@jackpark - I believe the sensors are identical.
@Agaricus
The overall benefit of such an effort is we could help QS community identify when they have insulin resistance (it is sometimes subtle and can be hidden by diet - it was in me) - and discover the best ways to reverse it thus helping prevent loss in vision, loss of kidney function, improve immune function, fitness and healing, and help reverse obesity.
One thought for tracking is that we could open up a shared Google spreadsheet. Each chart could represent a single food/beverage with curves from multiple people. Each row could be a unit of time. Each column a different person consuming that particular food with blood sugars recorded starting at time = 0. Then anyone can go in anytime and add their own data. Additionally, we can continually create new charts or new tabs with new food experiments. Each chart can have instructions how we consumed it (i.e. āentire bottle of Soylent drank within 5 minsā)
Ultimately these could be cross correlated with each personās HgA1C and 1 and 2 hour OGTT (oral glucose tolerance test) results representing various levels of insulin resistance.
Ultimately, for those who discover insulin resistance, the thing that might be ultimately useful is determining across the QS community how best it was reversed. Certainly HIT (high intensity training), fasting, better sleep, lower glycemic diets, aerobic exercise - all could make variable contributions - but one may make a difference better than any of the others. The best way to do this remains an open question - but these first steps might give us a better sense of the best way forward.
I think this type of approach could turn out to be highly useful. If I understand you right, we would be testing glycemic response to foods one food at a time.
Using the Freestyle Libre, Iām getting a measurement every 15 minutes. The readings are delayed by about 20 minutes, so that the 11:01 am reading appears in the file at about 11:21. I suspect there is a function to evaluate sensor readings at time t1 based on subsequent readings at time t2, t3, etc, but I havenāt gotten very deeply into this yet.
I think the first step would be for the two of us to try it. A couple of practical questions:
What will we be satisfied with for a āstableā starting point?
We would probably want to know the quantity as well as the type of food - I will have to get a food scale first.
1 Like
@Agaricus Yes, exactly. We would be testing glycemic response to food.
One goal to benefit the community is to determine what foods help keep blood sugar under 140 in order toā¦
- Prevent nerve damage: According to this articleā¦"University of Utah researchers studied people with painful sensory neuropathy, or nerve damage. They found that participants who had impaired glucose tolerance with glucose levels rising between 140 mg/dl and 200 mg/dl in response to drinking a glucose-rich drink were much more likely to have a diabetic form of neuropathyā¦The higher these OGTT numbers go, the more nerve damage is found, according to Johns Hopkins Hospital researchersā¦Glucose can also start killing beta cells at levels below 140. One study found that people with fasting blood glucose from 110ā125 (within the official āprediabeticā range) had already lost up to 40% of their beta cell mass.
- Prevent pancreas damage: āā¦Italian researchers found that even with glucose levels in the supposedly ānormalā range, beta cells started to fail. Ruhl says that researchers āfound that with every small increase in the 2-hour glucose tolerance test result, there was a corresponding increase inā¦beta cell failure. The higher a personās blood sugar rose within ānormalā range, the more beta cells were failing.āā¦Failing beta cells will lead to worsening diabetes, a truly vicious cycle.ā
- Prevent eye, heart, and kidney damage and strokes: āā¦Slightly elevated glucose has also been shown to cause eye damage (āretinopathyā) and increased rates of heart disease, kidney damage, and strokeā¦ā
Stable Starting Point: To standardize the data initially - we could start by testing for the first meal of the day (i.e. breakfast). In that way there is 6-8 hours of fasting to ācleanseā the system.
Foods to Try:
Here are a few ideas.
Breakfasts:
These are all gluten free.
-
Trutol. Not really a breakfast. It is the official OGTT (oral glucose tolerance test). This is the gold standard for testing glucose tolerance. It is 75 grams of Dextrose (D-Glucose) and tastes awful. Leave it to healthcare to create a drink of pure sugar that tastes terrible. I really donāt know how they managed it. Typically it is given to women during pregnancy to test gestational diabetes. I bought a carton and can send you some. We can list normal 1 hour and 2 hour clinical values for people in the spreadsheet. Additionally, we can test versus Kool-Aid, Tang, or soda like Coca Cola (it is equivalent to about 2 Cokes) - and see if the absorption curves are the same. In that way, we can figure out if people can avoid paying the $75+ to buy an entire carton of this beverage travesty.
