This extra chapter was cut from The Journey Back to keep the book focused. The physical paperback is $12.95 with free U.S. shipping. Any $17 PDF-and-audiobook bundle is a separate optional digital add-on, not the price of the paperback.
Because I have a degree in engineering, repeatability matters to me.
If I cannot reproduce a result, I do not have convincing evidence that the result is dependable for me.
Put that into an engineering context.
If I am helping determine whether a building can withstand a load without collapsing, one favorable result is not enough. I cannot test something once, watch it survive, and declare that the structure will always be safe.
Responsible engineering requires defined conditions, calculations, appropriate standards, margins, inspection, and evidence that does not depend on getting lucky one time.
The human body is not a building.
A glucose meter is not a structural laboratory.
But the discipline transferred.
I wanted to know whether something I observed was repeatable or whether I had merely captured one moment.
So I kept testing.
I even repeated meals that had already produced acceptable readings because an acceptable reading on Tuesday did not guarantee the same response on Friday.
That is where I began seeing the limitation of the way most people talk about food and blood sugar.
A person eats something, checks the meter, and says, ‘I ate this food and got this result. Therefore, the food caused the result.’
That conclusion is too clean.
The food mattered.
It was not operating alone.
The result belonged to a system.
By then, I had started organizing that system into the Four Loads: Storage, Regulatory, Inflammatory, and Glucose.
Most food experiments focus only on Glucose Load.
How many carbohydrates were in the meal?
How quickly were they absorbed?
What was the reading before eating?
How high did the meter go afterward?
Those are important questions, but they are not the complete experiment.
What happened under the other three Loads?
How did I sleep?
How much stress was I carrying?
Had I been active or sedentary?
What had I eaten earlier?
Was my routine different?
Was inflammation, illness, pain, recovery, hydration, or another condition changing the context?
Was I comparing the same portion, preparation, and timing?
If I ignored those questions, I could make the food responsible for everything the system produced.
That was the larger discovery.
A meal does not enter an empty machine.
It enters the body that woke up that morning.
It enters the sleep from the night before.
It enters the stress of the day.
It enters the movement that happened or did not happen.
It enters the medication context, hydration, prior meals, and timing.
The food is an input.
It is not the only input.
One of the clearest demonstrations involved pizza.
Earlier, I had tested thin-crust Domino’s pizza. I ate what I remember as one slice, and my glucose rose to approximately 180 mg/dL. I repeated a similar slice the following day and saw a similar result.
Then I ate another slice and took approximately a ten-minute walk. The rise appeared much smaller.
That experiment made me feel as if I had found a lever.
I could eat a slice of pizza, walk afterward, and influence the response.
I began feeling confident enough to say, ‘Charles has this covered.’
Then Alen and Brandon brought me pizza from Brixx.
The official restaurant name appears to be Brixx Wood Fired Pizza + Craft Bar, a Charlotte-founded chain.
These were my friends. They knew I had tested Domino’s. They knew I had eaten pizza, gone for a walk, and seen a result that encouraged me.
I ate the Brixx pizza believing I knew how the system would respond.
I told them, very matter-of-factly, that if my glucose rose above approximately 180, I would do one hundred air squats.
I was not quietly hoping for the best.
I announced the prediction.
Charles had it covered.
About an hour later, Alen texted me.
‘Check your glucose.’
So I checked it.
The meter showed approximately 223 mg/dL.
Oh my gosh.
That was not the result I had predicted.
Alen and Brandon were looking at it almost as if to say, ‘What happened?’
I had made the rule in advance, so I sent them pictures of myself doing the air squats.
The image is funny now.
I had confidently announced that I understood the machine, and a few hours later I was documenting the penalty clause.
But the number exposed something important.
My previous experiment had not given me control.
It had given me one promising pattern under a particular set of conditions.
The Brixx meal was not a clean repetition of the Domino’s experiment.
It was a different restaurant.
It was a different pizza.
The crust could have been different.
The amount of dough, sauce, cheese, oil, and toppings could have been different.
The slice size could have been different.
The portion I ate could have been different.
The meal timing and testing interval might have been different.
Even if both foods were called pizza, that did not make them the same input.
I also remembered that I had eaten Brixx pizza before without seeing such a large rise.
That made the result even more confusing.
If Brixx had produced one response before and approximately 223 this time, then the restaurant name alone could not explain the difference.
