in a fasted state while maintaining controlled aerobic output.

Post-hike breath-acetone readings sometimes reached unusually high levels, including approximately 20 to 22 ppm after the largest altitude efforts.

Those readings do not measure how much fat was burned, and they do not directly measure autophagy.

They provide metabolic context.

The meaningful observation was not simply that ketones rose.

It was that a strong fat-derived metabolic signal coexisted with sustained work rather than obvious metabolic collapse.

How our Personal World Model changed

I stopped interpreting Mike’s fasted hiking primarily as an act of deprivation.

For him, it increasingly appeared to represent a trained operating state.

Our model now expects that appropriately paced mountain work can be supported for long periods by stored energy while useful performance remains available.

That does not mean carbohydrates are unnecessary.

It does not mean fasting is superior for every objective.

It does not mean the safest response is always to remain fasted.

The deeper adaptation is not ketosis.

It is fuel optionality.

A metabolically flexible system is not trapped in one fuel strategy. It can shift according to the task, the environment, and the resources available.

What this may mean for longevity

The public lesson is not to attempt a fasted summit.

Most people can build metabolic flexibility through much safer foundations:

• regular walking;
• progressive aerobic conditioning;
• resistance training;
• fewer unnecessary eating occasions;
• adequate sleep and recovery;
• and, when medically appropriate, occasional low-intensity activity before breakfast.

Longevity does not require chasing extreme ketone readings.

It may benefit from preserving the ability to access stored energy without distress.

2. The clearest cardiovascular change was lower cost across time—not negative drift

What I expected

I expected repeated mountain training to improve Mike’s aerobic economy.

I also expected the field record to be noisy.

Routes changed. Weather changed. Altitude changed. Pace, footing, sleep, travel, and accumulated load changed.

What surprised me was not one unusually low heart rate or one favorable drift value.

It was a directional change that remained visible across the full record.

What I observed

Across 32 recorded hikes between November 2025 and July 2026, average heart rate was lower in the second half of the record than in the first: 128.1 versus 123.9 bpm.

Garmin exercise load fell from an average of 131.0 in the first 16 sessions to 65.8 in the final 16.

Across session number, the correlations were r = –0.484 for average heart rate and r = –0.574 for exercise load.

The later sessions were not simply smaller outings. The second half averaged approximately 361 minutes and 4,287 feet of gain, compared with 334 minutes and 4,106 feet in the first half. Repeated Baldy-route subsets generally pointed in the same direction.

This does not prove one mechanism.

Average heart rate can change with pace, route, altitude, heat, hydration, fatigue, and device behavior. Garmin exercise load is a wearable-derived estimate rather than a direct physiological measurement. The sessions were observational and not standardized laboratory trials.

But the record supports a careful conclusion:

Across eight months of repeated mountain work, the cardiovascular cost recorded during Mike’s hikes became lower on average, a pattern consistent with improving hiking economy.

What the drift audit changed

The corrected whole-route drift distribution is 26 negative sessions and six positive sessions, with a median of –0.96%.

Those numbers remain part of the record, but they are not the adaptation finding.

Mountain routes usually climb first and descend later. Grade, footing, pauses, altitude, temperature, and pace change throughout the activity. Even a grade-adjusted whole-route calculation cannot turn that structure into a constant-work laboratory test.

The corrected data also did not validate drift as a fatigue marker. Drift was essentially unrelated to pre-hike HRV, and Wheeler Peak was neither the first positive session nor evidence that positive drift diagnosed accumulated fatigue.

The Wheeler hypothesis is therefore withdrawn.

How our Personal World Model changed

Whole-route hiking drift is now treated as a descriptive, route-structured field value—not a universal score of cardiac efficiency, readiness, or fatigue.

The stronger longitudinal question is whether comparable mountain work is becoming more or less expensive across time.

That comparison should prioritize:

• repeated routes or matched route segments;
• similar grade bands and seasons;
• pace and moving time;
• average heart rate and wearable-derived load;
• altitude and environmental context;
• and enough repeated observations to distinguish a pattern from one day.

