I like the new layout of Glooko

It’s been a while since I performed an upload into Glooko, but I did so just now and noticed a new layout. Not sure if it’s new or not (it’s been a while since my last upload), but I like the information layout. I don’t recall seeing the below in the old format.



Wow, Liam really uses very little insulin compared to Samson nowadays. His programmed basal has been ramping up progressively, to about 5.3 units per day. And his bolus is typically about 4 to 6 units too. He eats about the same number of carbs – somewhere between 75 and 125 g of carbs most days.

1 Like

Not sure how, or why…but I guess even though it’s 2.5 years later, he’s still honeymooning…no other explanation for it I guess.

As for his carbs, they just end up being whatever they end up being. We don’t track carbs at all…we just make sure we bolus for w/e he’s eating. He’s not restricted at all…anything his brothers get, he gets (that includes cookies, cakes, just whatever they are having…I don’t want him to feel “left out”). So, as we’ve talked about…I just figure out how to “make it work”…it’s tough sometimes. lol

1 Like

yeah I mean we don’t restrict carbs but I think it averages out that he eats about the same number per day. Which i guess makes sense – they’re the same age and probably eating about the same number of calories every day to grow.

1 Like

I wish Glooka would display A1C estimate as the Dexcom site does.

That would really be nice.

Kind of a thought on this whole thing.

I don’t know how anyone does A1C estimates. But in general, if an estimate is only averaging BG’s they are missing something.

A very important factor is the length of a high BG.

So if I were coming up with an estimation algorithm, I would look at things like this:

If you have a 250 and a 80, that averages out to 165.

But being 250 and then the meter showing a BG of 80 an hour later, is much different than being 250 and the meter showing a BG of 80 4 hours later. (One of them is a short span of 250, while the other implies a much longer high and more time out-of-range.)

I assume Dexcom does this sort of thing with their A1C estimates, looking at times. But do any other meter thingies (phone apps, meters, glooko’s, etc.) do estimates in a smarter way than simply averaging BG’s?

@Chris, do you happen to know?

This accu-check site does an A1C estimate, but it’s a simple average BG calculation, so it’s kinda worthless:

1 Like

Agreed with this, but equally important to me is the length of a LOW BG…if you have a great A1C because you live in the 30’s, that’s no beuno.

1 Like

My underataning is most of these use the formulas like…

eAG(mg/dl) = 28.7 × A1C − 46.7
eAG(mmol/l) = 1.59 × A1C − 2.59

Where eAG is estimated average glucose. Now you have to do some algebra to go the other way… :grinning:

Nightscout says this…

This is only a rough estimation that can be very inaccurate and does not replace actual blood testing. The formula used is taken from:Nathan, David M., et al. “Translating the A1C assay into estimated average glucose values.” Diabetes care 31.8 (2008): 1473-1478.

1 Like

Sure, that isn’t great either.

But the problem with A1C is that it doesn’t have any way of knowing this.

I have had endo’s tell me my A1C indicates I am having too many lows. But that is just an assumption they make. So I try to explain to them in apples and oranges.

You give me 20 apples at the beginning of the week, and at the end of the week, I have 18 apples. So you yell at me, “You are eating too many oranges!!”

But you don’t know how many oranges I ate. You don’t know anything about oranges at all. You only know that I ate 2 apples.

That is always the assumption endo’s make with the A1C.

1 Like

@Eric The only way you can get a reasonable estimation is with CGM. I think Glooko uses (or can use) Dexcom data, as captured from Clarity. So it should be the same estimation as Clarity.

XDrip+ gives me an estimated A1c for 1,2,7,30,60, or 90 days. It’s generally pretty accurate with the lab reading.

1 Like

*Assuming your CGM is accurate…ours is off more times than it isn’t unfortunately. So I actually don’t base any “real” evidence off of the CGM data. I base it off of the PDM data (Glooka) because those are at least actual blood checks.

Glooka uses the data directly from the PDM (at least in our case). I think there’s a way to connect the two, but I’ve not looked into it because I don’t want to mix reliable data (PDM) with unreliable data (CGM).


Do you know the formulas they use? I assume they use some type of “time” at different BG’s to factor it.

If someone does enough BG tests, you could also come up with some more reasonable estimates. At least better then the current calculations they use with only “average BG”. Certainly there is some way to look at time between BG tests to come up with better calculations.

@ClaudnDaye Good caveat! Garbage in, garbage out.

1 Like

@Eric I think they get the average (of every 5 mins) and calculate from there using a formula like the one @Aaron pointed out.

1 Like

To test a theory…

My 90 day mean glucose from xdrip+ is 5.89 mmol/l.

Xdrip+ says my estimated A1C is 5.3%.

eAG(mmol/l) = 1.59 × A1C − 2.59

So… plugging in my eAG AND solving for A1C…

A1C= 5.333333333333333% which matches what xdrip says to the first decimal place…


xDrip+ uses the following to calculate a1c:

public String getA1cDCCT(){
    if(getTotalReadings()==0) return "A1c:?%";
    return "A1c:" + (Math.round(10 * (avg + 46.7) / 28.7) / 10d) + "%";