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.
A lot of my early experiments were documented in an app.
I logged the food in Carb Manager, checked my glucose, and tried to connect what I had eaten with what the meter showed me afterward.
But I discovered an important asymmetry in the method.
One meal could give me a reason to say, ‘That response was bad for me under these conditions.’
One meal could not give me enough evidence to say, ‘This food is good for me.’
A high reading was a warning that deserved investigation. A low reading was only one successful result.
The difference mattered because it would have been easy to turn every acceptable number into permission.
If I ate one food once and my glucose did not rise very much, I could not declare that the food was harmless. The portion might have been different. The timing might have been different. I might have slept better, walked more, eaten it with another food, tested too early, tested too late, or received a reading within the ordinary limitations of the meter.
A food did not earn a permanent green light because it passed one test.
It had to survive repetition.
The first repeatable pattern did not come from anything exciting.
It came from boring meals.
For breakfast, I ate three eggs with a little salt and pepper.
At that point, even the salt made me nervous. I had just left the hospital after dangerously high blood pressure, and I was terrified of doing anything that might push the numbers in the wrong direction. I was treating seasoning as if one careless shake could undo everything.
But the breakfast itself was simple.
Three eggs.
A little salt.
A little pepper.
Then I tested.
The eggs did not produce the kind of glucose rise I was watching for.
That did not prove eggs were a perfect food. It did not answer every question about cholesterol, calories, saturated fat, sodium, or how another person might respond. It told me something narrower and more useful.
When I ate that breakfast in that portion, under those conditions, my meter did not show a large post-meal glucose rise.
Then I did it again.
Three eggs became the first meal of the day over and over. The repetition mattered more than the novelty. I did not have to wake up every morning and invent breakfast. I already knew what I planned to eat, and I was collecting another observation under conditions that were at least reasonably similar.
Lunch became another repeatable meal.
I went to Walmart and bought a bag of frozen salmon fillets with the skin on. I also bought tilapia because I enjoy fish and could imagine eating it frequently.
I seasoned it with a measured amount of Old Bay, cooked it, and ate it with a large serving of broccoli.
Because I am a dipping person, I used a small amount of guacamole with the broccoli.
Salmon.
Broccoli.
A little guacamole.
A little Old Bay.
Then I tested.
I did not see a large glucose rise.
Again, that result did not allow me to say, ‘This meal is good for every person in every situation.’ It allowed me to say that this combination appeared to work predictably for my glucose under the conditions I was creating.
I could repeat the meal without feeling as though I was gambling every afternoon.
Dinner started with a bagged salad.
I used a very small amount of vinaigrette because I had become aware that dressings could contain sugar. The salad gave me volume. I could eat a great deal of it, feel physically full, and still see a relatively level glucose response on my meter.
That led to the next experiment.
If the salad appeared to work, could I turn it into something that felt more like a complete dinner?
I took a low-carbohydrate keto-style wrap and laid it flat. I spread a small spoonful of guacamole across it. Then I folded or placed a piece of seaweed on the wrap, added part of the bagged salad, and topped it with breast meat I had pulled from a rotisserie chicken.
I finished it with a small amount of balsamic vinegar. I knew balsamic vinegar could contain sugar, so I did not pour it on without measuring or paying attention.
Then I made a second wrap.
After two of them, I felt as if I had eaten Thanksgiving dinner.
That was important because I was not merely asking whether a meal produced a glucose rise.
I was asking whether I could live with it.
A meal that kept the meter level but left me hungry, miserable, and thinking about food all evening would not have been a useful long-term answer for me.
These wraps gave me volume, texture, protein, the experience of dipping and dressing food, and a sense that I had eaten a substantial meal.
The meter did not show the kind of glucose rise I had seen with higher-carbohydrate foods.
I need to be precise about the language here.
At the time, I might have said, ‘It did not spike my insulin.’
I was not measuring insulin.
I was measuring glucose.
A finger-stick glucose meter cannot tell me exactly what my insulin did after a meal. It can only show the glucose reading at the moment I tested. The absence of a dramatic glucose rise does not prove that insulin remained unchanged.
What I could honestly say was that I did not observe a large glucose rise with those repeated meals.
The three eggs, the fish with broccoli and guacamole, and the salad wraps created the first dependable structure of my day.
The pattern was not based on one spectacular number.
It was based on the absence of a surprise across repeated ordinary meals.