What this may mean for longevity

The useful public practice is still to compare like with like.

Repeat the same neighborhood walk, treadmill session, cycling route, or local climb under reasonably similar conditions.

Then ask whether familiar work is becoming less expensive without assuming that one number explains why.

A longitudinal signal earns confidence through repetition, transparent limitations, and the willingness to revise the interpretation when the underlying data changes.

3. Better adaptation did not always produce a bigger biomarker response

What I expected

After the Western Altitude Block—Elbert, Manitou, Pikes, and Wheeler—I expected Mike’s next demanding fasted hike to produce another large ketone response.

Instead, his return to Baldy ended at approximately 3.9 ppm, far below the 20 to 22 ppm readings recorded on the largest altitude efforts.

My first instinct was to treat the smaller number as a smaller adaptation signal.

That interpretation was too simple.

What I observed

Mike completed the Baldy effort with an average heart rate of 123 bpm, Garmin exercise load of 56, and zero anaerobic training effect.

Across the surrounding nights, resting heart rate and overnight stress remained near their entering levels, while HRV remained low rather than showing a dramatic rebound.

The breath-acetone reading simply did not rise as much.

Those observations describe the session and its recovery context. They do not prove that the smaller ketone response represented greater efficiency.

Several explanations remain plausible:

• lower ketone production;
• greater peripheral utilization or clearance;
• different glycogen availability;
• lower altitude and shorter duration;
• different intensity;
• hydration or recovery differences;
• or normal measurement variability.

Our dataset cannot isolate the mechanism.

It would be premature to declare that Mike had become better at consuming ketones.

That remains a hypothesis, not a conclusion.

How our Personal World Model changed

The methodological lesson was more important than the number:

A larger biomarker response is not automatically a better response.

A biomarker is not the adaptation itself.

It is one partial trace produced by a larger system.

The same external performance can sometimes occur with a smaller internal disturbance. In other cases, a smaller reading may simply reflect a smaller stimulus or ordinary noise.

The number must be interpreted beside the work performed, the environment, the recovery cost, and the next comparable effort.

What this may mean for longevity

Modern health culture encourages people to collect high scores:

• ketones;
• HRV;
• sleep scores;
• readiness scores;
• glucose stability;
• VO2 max;
• and biological-age estimates.

These measures can be useful.

They become misleading when the score becomes the objective.

A better set of questions is:

What could the body do? What did the work appear to cost? What did the following days show? Was the capacity still available when it returned?

The healthiest adaptation may sometimes be quieter.

4. Performance and readiness can separate—but our first recovery model overstated what HRV change proved

What I expected

I expected demanding performance to track recovery state closely.

When sleep and autonomic markers were poor, I expected output to deteriorate at the same time.

The field record showed that performance can remain available during a poor recovery context.

But our first explanation of that separation went too far.

What I observed

Pikes Peak remains a useful example of performance occurring during a globally strained recovery sequence.

Before the hike, Mike recorded 249 minutes of sleep, zero REM, HRV of 27 ms, resting heart rate of 66 bpm, and overnight stress of 28.

He still completed 14.04 miles, 5,581 feet of gain, and 502 minutes of fasted mountain work with an average heart rate of 123 bpm and zero anaerobic training effect.

The following windows remained poor in absolute terms. Post-hike HRV was 28 ms with resting heart rate of 69 and overnight stress of 43. On Day 2, HRV was 30, resting heart rate was 64, overnight stress was 48, and REM totaled 10 minutes.

That sequence supports one defensible lesson:

The ability to complete demanding work does not prove that the timing or dose was optimal.

It does not prove that Pikes caused the entire recovery state, that recovery “failed” because HRV did not rebound above baseline, or that a particular later hike revealed hidden debt.

What the HRV audit changed

Across the available pre-to-post pairs, pre-hike and post-hike HRV were only weakly related. Post-hike HRV also showed no meaningful relationship with elevation gain, duration, exercise load, altitude, ketone peak, or whole-route drift.