I started looking beyond the plate.
The night before, I had slept badly.
I was also under significant stress because of deadlines.
Those conditions belonged to Regulatory Load.
The pizza belonged primarily to the immediate Glucose Load.
The movement I planned afterward belonged to the way I was using muscle and activity within the larger system.
Other conditions could have been present too.
When I pulled the picture together, the 223 reading stopped looking like an inexplicable act of betrayal.
It became evidence that the context had changed.
I need to be careful with the next sentence.
At the time, I might have said that one poor night of sleep could create a twenty percent increase in glucose and insulin.
That is too precise to apply as a universal rule.
Controlled studies do show that sleep restriction can reduce insulin sensitivity and impair glucose regulation. In one small laboratory study of nine healthy adults, a single night restricted to four hours of sleep reduced the glucose infusion rate during a clamp study by approximately twenty-five percent. That is not the same as proving that every person’s glucose meter will read twenty or twenty-five percent higher after one bad night.
It also does not prove that my bad night caused the 223 reading.
My sleep and stress made the result more biologically plausible.
They did not isolate the cause.
That distinction is the difference between using the Four Loads as a framework and using them as an excuse.
The framework tells me what variables to investigate.
It does not allow me to assign blame without evidence.
I could not honestly say, ‘The pizza did it.’
I also could not honestly say, ‘My sleep did it.’
I could say that a different pizza, on a poorly rested and stressful day, produced a much larger glucose reading than I had expected.
That was the observation.
The explanation remained a hypothesis.
The experience showed me why people become so frustrated with blood sugar.
They can eat a meal one day and see a number they consider acceptable. They can eat what appears to be the same meal later and see a much higher number.
From their perspective, the meter looks irrational.
They begin saying, ‘My blood sugar has a mind of its own.’
Or, ‘I have no idea where these numbers are coming from.’
The frustration is real.
The missing information may be that the meal was never the only variable.
If someone logs only the food, two rows in the diary may look identical.
Pizza.
Pizza.
But the actual experiments may be completely different.
One followed a full night of sleep.
One followed a short, interrupted night.
One occurred on a calm day.
One occurred while deadlines were pressing.
One involved a ten-minute walk.
One involved sitting.
One slice was thin and small.
Another was larger or built on a different crust.
One test happened at sixty minutes.
Another happened later.
The food log says the meals were similar.
The body experienced different systems.
That is why I used the Oura Ring Alen had given me.
The ring helped me record an estimate of my sleep instead of relying only on the story I told myself the next morning.
Later, Brandy gave me a COROS PACE 3, which I absolutely love. It also records sleep-related information.
When I compare the Oura and COROS results, they may be close, but they are not identical.
That does not surprise me.
There is no perfect consumer sleep tracker.
These devices estimate sleep using their sensors and algorithms. They do not turn my bedroom into a clinical sleep laboratory, and they should not be used to diagnose a sleep disorder.
The disagreement between them did not make both devices worthless.
It reminded me to use them for patterns rather than pretend that every minute and every sleep stage was exact.
If both devices, my bedtime, and my own experience suggested that I had slept poorly, that gave me relevant context for the next day’s observations.
It did not allow me to calculate the exact number of glucose points caused by sleep.
That is the level of honesty I want the Four Loads to preserve.
I am not trying to give people four new boxes into which they can force every result.
I am trying to stop them from forcing every result into one box labeled food.
The Four Loads widen the investigation.
Glucose Load asks what entered the bloodstream and how the immediate response appeared.
Regulatory Load asks what sleep, stress, hormones, and timing may have contributed.
Storage Load asks how movement, muscle, energy use, and the body’s capacity to handle fuel fit the pattern.
Inflammatory Load asks whether food quality, illness, recovery, hydration, and other inflammatory conditions may be part of the context.
The categories interact.
They are not four sealed rooms.
The pizza experiment mattered because all four rooms had doors open at once.
The mistake would have been to look only at the pizza box.
The other mistake would have been to see the bad sleep and declare the pizza innocent.
Both would have been oversimplifications.
The meal supplied a substantial carbohydrate input.
The sleep and stress may have changed the way my body handled that input.
The absence or timing of movement may have changed the response again.
The glucose meter captured the combined result at one moment.
It did not print an itemized receipt showing how many points belonged to the crust, how many belonged to sleep, how many belonged to stress, and how many belonged to the measurement itself.
