Improving my Sleep with a Glucose Monitor

Do bad glucose levels lead to bad sleep?

Does a bad night sleep impact blood glucose levels the next day?

At the recent Quantified Self Amsterdam conference, we had a workshop on metabolism and sleep. During the workshop, we got a chance to meet and talk with type 1 diabetic patients who have been using continuous glucose monitors for years — and know deeply how sleep and glucose levels are related.

It turns out there’s so much more to glucose than just what we eat — sleep is a fundamental part of the equation.

Measuring Sleep

I’ve measured my sleep with the Fitbit and the emFit for the last year. Both are great — giving insights into sleep, and giving a history of times slept/how well we slept over time.

Fitbit Charge HR 2

While the Fitbit is great for starting out looking into sleep, the emFit is a prosumer device, giving more detail with minute-by-minute heart rate, Heart Rate Variability (how well rested the body is — see below section for more) and breathing patterns. The emFit is even used for medical research and for managing chronic diseases like epilepsy.

Does Blood Glucose Impact Sleep

Some of the worst nights sleep I’ve had is when my glucose levels have gone too low during the night (hypoglycaemia).

Glucose Over the night. Hypoglucemia is under 3 mmol/L

I’m restless. I’m cold. I wake up with a headache and feeling groggy the next morning.

When blood sugar goes too low, the liver will release new glucose, but it looks like this only happens when I’m awake. In above, the levels only went back up when I woke up.

This looks like what’s called ‘rebound hyperglycemia’.

We talked about glucose levels impacting sleep during the workshop in Amsterdam. Some of the type-1 diabetics in Amsterdam would sometimes load up on carbs before bed to improve sleep.

Seth Roberts has done a lot of self-experimentation with glucose and sleep, and has tried taking honey before bed to improve sleep.

For me, I feel like I’ve slept better when glucose levels are higher.

From the data, there’s a sweet spot of blood glucose to sleep score — at around 5. If it goes lower or higher, the sleep score is worse.

Sleep Score vs Overnight Average Glucose Levels (mmol/L)

Does Bad Sleep Impact Glucose Tolerance?

There’s a lot of research on how sleep deprivation can lead to insulin sensitivity.

At the QS conference, the type 1 diabetets have noticed big impacts on glucose control after little sleep. Gary Wolf also noticed in his show & tell talk.

I can’t see any patterns in my own data. This is probably because I don’t have enough nights with little sleep in the data. I haven’t been travelling much recently, no jet lag, few late nights.

I’ll update the analysis later if that changes.

Glucose and Sleep Stages

I’m seeing some big dips in glucose levels and was wondering if they’re related to sleep stages.

Glucose Levels Overnight

In above chart, glucose levels go low over a few hours, and only shoot back up again when I wake up about 4:45, when the liver kicks in and increases the glucose levels in the blood.

Fitbit recently updated their app to give sleep stages, and it’s nice. For the price, it’s one of the best on the market.

From the Fitbit chart, the drop seems to line up with a REM and deep sleep phase — starting about 03:15.

Sleep Stages on the Fitbit

The emFit and Fitbit give similarit’s sleep stage graphs.

I’ve been looking around online, and there does seem to be some research saying that the brain uses REM sleep phases use a load of glucose.

Note about tracking sleep stages — while the Fitbit and emFit are accurate for tracking ‘time asleep’, and how restless your sleep was, they may not give accurate figures for sleep stages (REM, light and deep sleep), even if that information is in the app.

Unless a device is tracking brain waves (the Zeo Sleep Manager — now out of business used to track brain waves), it’s not possible to get an accurate measure of sleep stages.

What the emFit and Fitbit give here is just a ‘best guess’.

Glucose goes up in the morning

When I wake up in the morning, glucose levels shoot. This is probably because waking up triggers a cortisol release, which then signals to the liver to release glucose.

I get another boost when I drink a cup of coffee. Again likely due to the liver releasing more glucose.

The ‘dawn phenomenon’ is a symptom of diabetes, where glucose levels go high a few hours before waking up. It happens when cortisol triggers glucose to be released by the liver.

For me, it’s rare, only when my glucose levels go too low overnight.

High Glucose Makes me sleep after lunch

Something I noticed recently. The bigger the post-lunch glucose spike, the bigger the crash I feel when the glucose levels come down.

The combination of no activity after a high carb meal spike my blood glucose the most.

Glucose and Heart Rate Variability

Heart Rate Variability(HRV) is a measure of how stressed the body is.

HRV is lower When we’re tired after a gym session, when we’re sick or when we’re stressed HRV is lower. When we’ve recovered well, HRV will be up (more on HRV here.)

Measuring the difference between the evening HRV and morning HRV gets a good indication of how well recovered during our sleep. emFit gives this value.

Good Recovery on the Heart Rate Variability Overnight on the emFit

From the data, low blood glucose overnight impacts HRV recovery. Again some of the worst nights HRV values have been at the same time as low glucose levels.

What did I learn?

· Hypoglycemia is very bad for sleep. Maintaining stable glucose levels during the day is important.

· Too much glucose is also not good. Getting it into a stable range is most important.

· There’s a strong signal between HRV values and glucose levels.

· For me, bad sleep doesn’t seem to impact my glucose control. However — this is likely more related to lack of data. This is something I will investigate further.

Having hypoglycemia stages during the night is worrying. It’s a symptom of diabetes, so it’s something I need to take seriously.

Getting this data has changed my behaviour. I’m no longer skipping meals, intermittent fasting. From the data, it hurts my sleep quality. I might revisit later when I know more, however.

But most of all, drawing conclusions here is hard. There are so many factors going on. Six weeks of data is not nearly enough to draw anything conclusive. But it does give us useful information.

Where Next?

Lots more research. Figure out what other factors are involved in sleep. There’s so many glucose metabolism processes going on here. Likely I’ve only touched the surface and I’m missing big parts of the picture.

More analysis on the data. These devices produce so much data. I’ve only just scratched the surface in the analysis. Sleep stages and movement overnight looks really interesting.

I’m also reading Mark Hyman’s excellent book ‘Blood Sugar Solution’ is giving me great ideas on how to get my glucose levels under control, and it’s giving me some great info on where to go next.

Blog Post Series

This post is the second in a series of blog post where I’m digging in deep into how glucose is impacting my life.

  1. First four weeks of continuous glucose monitoring.
  2. Glucose and sleep.
  3. Glucose and activity.
  4. Glucose and food.
  5. Glucose and stress.
  6. Lowering my HBA1C

Links