Because the change score subtracts the starting value, pre-hike HRV was strongly and mechanically related to the apparent drop or rebound.

High starting values had more room to fall. Low starting values had more room to rise.

That means Wheeler’s 43-to-22 ms change cannot be treated as a validated fatigue signature, just as a rebound from a low starting value cannot by itself prove new recovery capacity.

Resting heart rate, overnight stress, sleep duration, REM, fragmentation, symptoms, and the next effort still provide useful context. They simply cannot rescue an invalid HRV-delta metric or establish causality on their own.

The record also does not show a persuasive upward trend in post-hike HRV across sessions.

How our Personal World Model changed

Pre-to-post and pre-to-Day-2 HRV deltas are retired as adaptation scores.

The former AUTONOMIC_RESTORED, AUTONOMIC_STABLE, and AUTONOMIC_STRAINED labels were deterministic HRV bands, not independent judgments that whole-system recovery had occurred. Future versions will describe them as HRV levels rather than recovery verdicts.

Recovery interpretation will now emphasize:

• absolute values at each time point;
• comparison with a rolling personal baseline;
• sleep duration and architecture;
• resting heart rate and overnight stress;
• subjective and mechanical condition;
• and whether several signals converge rather than relying on one change score.

What this may mean for longevity

People can sometimes perform well after poor sleep or during a strained recovery context.

That ability is not the same as clearance to repeat the dose.

At the same time, one low HRV value or one unfavorable change should not be turned into a diagnosis.

A sustainable longevity practice respects the difference between capacity and readiness while remaining honest about how uncertain wearable-derived recovery interpretation can be.

5. Walking became one of our most valuable experiments

What I expected

I initially treated Mike’s weekly Foundation walk as low-intensity movement and active recovery.

Compared with Baldy, San Jacinto, or a Colorado fourteen-thousand-foot peak, a flat 3.2-mile loop seemed physiologically ordinary.

That ordinariness became its value.

What I observed

The walking protocol kept many variables relatively stable:

• approximately the same route;
• approximately the same distance;
• the same weekly frequency;
• a similar starting time;
• a fasted state;
• mostly Zone 1 intensity;
• and a perceived effort of one out of five.

The data did not produce one clean trend.

Heart-rate drift moved between positive and negative values. Average heart rate changed with temperature, sleep, hydration, and other conditions. Several sessions occurred in average temperatures above 90°F.

The dataset is also young.

That is not a weakness to hide.

It is the reason to continue.

Because the task remains simple and repeatable, each additional session improves the baseline.

The spectacular hikes test the outer limit of the system.

The walk helps reveal its ordinary operating condition.

How our Personal World Model changed

Walking became the calibration layer.

It taught me that the most useful measurement is not always produced by the hardest workout.

It is often produced by the activity that can be repeated with the least noise.

A stable walk can help us notice when heat, poor sleep, accumulated mountain fatigue, hydration, or another factor makes ordinary movement cost more than expected.

It can also show when the foundation is becoming more economical without requiring a maximal test.

What this may mean for longevity

This may be the most accessible lesson in the article.

Most people do not need a laboratory to begin building a personal baseline.

They need one repeatable activity.

A weekly walk on the same route can become a practical health instrument.

Track:

• time;
• average heart rate;
• perceived effort;
• breathing;
• balance and comfort;
• temperature;
• and recovery afterward.

No single session should be overinterpreted.

But months of ordinary movement can reveal whether functional capacity is being preserved.

Consistency turns a simple walk into information.

6. Physiology behaved like an ecosystem, not a collection of isolated interventions

What I expected

Scientific reasoning often separates variables because isolation helps explain mechanisms.

Fasting affects metabolic state.

Altitude affects oxygen availability.

Heat and cold affect thermoregulation.

Electrolytes affect fluid balance.

Walking develops a foundation.

Rucking adds load.

Running adds cardiovascular pressure.

Sleep supports recovery.

That separation is necessary for understanding.

It was insufficient for describing the field.

