Hi
So I’m focusing more on my morning fasts… I took my BG before coffee (BLACK only) after having 1/2 cup of coffee - I went from a 90 to a 97… Would you say coffee broke my fast?
Is breaking a fast mean any movement in blood sugar?
Thanks!
Checking Glucose number after coffee
SomeGuy
(Joey)
#2
@ketoartsy If all you had before vs after was 1/2 cup of black coffee and you measured glucose going from 90 to 97, then you’ve successfully demonstrated how large +/- testing errors can be … not your fault, but the limits of accuracy of measurement gear.
Besides, as many folks with continuous glucose monitors can attest, the serum glucose readings do a fair amount of wandering as a result of normal metabolic activity.
FWIW, I’ve measured my morning blood glucose both before and after black a full cup of coffee with my usual heaping tablespoon of 100% unsweetened cacao powder… it didn’t move a bit. More measurement error? Perhaps 
ctviggen
(Bob M)
#3
Not to mention that your blood sugar might go up regardless of whether you have coffee or not, which is what happens to me:
I have delayed drinking coffee until later in the morning…and I get the same effect. Any blood sugar increase due to coffee would be “on top” of this, and would be hard to determine, as if there is an increase, it seems to be small.
amwassil
(Michael - When reality fails to meet expectations, the problem is not reality.)
#4
I’ll add only that it’s also remotely possible that the caffeine stimulated a cortisol response which in turn triggered a GNC response. You could test this by drinking decaf for few days. Personally I’d go with range of meter error or your normal morning glucose rise. I likely drink more coffee than any 2-3 other people and have zip glucose response.
lfod14
#5
7 is beyond ignoreable, there is always a response to taking stuff in.
Is breaking a fast mean any movement in blood sugar
Not really, being stressed, getting pissed off, feeling pain, getting sick all plays with it. Especially if you’re fasting for fat loss, way more wiggle room than if you’re going for autophagy and even then there’s debates how strict you need to be.
SomeGuy
(Joey)
#6
@ketoartsy Cheryl - Based on the replies thus far, it would seem safe for you to finish up and enjoy that second 1/2 cup of your black coffee. 
ctviggen
(Bob M)
#7
I did measure my insulin after drinking 16 ounces of coffee and two teas, and even it was not that high:
PaulL
(Bacon is a many-splendoured thing)
#8
Most home glucometers have an accuracy rate of ±20%, so a reading of 97 could mean a true value of anywhere from 80.8 to 121.25 (or 77.6 to 116.4, depending how you interpret the phrase “accuracy rate”). In any case, 97 and 90 are well within each other’s range of accuracy and are hence essentially the same reading.
ketoartsy
(cheryl)
#9
Interesting… I just rechecked today … BG before 95… then after coffee it was at 86…
So, I am thinking it’s not the coffee…
Other fun facts… took BG at 8:00… I was at 91… so fluctuation seems to be what happens on it’s own or the meter has some fluctuation as well…
I might test again tomorrow… but am thinking I’m not breaking a fast with coffee black…
ctviggen
(Bob M)
#10
A 100% accurate meter is only accurate to within 15% of the actual value. That’s the requirement, and a lot of meters don’t meet that.
So, if your blood sugar is 100, and your meter shows between 85 and 115, that’s “accurate”.
I use these more as tools to see trends. But those can be hard to see if you get plus or minus 10 points every morning.
CFLBob
(Central Florida Bob )
#11
7? So if the real value was 93-1/2 that’s plus or minus 3.5 points or 3.7%? No production meter is that accurate.
Some time in the last couple of years, someone around here posted a paper comparing meters on the market at the time, (the paper is dated October 2016) and the best meter on the market had an accuracy of a little over 5% (the Bayer Contour Next). The rest went up to a high of 20%. The paper is, “Comparative Accuracy of 17 Point-of-Care Glucose Meters” from the Journal of Diabetes Science and Technology 1–9. I have the paper, but since I got it from here, I probably shouldn’t post it again. It’s probably somewhere around here. After reading the paper, I switched to the Contour Next.