I had to reconstruct the conditions and remain humble about what I could prove.
That is what a failed repetition taught me.
Failure did not mean the experiment was useless.
The failure revealed which assumption needed to be challenged.
I had assumed that my earlier Domino’s experience gave me a general pizza rule.
It did not.
I had assumed that a walk or a set of air squats could reliably control the response.
That had not been established.
I had assumed that foods with the same name were close enough to count as the same input.
They were not.
I had not given enough weight to the sleep and stress that preceded the meal.
The 223 reading forced those hidden assumptions into view.
Even the air squats require honest interpretation.
I did them because I had promised my friends that I would. They were a personal response to the number and a visible reminder that I had made a prediction.
They were not a medically established punishment for a reading above 180.
I do not want readers inventing an exercise penalty every time a meter surprises them. Exercise decisions can depend on the glucose level, symptoms, medications, hydration, cardiovascular status, injuries, and medical guidance.
The useful part of the scene is not that one hundred squats corrected the meal.
The useful part is that I had to confront a prediction that failed.
Alen’s text did not let me forget the measurement.
‘Check your glucose.’
The meter answered.
My model had to change.
That is how engineering is supposed to work.
A failed test is not an insult.
It is information about the model.
If the result does not match the prediction, I can defend the prediction, blame the instrument, or investigate the assumptions.
I chose to investigate.
That did not always give me a final answer.
Sometimes the most accurate conclusion was, ‘I do not know exactly why this happened.’
That sentence is not weakness.
It is a boundary around the evidence.
In the Brixx case, I had several reasonable suspects.
Different pizza.
Different crust.
Possibly different portion.
Bad sleep.
Deadline stress.
Timing.
Activity.
Normal meter limitations.
But I had not held the other variables constant, so I could not identify one culprit with confidence.
The conclusion was not, ‘Sleep caused 223.’
The conclusion was, ‘My food-only explanation failed.’
That was enough to move the framework forward.
It showed me that repeatability requires more than repeating the name of the food.
To compare two results honestly, I needed to record the surrounding Loads.
How much did I sleep?
How did the wearable estimate compare with my own experience?
What stress was present?
What movement occurred before and after the meal?
What exactly did I eat?
How much did I eat?
What was the starting glucose?
When did I begin eating?
When did I finish?
When did I test?
What medication had I taken according to the prescribed schedule?
What happened at the next reading?
Without that context, a number can be accurate and the conclusion can still be wrong.
That is one of the messages I most want to get to the public.
If your glucose response feels unpredictable, it does not necessarily mean that you are failing or that your body is behaving randomly.
It may mean that the variables you are recording do not yet include enough of the system.
That does not promise that every surprising number can be explained.
Bodies remain complex. Meters have limitations. Illness, hormones, medication, digestion, and factors we do not see can affect a reading. Some results remain unexplained even after careful review.
But looking at the Four Loads gives a person better questions.
What changed besides the food?
That question can turn frustration into investigation.
For me, Brixx was the meal that broke the illusion of a single-variable answer.
I had pizza before.
I had walked before.
I believed I knew what would happen.
Then the meter read approximately 223.
The surprise did not destroy the method.
It improved the method.
I stopped asking whether pizza always did one thing.
I started asking what combination of Loads had produced this response on this day.
I stopped treating an earlier successful walk as a guarantee.
I started treating it as one intervention whose effect had to be retested under changing conditions.
I stopped expecting the Oura Ring or COROS watch to deliver perfect sleep truth.
I started using their estimates as context within a larger record.
I stopped believing that repeatability meant forcing the body to produce the same number.
I began understanding that repeatability meant documenting the conditions closely enough to know whether I had actually repeated the experiment.
That may be the most important distinction.
I thought I was repeating pizza.
I was not repeating the night before.
I was not repeating the stress.
I was not repeating the crust, portion, timing, or full pattern of movement.
The food looked similar.
The experiment was different.
The meter did not betray me.
It exposed the variables I had left out.
Want to see how I connected the dots?
I wrote The Journey Back to share my experience and explain the Four Metabolic Loads framework. The physical paperback is $12.95 with free U.S. shipping.
→ See The Journey Back Paperback ($12.95)
Educational only, not medical advice. Charles’s experience was alongside physician-supervised care and prescribed medication. Individual results vary. Talk to your doctor.