Listen to this story



Fasting and Ketones: Data from 3 Days with No Food, Only Wine

Mark Moschel
Jul 26, 2017 · 8 min read

I’m obsessed with eating.

Probably addicted. I do it multiple times every day.

I finally had the good sense to quantify just how obsessed I am. I collected lots of data, ran that data through a complex regression algorithm, and plugged it into a state-of-the-art graphing application.

Here’s what popped out:

That graph shows very clearly that, yes, I do appear to be obsessed with food. That’s especially true as the day goes on.

Below is a story that I originally told at my local Quantified Self meetup. The data and graphs get a lot more serious than the one above. And even though I love eating as much as the next person, I am going to talk about abstaining from food.

Two Insights

This article is a continuation of an experiment I did last year .

[Editor’s Note: The goal of a ketogenic diet is to achieve a metabolic state that emphasizes burning fat as the primary source of energy. It is often marked by the presence of ketones in the blood. This family of diet is most often performed either with regular periods of fasting or by restricting carbohydrates, as with the slow-carb diet or Atkins.]

In that article, I shared data I collected while following a ketogenic diet for three months. Two insights from that data inspired this new experiment with fasting.

Insight #1: I Felt Better When My Ketones Were Higher

The following graph compares my average ketone readings (more than 100 samples) to how I felt at the time of testing.

  • On average, when I felt like a five (SUPER HIGH ENERGY!!), my ketones were at 3.4.
  • When I felt like a three (normal energy), my ketones were at 1.4.
  • When I felt like a one (low…en…er…gy……), my ketones were at 0.5.

This shows, on average, that I felt better when my ketones were higher.

Insight #2: My Ketones Were Highest at the End of Intermittent Fasts

The following graph shows four consecutive days of ketone samples. You can see that the measurements peaked in the early afternoon toward the end of my intermittent fast.


  1. If ketones are highest at the end of a short-term fast, and
  2. I feel best when ketones are highest, then…

What would happen if I extended the fast? Would my ketone levels continue to increase? Would I feel SUPER AWESOME?!

First Experiment: Three-Day Fast

To test this theory, I fasted for three days while measuring my ketone and glucose levels using a . If you buy one, remember that you also need to buy strips and lancets. The whole purchase should run about $120.

Let’s look at the data.


My ketone levels trended upward each day. Here are my average readings:

  • Day 1: 1.5
  • Day 2: 2.2
  • Day 3: 3.0

My ketones peaked at 4.4 in the afternoon of day three. Surprisingly, my ketones remained high even a day after the fast ended (day four in the graph).

Another notable finding is that my ketones dipped after each weightlifting session.


The next graph adds my glucose readings (I started tracking glucose on day two).

In this graph, glucose is often high when ketones are low, and vice versa. This implies a possible inverse correlation between the two.

Also, similar but opposite to ketones, my glucose spiked during weightlifting workouts.


This graph compares my ketone levels to how I felt in terms of energy. I recorded my energy level before taking each blood sample using a scale of one to five:

  • 1 = [yawns] I want to lie on the couch and [yawns again] do nothing but watch Black Mirror.

My energy was surprisingly highest on the last day of the fast.

I averaged a 3.0 on days one and two and a 3.8 on day three. Dips happened each night, probably because I was tired and couldn’t eat food to boost my energy.

This graph also shows a possible correlation between ketones and energy. There are too many variables to know with any certainty (subjective measurements like “how do I feel” have so many confounding factors), but it’s another data point that aligns with data from my prior experiment.

Lessons from Fast #1

  1. My ketone levels increased each day of the fast.
  2. A heavy workout decreased ketones and increased glucose.
  3. Ketones and glucose seemed to have an inverse correlation.
  4. Ketones and energy seemed to have a direct correlation.
  5. Ketones remained high even a day after the fast ended.

This data is promising!

During these 72 hours of abstaining from food, my glucose stayed fairly stable, my ketones steadily rose, and my energy increased toward the end of the fast.

But this was just one short experiment. If I repeated the experiment, would I see the same results?

Second Experiment: Three-Day Fast (+ 🍷!)