-
Soylent. It is advertised as a low glycemic, full meal. I have tried it. My blood sugar curve looked like it almost matched Trutolās curve - but kept my blood sugar up longer. It might not be as good as it seems. In fact anything ground to a powder has so much surface area, it will probably have a high glycemic index.
-
Bobās Red Mill Extra Thick Rolled Oats Oatmeal (gluten free) 1/2 cup + 3 eggs. The thickest cut, lowest glycemic index oatmeal I have yet found. One minute in the microwave to cook. Salted and with eggs it makes āsavory oatmeal.ā
-
5 eggs (2 with yolk, 3 without) + 3/4 cup blueberries. Blueberries and strawberries reportedly have the lowest glycemic index of any food - but I have yet to test strawberries. Reportedly any fruit where you canāt routinely buy choose from it is considered low glycemic index. Some think mangoes might also have a low glycemic index because of the fibre. I am wondering if it will actually turn out to be just like eating candy.
Lunch or Dinner:
Meal kits might be an interesting and easy way to standardize blood sugar trends across a variety of foods and meals.
If you enjoy cooking or have wanted to learn, a simply great service for us has been HomeChef.com. (No financial relationship.)
My upcoming May 25th order includes these three meals:
- Jalapeno Popper-Stuffed Pork Tenderloin
- Chicken Chopped Salad
- Shrimp with Tiger Sauce and charred sugar snap peas
My June 5th order has these:
- Crispy Honey-Soy Barramundi (w/ ginger roasted radishes, chayote & Swiss chard)
- Piedmont Chicken Breast (w/ mozzarella, pesto, pesto, and balsalmic reduction caprese salad)
- Pork Chop with Pine Nut & Parmesan Butter
Each meal is ~$20 for two servings. Hereās my referral link that I think gives you $30 off: https://www.homechef.com/invite/yrKZ6f489M6n
Standardizing across meal kits would track blood sugars and meals in a way that people really havenāt done before.
Theoretically, if enough people did this, the data could help machine learning engines better predict the responses of blood sugar for diabetics. The blood sugar responses could also be interesting to those designing meal kits to adjust the glycemic impact to make them less likely to induce diabetes.
1 Like
The Libre and the Libre Pro should have equivalent accuracy.
I havenāt seen the Libre Pro on eBay. As I understand it, Abbott is clamping down because they are doing clinical trials to prove its validity for the FDA.
@Praxiteles Right now, I suspect Pro is off the table. Using Ebay as a source is also nearly off the table. I found one seller in the UK, and used a Contact Seller button to ask if the reader was English. He replied it was. I clicked the Reply button and was taken to a message at ebay that the message had an invalid id. And, Contact Seller is no-longer visible. Sellers on ebay tend not to tell the dates on the sensors, and ebay sends me messages with links to sensors which donāt exist. The entire process of getting started is being terribly problematic. If I had a reliable source of sensors, Iād pick up that reader and get started.
The breakfast ideas seem like a good place to start. Itās logical to run these tests fasting. Is there a reason (other than enjoyment of eggs) that there are so many eggs in breakfast 3 and 4? Iād be tempted to run test 3 on oatmeal only to give it a clear run, then test 4 on fruit only, then test 5 on a protein/fat breakfast like 2 eggs, olive oil, and cheese (or something like that).
From my experience so far with the measurements, most food related spikes wash out within 90 minutes, so if we donāt eat anything else for a couple of hours we out to see the whole curve pretty clearly, assuming that there arenāt other, non-food related reasons for spikes and dips.