That may be one of the least glamorous discoveries in the entire journey.
Reliability often looks boring.
The same breakfast.
The same approximate portion.
The same basic lunch.
The same kind of dinner.
The same testing process.
The same log.
That repetition gave me a baseline.
Once I had a baseline, I could change one thing and see whether the response changed with it.
That is when Domino’s entered the experiment.
Alex had bought a thin-crust pizza from Domino’s. I ate one slice and checked my glucose.
The reading rose to approximately 180 mg/dL.
At the time, I knew people who had experienced much higher readings, so part of me thought, ‘That is not bad bad.’
But comparison with somebody else’s worst story was not a useful standard.
My question was not whether another person had ever seen a higher number.
My question was what that slice did in my own experiment.
The rise was meaningfully different from the responses I had been seeing after my eggs, salmon, broccoli, salad, and wraps.
So I tested it again.
The next day, I reheated another slice in the microwave and ate it. My glucose again reached approximately 180 mg/dL.
That repetition changed the meaning of the first result.
One reading near 180 might have been noise, timing, technique, or an unusual day.
A similar response after a similar portion on the next day did not prove that every slice of pizza would always produce exactly the same number. But it made the pattern harder to dismiss.
The pizza experiment had now repeated closely enough to earn my attention.
Then I changed one variable.
I ate what I remember as another single slice and went for approximately a ten-minute walk afterward.
When I tested, the glucose rise was much smaller than the rises I had recorded without the walk.
That felt enormous.
I had not discovered that pizza no longer contained carbohydrates.
I had not proved that walking canceled the meal.
I had not demonstrated that a post-meal walk made pizza nutritionally equivalent to salmon and broccoli.
I had discovered that movement appeared to change my glucose response to that meal.
That was a new degree of freedom.
Before that experiment, I had been thinking mostly in terms of food categories.
Eat this.
Never eat that.
The walking result suggested that the response was not determined only by what sat on the plate. What I did after the meal could become another input.
I thought about the muscles in my calves, including the soleus, working as I walked. But my home experiment did not isolate one muscle or prove a specific physiological mechanism. I was using my legs, not conducting a muscle-biopsy study.
The honest observation was simpler.
A short walk after the pizza was associated with a smaller glucose rise in that test.
Research conducted under controlled conditions has also found that walking after meals can reduce post-meal glucose exposure in people with type 2 diabetes. That supports the plausibility of what I saw. It does not validate every detail of my personal experiment or guarantee that everyone will receive the same result.
I also had to acknowledge the weaknesses in my test.
Was each slice exactly the same size and weight?
Did each slice have the same amount and distribution of crust, sauce, cheese, and toppings?
Did I begin each test at the same glucose level?
Had I slept the same amount?
Was my stress the same?
Had I eaten the same earlier meals?
Did I test at precisely the same number of minutes after the first or last bite?
Was the ten-minute walk the same pace throughout?
What was the reading before the walk?
What was it immediately afterward, and what happened later?
Those details determine how confidently I can compare the results. They need to be reconstructed from the app, meter history, videos, and any written notes that remain.
Still, the sequence gave me a method.
First, repeat the meal without the intervention.
Then change one variable.
In this case, the variable was walking.
If the result changed, repeat the comparison before treating it as a dependable pattern.
That last step is important because the first walk after pizza was promising, but one walk was not enough to establish a personal law.
The repeated eggs, fish, vegetables, and wraps gave me more confidence than the single walking result because I had seen those meal responses across many days.
The pizza result was the beginning of a new hypothesis.
The hypothesis was not, ‘I can eat bad food and be okay if I walk.’
That language gives the walk more power than the experiment established.
A more accurate hypothesis was, ‘Post-meal walking may reduce my glucose rise after some meals, including meals with more carbohydrate.’
That was something I could test again.
It also changed the question I asked about the meals that were already producing relatively level readings.
If walking appeared to help after pizza, what would happen if I walked after salmon and broccoli, eggs, or one of the salad wraps?
Could a useful response become even steadier?
Could movement after meals become part of the routine instead of emergency damage control after a food I regretted?
That was a much more powerful direction.
The walk did not have to serve as payment for pizza.
It could become a normal input in the system.
This is where my methodology became slightly different from the way people often talk about food.
People want a list.
Good foods on one side.
Bad foods on the other.
My meter would not give me a list that simple.