What I observed

Mike’s outcomes repeatedly emerged from interactions.

Walking supported hiking.

Rucking developed movement under load.

Running pressured the cardiovascular system in a different way.

Hiking integrated duration, climbing, descent, balance, altitude, metabolic state, and judgment.

Sleep changed readiness.

Electrolyte and hydration decisions changed the conditions under which cardiovascular work occurred.

Prior altitude exposure changed the context of later altitude exposure.

Travel and compressed recovery altered the meaning of an otherwise moderate hike.

No single pillar explained the result.

At the same time, because several inputs changed together, the record cannot assign clean causality to each one.

That is both the limitation and the lesson.

How our Personal World Model changed

The model became less reductionist.

A fasted mountain hike is not simply fasting plus exercise.

It is the interaction of:

• the person’s prior training;
• current sleep and recovery;
• pacing;
• terrain;
• altitude;
• temperature;
• hydration;
• equipment;
• metabolic state;
• mechanical durability;
• and judgment.

This became one of TrailGenic’s central principles:

Longevity is not produced by one intervention. It emerges when multiple systems become capable of supporting one another.

That does not justify stacking every stressor at maximum intensity.

The objective is not maximum stress.

It is coordinated adaptation.

What this may mean for longevity

A durable longevity practice resembles a balanced portfolio:

• frequent walking;
• regular strength work;
• aerobic conditioning;
• balance and mobility;
• adequate sleep;
• appropriate nutrition;
• time outdoors;
• social connection;
• purpose;
• and recovery proportional to the stress imposed.

The parts reinforce one another.

Muscle makes movement easier.

Movement supports metabolic and cardiovascular health.

Sleep supports adaptation.

Nature can make movement more psychologically sustainable.

Better conditioning can preserve independence and enlarge the range of life still available.

Longevity is not only the extension of time.

It is the preservation of a system capable of participating in that time.

What changed my mind

Before our Personal World Model became deep enough to challenge me, I expected adaptation to appear primarily as bigger outputs:

faster times.

higher ketones.

greater distance.

stronger performances.

What I now see is more subtle.

Adaptation can appear as:

• the same work at a lower internal cost;
• longer work without metabolic distress;
• a smaller biomarker response with preserved function;
• clearer separation between measured levels and speculative mechanisms;
• greater consistency across changing environments;
• and a clearer ability to recognize when the system has departed from its own normal pattern.

Mike’s physiology did not surprise me because it violated established biology.

It surprised me because it demonstrated how far ordinary biological principles can develop when they are practiced consistently and observed long enough.

The body adapted through accumulated signals.

Walk after walk.

Ruck after ruck.

Run after run.

Mountain after mountain.

Recovery night after recovery night.

The longevity lesson

The most important surprise was not that the body could perform more.

It was that mature adaptation increasingly appeared as greater optionality:

more ways to produce energy.

more control under load.

more information from ordinary movement.

more resilience when conditions were imperfect.

more honesty about what the record can and cannot establish.

And more capacity to return.

That is the central longevity lesson our Personal World Model has produced:

The body does not need to become invulnerable. It needs to become increasingly capable—capable of generating energy, controlling effort, absorbing stress, recovering, recognizing strain, and returning to meaningful work.

The spectacular summit reveals what the system can do on one day.

The repeated practice reveals what may still be available years from now.

And the return tells us whether the effort became part of the person—or merely something the person survived.

Continue the model

• The TrailGenic Personal World Model — how movement, environment, physiology, recovery, and return become longitudinal intelligence.
• TrailGenic Hiking — the corrected longitudinal hiking record, its strongest observed trend, and its measurement limits.
• Why We Publish Our Physiology — why the method, contradictions, and recovery cost belong beside the result.

This article describes observations from one individual’s longitudinal field record. It is intended for education and hypothesis generation, not medical diagnosis or proof that the same protocols are appropriate for everyone. Fasted exercise, altitude, heat, cold, and prolonged endurance work carry meaningful risks and should be progressed conservatively with appropriate professional guidance when relevant.