For the OP, I don’t think there’s a universal definition for breaking a fast short of a four course meal including dessert and all the trimmings. Instead of black coffee, I have a teaspoon of heavy whipping cream in a 12oz mug of coffee and two mugs of that. Back when I was taking a prescription medicine that upset my stomach, I’d have as much as a tablespoon of butter in the evening. I’ve heard a limit of 400 calories a day to stay under and still be fasting. (I have that associated with Dr. Fung in my mind, but maybe not?)
PaulL
(Bacon is a many-splendoured thing)
#12
I believe the real question to ask is why one is fasting in the first place. It is my impression that Dr. Fung has indeed said there is a certain amount of calories one can have without inhibiting autophagy. And lean people sometimes have to have a certain number calories in the form of fat, if they want to fast. As long as the calories are minimal and don’t come from carbohydrate, I suspect that weight loss won’t be inhibited too much, either.
ctviggen
(Bob M)
#13
That’s the one I use, because of that paper.
MARD is basically how far from “actual” it is.
I still like CGMs the best, as they seem to be more accurate, relatively speaking. At least for trends, they are great.
CFLBob
(Central Florida Bob )
#14
Since I managed to prevent being formally diagnosed diabetic, a CGM isn’t likely, but I’d find it interesting for a couple of sensors. It would be interesting to test different foods and things and not have to stick a finger four or five times for every item.
SomeGuy
(Joey)
#15
I’ve read the paper/methodology and have to disagree with the impression I think is being taken from this discussion…
The Mean Absolute Relative Difference (MARD) is being calculated versus a laboratory standard. The the 95% confidence limits (shown in the above graphic as a high-low band above/below the result for each bar) shows a fairly narrow range of testing result for each device.
In other words, although the mean test result of each device can vary markedly from the laboratory standard, the repeatability of each device (consistently being different than the benchmark) appears to be rather high (given the narrow range around its resulting error).
If I’m understanding the write-up correctly, you can count on the Nipro SideKick, for example, to be 20% wrong quite consistently. So even though your glucose reading might be wrong, it’ll tend to be wrong (within 95% c.f.) by a similar amount each time you test the same blood.
Again, assuming I’m not misunderstanding the paper’s methodology, the ACCURACY of over-the-counter glucose monitor devices is pretty poor. But the PRECISION with which they produce inaccurate results is quite good.
Anyone care to tell me if I’m all wet on this conclusion? 
CFLBob
(Central Florida Bob )
#16
At the risk of boring too many people to sleep, (one of my main life skills) I’m going to use a frequently used analogy for accuracy and precision. It comes from archery or target shooting (which are essentially the same thing).
One could mentally add a fourth picture of accurate but not precise, which would be harder to draw, but it would be a pattern of dots that puts the average position of all the blue dots on the target center, but they’re spread out over a wider area. A circular pattern, say, five times the diameter of the center spot with all impacts equidistant from the center would be accurate on average but not precise.
The way I’m reading what you’re saying is that the meters seem to follow the middle image; they precisely, repeatably give a wrong answer. In the sports that use this analogy, inaccurate but precise then leads to knowing how to correct the shot so that it becomes accurate and precise. In the picture, you can see that “all they gotta do” is move their point of aim to move the group downward.
There’s a scatter plot of the readings on the various meters that show the spread of their measurements (Figure 2, page 5). I’m going to snip out the middle to show the low MARD (top) and high MARD meters (bottom) so you can see their precision.
Look at the top left and bottom left scatter plots. The top left is the Contour next, the best MARD and the bottom is the Nipro Sidekick. The scatter, or lack of precision, is pretty obvious.
With a BG meter, moving the reading to agree with a calibrated measurement just ain’t happening. First of all, we don’t know what the right answer is, unlike the target illustration. Even if we did know, we don’t have a knob to adjust the meter to move the precise grouping of readings down toward the target as in our picture, but what we want out of a meter is to be both accurate and precise.
The way I read that scatter plot, the meters with the lowest MARD are both more accurate and more precise.
SomeGuy
(Joey)
#17
Indeed, this is a precisely accurate description of what I was trying to say! …
… and taking your comments to heart, I now see that although what I was trying to say was said precisely, nonetheless it was highly inaccurate.
Many thanks for following up! 