Iāve been showing to changes in blood glucose at night, which are a bit mysterious. For instance, see below:
18675 2017/05/05 00:01 0 105
18676 2017/05/05 00:16 0 109
18677 2017/05/05 00:31 0 127
18678 2017/05/05 00:46 0 135
18679 2017/05/05 01:01 0 136
18680 2017/05/05 01:16 0 138
18681 2017/05/05 01:31 0 139
18682 2017/05/05 01:46 0 132
18683 2017/05/05 02:01 0 121
18684 2017/05/05 02:16 0 107
18685 2017/05/05 02:31 0 92
I was asleep during those hours.
Hopefully several of your test meals would support glycemic load calculations.
Digestive enzymes could alter the uptake in your digestive track.
You may also need to standardize the day before the test (diet and exercise).
Good luck
Agaricus, do you know what sleep stage you were in during your blood sugar changes?
I believe the rise in glucose levels in the middle of the night is a circadian rhythm thing (and a problem for people with diabetes, or jet lag). Do you see a similar pattern every night?
@OP_Engr I donāt know what sleep stage it was. I did a somewhat superficial look (PubMed, Google Scholar) on the topic of sleep stage and blood glucose but didnāt see anything very helpful. Is there something youāve seen you can point me to?
@ejain No. In the 12 nights for which I have data there have been two with spikes like this.
I donāt have a specific article. I was wondering about sleep state and hormone production (adrenalin during a vivid dream or growth hormone during fasting).
Any relationship of your blood sugar increase to noticeable events in a dream journal or diet the day before?
A morning saliva cortisol test could provide more insight (probably more questions than answers).
@OP_Engr - Both nights were during separate, short-term stressful 24 periods, so this may be a clue; it would be great to be able to do some hormone tracking but I donāt know any tech for high resolution hormone measurement.
Iām experimenting with using Fluxtream to look at my blood glucose data and food notes. Havenāt gotten very far yet, but my first glance at the data showed something interesting. I noted the time on a sudden wake up early in the morning (4:40 am), suddenly alert. Look at the blood glucose data, which I didnāt see until I downloaded it later in the day. My blood glucose went from 84 to 108 (red vertical line) in 15 minutes before peaking at 120. I wake up during this spike.
@Agaricus, if you notice that effect consistently (gradual decline in blood glucose, then a fast rise that corresponds with you waking up), I speculate itās adrenaline released in response to a perceived hypoglycemic state. Itās an idea Iāve gotten from the offbeat world of Ray Peat, but I think it can probably be supported independently (adrenalineās function for releasing glycogen is described here, for example).
Improving my thyroid function and keeping glycogen high seemed to dramatically cut my sleepwalking and nightmares, with I suspect was due to minimizing adrenaline release.
1 Like
@danwich Thatās an interesting theory - Iāve always been prone to sudden wakeups and when I was younger I did some serious sleepwalking.
I think the 15 to 20 min latency for this sensor makes it difficult to determine which came first (your awakening or the blood sugar spike).
Sleep apnea can also cause adrenalin spikes.
Do you have the ability to measure sleeping heart rate and respiration?
How about snoring? (The Snore Lab app has the capability of recording sounds while you sleep).
Any muscle pain/cramps?
Any possibility of external triggers (sound, light, etc.)
1 Like
Yes, I agree, questions about latency make minute by minute inferences dubious. Iāll post some more data as I get it. Although precise conclusions are probably out of bounds, there continue to be some quite clear patterns of high #s at night. I peaked last night at 152, about 90 minutes after going to sleep and about 4 hours after eating. I do snore sometimes but donāt suffer from sleep apnea. The idea that snoring could cause blood glucose spikes in intriguing and, if snoring does turn out to be involved, probably affects lots of people. (At this point I have no idea if my numbers are idiosyncratic or typical of some substantial subgroup of humans.) It seems like an easy experiment to do audio measurement of snoring and see if it shows up during the spikes. Minute by minute accuracy wouldnāt be necessary; snoring during the 1-2.5 hour window of the spikes would be a clue.
Add body position to my list of measurements.
Does body position change the sensor results?
This sounds like classic Dawn Phenomenon. I saw the same thing each morning between 4am-7am while wearing a Freestyle Libre CGM for two weeks. I did a short summary with my charts. (I was following a cyclical ketogenic diet at the time).
1 Like