The second experiment was another three-day fast. But this time, there was a twist.

You see, I’m in the wine business now. This means I drink wine almost every night (occupational requirement). I follow a ketogenic diet and do frequent blood testing, so I know the wine I drink doesn’t kick me out of ketosis. However, I always drink wine with food.

Naturally, I was curious: What would happen if I drank wine ONLY while not eating food?

Like any good biohacker, I decided to experiment (for science, of course).

My hypothesis: I’ll get a super-awesome buzz!

I started another three-day fast, but this time, after two days of fasting, I sat down with a bottle of wine and started drinking. I tracked how much I drank and how inebriated I felt.

At the end of the night, I collected all this data, ran it through multiple complex algorithms, and graphed it.

On a side note: I am quite impressed with this new graphing application I’ve been using.

Here’s what I found:

My level of inebriation is measured in a technical unit known as a “buzz.”

After the first glass of wine, I felt great!

And then things quickly became a little blurry. There were some stick-figure drawings, a round of pushups, and eventually falling asleep at the kitchen table.

Overall, a very good time.

The wine in my experiment was a natural wine from the Loire Valley in France. It’s one that we’ve sold at Dry Farm Wines, so it has been lab tested. I therefore knew it was organic, sugar-free, 12 percent alcohol, very low in sulfites, and free of any additives. I would not have done this experiment with commercial wine, and I would strongly discourage trying this experiment with any wines that aren’t sugar-free or guaranteed to be clean.

The Real Data

Here’s the hard data, starting with my ketone readings.


This graph has a similar shape to the first experiment’s ketone graph.

Ketones dipped after working out, peaked in the afternoon on day three, and stayed high even a day after the fast ended.


This data is also similar to the prior experiment.

Glucose typically goes up when ketones go down, and vice versa.

When Drinking Wine

Zooming in on day two, we get a closer look at the data around each glass of wine.

After the first glass, there was no change in ketones or glucose. After the second, ketones dropped a little and glucose rose a little. After the third glass, ketones dropped again (though still comfortably in ketosis at 1.4) and glucose actually dropped a little (not sure why). Another blood test a couple hours later (around 2 a.m.) showed that ketones were still at 1.4 and glucose had risen slightly to 82.

When I woke up the next morning, my ketones were back to 2.3 and glucose was 70. By early afternoon, ketones were in the fours and glucose was down to the 40s.


Comparing my ketones to how I felt, there is a similar correlation between the two, as we saw in the first experiment.

One notable observation is that my energy dropped when my ketones were their highest (approaching a 5.0 ketone reading). I’m not sure if that’s an anomaly or if 5.0 is my threshold and anything above that is too high for me. I don’t have enough data to know.

Learned from Fast #2

  1. This experiment supported all the same lessons from the first fast: Ketones increased each day, ketones and glucose showed an inverse correlation, ketones and energy showed a direct correlation, ketones dipped after workouts, and ketones remained high even after the fast ended.
  2. Drinking three glasses of wine while fasting moderately affected my ketones and glucose, which quickly recovered the next morning, and created a very gnarly buzz (a technical term).

What’s Next?

What would happen if the fast extended beyond three days? Would ketones continue to rise into days four and five? Would they eventually level off? Would I continue feeling better each day, or would my energy start dropping?

These are my lingering questions. The next experiment will be a longer fast.


  1. Read the original ketone experiment:
  2. Learn more about the ketogenic diet: (plenty of stick-figure drawings)
  3. Read more about fasting:

Better Humans

Better Humans is a collection of the world's most…

Thanks to Niklas Göke

Mark Moschel

Written by

Partner and Health Evangelist at @DryFarmWines. Aspiring writer with 3rd-grade drawing abilities. @Bulletproofexec conference emcee. Previously CTO @Factor75.

Better Humans

Better Humans is a collection of the world's most trustworthy writing on human potential and self improvement by coaches, academics, and aggressive self-experimenters. Articles are based on deep personal experience, science, and research. No fluff, book reports, or listicles.

More From Medium

More from Better Humans

More from Better Humans

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade