Rebooting with a Juice-Fast. As a person with Type 1, what are the effects?

Rebooting with a Juice-Fast. As a person with Type 1, what are the effects?
Rebooting with a Juice-Fast. As a person with Type 1, what are the effects?

Having let the good life get to us a little too much, my partner and I decided to try a “juice-fast” reboot. If you’ve seen “Fat, Sick and Nearly Dead…“, this approach is basically to eat only juice for a period of time, in order to “reboot” yourself. In our case, we decided to use a pre-prepared one we could buy, as it was easier than preparing everything ourselves. For this, we decided to take on a three day option from Press. The details can be found here.

This is a restricted calorie, 3 day, juice and nut milk composition. When I say restricted calorie, I mean 955 calories per day, and given that they missed one of the juices in the pack, more like 855 calories per day.

So first up, this is hard work. By any stretch of the imagination, 855 or 955 calories a day is not a lot, and it has a commensurate effect on your weight. Over the course of the three days my weight has done this:

That’s a drop from 96.3kg to 93.1kg in three days. Some of this will be water weight, but it demonstrates, to some extent, why the Newcastle diet works well in T2.

These types of effect are well documented in various circles, so for me the interesting question was “What would the metabolic effects be in T1?”. Well, the Nightscout traces are below:

The first thing I noted was that my daily insulin needs were fairly low. Given my daily carb intake was, according to the bottles, around 90g (not all noted to OpenAPS, for reasons I’ll go into shortly), this seems reasonable. However, there are some anomalies on this.

The two nut milk concoctions stated that they should have 18g carbs per bottle, but my glucose reaction suggested somewhat differently, with noticeable spikes and OpenAPS having to work hard to handle it. Likewise, on the first day, I rather rage bolused, which wasn’t helpful. Some of the other carb counts also weren’t tremendously useful, so I was playing it a little by ear to try and manage how this worked. Of course, the old adage of “Rubbish in, rubbish out” meant that OpenAPS also struggled a little, and in some cases I didn’t tell it about supposed carbs and in others, nudged it with small boluses to try and help out.

Let’s just say that it was more difficult than I expected, with reaction not matching carb calculations, and you can see this in the data below:

Still, a standard deviation of 1.8 on an average of 6.0 isn’t something to complain about.

Now, if you read things like Mango Man’s website, he advocates a high carb, low fat, vegetarian approach to T1 to give huge increases in insulin sensitivity, when his 60 day program is followed, so I’d expect that a high-ish carb (considering the calories) diet would provide some indication of whether this might work.

My variation in Insulin Sensitivity over the fast period looks like the below:

What this suggests is that over the period of the reboot, my insulin sensitivity was improved, (bearing in mind there is a time lag) and that it dropped to my normal, with exercise levels, (whilst I was doing no additional exercise) when consuming very few calories and most of them from carbs. Now a three day reboot is not enough time to confirm whether the full program that is described by Dr Khambatta would significantly change insulin sensitivity, but it certainly raises question about what might be found, and I may undertake a longer version of this to see what happens.

The other thing I thought worth monitoring was ketones. A lower calorie diet means a calorie deficit, and if I’m not getting enough energy to keep my body moving, it has to come from somewhere. My initial suspicion was that I would see ketones, as this process is partly about reduced calorie and therefore starvation. At the end of day 1, I showed 0.2 mmol/l, day 2 was 0.4 mmol/l and day 3 was 0.5mmol/l.

These levels are consistent with what the Volek & Phinney charts show to be entering Nutritional Ketosis.

With the amount of carbs being consumed, I question whether this is the case, but a longer trial would give a better demonstration of what was going on.

In association with this, I’m not going to pretend that the first two days of this process were fun. By the end of the first day, I had a raging headache, which repeated itself at the end of day two, and woke up on days two and three with muscle aches that were consistent with what is experienced in DKA. By the end of day three, it felt more as though I’d reached a level of equilibrium, so that did at least help the situation. In order to feel full, I also drank a lot of water throughout the three days.


I can understand where the term reboot comes from, but I’m not going to pretend that it’s easy or fun. What it certainly is is interesting. What I’ve taken away from it is:

  • It seems like a very quick way to lose weight, although what happens afterwards is obviously dependent on what you eat.
  • The initial observations suggest that it may very well improve insulin sensitivity, but I’d need to do further observational study to see about that.
  • The carb counts in the Press pack may be correct, but the low calorie state potentially means that your body reacts differently to what is expected, making managing it significantly harder than expected.
  • It reinforced the “be careful with corrections” mantra relating to OpenAPS and SMB.

This is another of those areas where a clearly documented study would be incredibly useful, where a cohort using OpenAPS would be recruited and split into three groups. One would be eating an “Eatwell” type plate, one a low carb diet and one a low fat diet.

Each of these would be rotated through, and the metabolic indicators captured, including blood lipids, Hba1Cs and glucose, food and OpenAPS data (including sensitivity). A comparison would then be made between the different approaches, and a wider set of conclusions could be reached.

Until then, however, I’d say that a “Juice Fast” is a good way to lose weight quickly, rather like the Newcastle diet. It’s the follow up to this that would be more important and provide the real benefits.

1 Comment

  1. The consumption of fat and protein do affect my BG a lot when I stick to 15g CHO/meal. My insulin sensitivity diminishes, but the BG doesn’t spike near as much. 1) Is there some reference or ratio for insulin dosing where most calories are from fat or protein… 2) When not on the low carb diet, My BG seems to ramp almost linearly upward (very fast) or linearly downward (half as fast) – not a normal decay rate. 3) However, at higher BG levels I suspect the meter is not reading as high as BG really is. I’m at 2:1 sensitivity longer acting when high and 1.5:1 sensitivity shorter acting when normal BG. 4) The insulin absorption rate was supposedly very fast for those high-pressure spray pens that didn’t use a needle but jetted a spray through the skin. Likewise, when I inject in an overly used area, I will get zero absorption. I’d like you to hear your comments on the possibility of the closed loop feedback models to use these “constants” as potential variables: insulin adsorption and BG decline rate and duration rather than IOB type of calculatons.

Leave a Reply

Your email address will not be published.