A result belonged to a portion, a combination, a preparation, a time, a sequence, a night of sleep, an activity pattern, a medication context, a stress level, and a particular human body.
The same meal could produce a different response on another day.
That did not make the data useless.
It meant the claim had to remain proportional to the evidence.
If one meal produced a reading I considered too high, I could decide not to repeat that meal while I investigated. That was a protective decision. I did not need a scientific paper about my exact slice before choosing not to eat another one.
But if one meal produced a low reading, I needed more restraint before declaring victory.
A single acceptable number can create false confidence.
It can make a person stop measuring, increase the portion, add another food, or assume the same result will occur at every time of day.
That was why ‘good’ required repetition.
Even after repetition, I was not proving that the meal was healthy in every sense.
The meter could not evaluate the full nutritional quality of a bagged salad, the sodium in a rotisserie chicken, the fat profile of a meal, the effect on triglycerides, the effect on blood pressure, or what years of eating it might mean.
It was one instrument observing one part of the system.
The glucose result mattered.
It was not the entire verdict.
That is also why fullness mattered to me.
The salad wraps did more than produce an acceptable reading. They made me feel as if I had eaten Thanksgiving dinner. They reduced the likelihood that I would go searching for food an hour later.
That experience could not be read from the glucose meter.
I had to record it from the other instrument in the experiment: me.
Was I hungry?
Was I satisfied?
Did I enjoy the meal?
Could I prepare it consistently?
Did it fit the life I was actually living?
Could I eat it again without feeling punished?
A repeatable meal had to work both on the meter and in my life.
That is why the basic meals became so valuable.
Three eggs were easy to repeat.
A piece of frozen salmon and a bag of broccoli were easy to measure.
A small amount of guacamole gave me the dipping experience I liked.
A bagged salad could be turned into two large wraps with seaweed and chicken.
The ingredients were not gourmet. The structure was dependable.
Dependability gave me something to compare the experiments against.
When an experiment failed to repeat, I did not automatically assume the body had betrayed me or the entire method was wrong.
I looked for the changed variable.
Was the portion larger?
Was the label different?
Did the product manufacturer change the ingredients?
Did the rotisserie chicken contain something I had not logged?
Did I use more dressing or balsamic vinegar?
Did I start at a different baseline?
Was the testing interval different?
Was I sick, dehydrated, stressed, or poorly rested?
Did I wash my hands and obtain an adequate sample?
Was the strip stored correctly?
Did the unexpected result appear again?
Sometimes I could find the likely difference.
Sometimes I could not.
An unexplained result did not have to become a crisis. It could remain an unexplained result until more evidence appeared.
That may be the hardest discipline in self-experimentation.
You have to be willing to say, ‘I do not know yet.’
The marketer in me likes a clean conclusion.
The engineer in me has to respect uncertainty.
The man who had just left the hospital wanted an answer immediately.
The method required me to earn confidence through repetition.
That is what the ordinary meals gave me.
They were not exciting, but they created a stable platform.
Breakfast did not surprise me.
Lunch did not surprise me.
Dinner left me full without producing the kind of rise I was trying to avoid.
Once those patterns repeated, I could experiment from a position of greater stability.
Then pizza showed me the other side.
The first slice produced a large rise.
The second similar slice produced a similar rise.
The walk appeared to change the next response.
That sequence taught me more than a list of approved foods ever could.
It taught me to separate observation from conclusion.
It taught me to compare a result with my own baseline instead of somebody else’s horror story.
It taught me that movement could be an input, not merely an attempt to burn off guilt.
It taught me that a low glucose reading did not certify the entire meal as healthy.
It taught me that one result could ask a question, but repetition had to help answer it.
Most importantly, it gave me more freedom without giving me false permission.
Freedom did not mean I could eat anything I wanted and use a ten-minute walk to erase it.
Freedom meant I had more than one lever.
I could change the food.
I could change the portion.
I could change the combination.
I could change the timing.
I could add movement.
Then I could observe what happened and decide whether the pattern deserved trust.
That was the first real breakthrough in repeatability.
I stopped asking, ‘Is this food good or bad?’
I started asking, ‘What happened when I ate this amount, in this combination, under these conditions, and does it happen again?’
The answer was not printed on one label.
It was built across repeated meals.
One reading gave me a clue.
Repetition gave me confidence.
Changing one variable gave me understanding.
And an unexpected result gave me the next experiment.
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.