I was very excited when I first ran across the glucose-ketone index (GKI), as I'm a bit of a data junkie. I'm not yet enough of a data junkie to shell out for a keto meter, especially given the cost of the strips.

There are some implications to the GKI that are interesting, but those little charts out there that tell you whether you're in weight-loss mode, or the best mode for diabetes or in therapeutic mode, are a bunch of nonsense; I can find no research to back this up except for research on a particular form of brain cancer. That particular cancer feeds on glucose, whereas most of the brain can use ketones, so it made a lot of sense there. But all these little charts you see telling you what the best GKI is for your purposes are largely just a keto myth.

But let's back up and discuss how to do the measurements in the first place before exploring what they actually mean.

bG meters


The bG meter is so common now, it's hard to remember that they are fairly new; when I was diagnosed with T2, only doctors could buy a bG meter. Now you don't even need a prescription, except to get insurance to cover it.

If you're not diabetic, you have to pay out of pocket. Some meters are actually given away by manufacturers with coupons - free! Seems like a great deal, until you see what the test strips cost. Many test strips are over a buck each :(

If you're interested in testing bG and have to pay out of pocket, I would recommend either the Walmart or Walgreens generic meters as the strips are much more affordable.

At this writing, you can get the Walmart Relion meter, with a lancet device, lancets and 25 test strips for around $50; additional strips run under $30 for 100. The equivalent Walgreens True Metrix kit is only $7 right now, with 100 test strips under $15. Check to see which is the better deal before you order.

What I really recommend for bG testing is the Freestyle Libre system. Unfortunately, it is currently only available by prescription (though I'd question a doctor who was unwilling to write a prescription for it, even if you're not diabetic, why not?) The system involves a sensor which you attach to your arm and works for 2 weeks (no more finger poking!) and a reader that you swipe to get specific bG readings (though it reads continuously). Unfortunately, if insurance doesn't cover it, both the reader and replacement sensors are rather expensive, the reader running around $80 and the sensors around $40.

Though my Medicare covers regular bG test strips for free, cheap as I am, I pay the $18 copay to get my sensors every month. I get *much* more information from my Libre than from a regular bG meter. For example, some meals might not spike my bG too much on a postprandial test, but my bG will run 15-20 points higher the next day. I did not know my bG ran under 70 many nights without my Libre. I am much, much more informed about what my bG does than previously.

I wish I could give one to everyone whether diabetic or not; my Libre keeps me accountable. When that devil on my shoulder is telling me that no one will know if I eat a cracker or cookie, I know my Libre will know.

ketone meters


There's 3 basic ways to check your ketones, depending on how accurate you want to be.

The first are ketostix, which just test your urine, so no poking required. However, they really only tell you if you're in ketosis or not, that they are darker for more ketones doesn't mean you're making more, as hydration has much more to do with how dark they are than how many ketones you're making. Also, as you become more efficient at using ketones as fuel rather than glucose, these strips may even indicate no ketones in spite of you making lots.

The second is a breath meter. This is much more costly than ketostix and I'm not sure how much more accurate it is; again you are only measuring ketones being lost, not those your body is making, so your results will decline as your body is more efficient at using ketones. On the plus side, there aren't ongoing costs like with a blood meter.

The only way of measuring that is accurate enough for the GKI to have any meaning whatsoever is to use a blood ketone reader (I had a tongue-in-cheek joke here about buying it off my wish list, and someone did. Thanks Anna!) Given that the strips are the most expensive part of using one and they sell them via a subscribe-and-save model on their web site cheaper than anyone else, this would be my choice. This or a similar meter would be what you'd need to calculate a GKI.

how to calculate a GKI


In the US, we measure bG in mg/dL, but everyone else in the world measures it in mmol/L; even when we're using metric, we do it wrong! So the first step is to convert our measurements to the right ones, which is to divide by 18; e.g. a 102 mg/dL reading is 5.66 mmol/L.

Ketones are measured in mmol/L everywhere, so no conversion needed there.

The basic idea is your body needs fuel. Many tissues prefer and will use ketones as fuel, though a few tissues require glucose. So overall, the idea is that the more ketones and less glucose you have, the better.

Note:It is important to know if your bG is elevated, over 240 mg/dL or 13.3 mmol/L, you should not have ANY ketones. If you do, even on ketostix, then you may have DKA, a life-threatening emergency. Get thee to an ER NOW!

The GKI is calculated: glucose level (in mmol/L) ÷ ketone level

I think knowing your number is useful in that you can know if you're doing better or worse over time; if I weren't so frugal, I'd probably follow mine, but as it is, I extrapolate from bG to decide how well I'm doing. As of this writing, my bG usually runs 65-85, so when it's running 75-95 for a day, or I see any number over 100, I know I need to buckle down.

But the charts that tell you what mode you're in are somewhat silly. This chiropractor never scores properly except when doing extended fasts, yet has been steadily losing weight for a year.


What about exogenous ketones?


IMO, raising your GKI via exogenous ketones is not very worthwhile, unless you're specifically fighting cancer and need a ridiculous GKI to extend your life.

For me, the goal is not having a good GKI; the low glucose from reversing diabetes and the elevated ketones from burning my fat are goals.

Exogenous ketones may be useful for getting into ketosis faster and teaching your body to use ketones as fuel, especially at the beginning. But your body can do this perfectly well without buying any products.

I know of nothing that will get you there faster than an extended fast followed by eating like a bear.

But if you really want to spend money on this, here's some highly rated exogenous ketones.

What are the implications of the GKI?


According to my Libre and the FDA, 70-140 mg/dL is normal bG.

I have not found this to be true for myself. My bG often goes below 70 since I began this way of eating/not eating, but without hypoglycemic symptoms. I have entire nights when it stays below 70 for hours at a time and even a few mornings when it stays there.

It seems that bG can drop below 70 for many folks, even on no meds. When you're in ketosis, you have both ketones and glucose in your blood available for energy; so your body doesn't consider it an emergency if glucose lowers because it has the backup ketones available. We're trained to think bG under 70 is dangerous, which it is when you're not in ketosis.

On the few ocassions I've had "hypoglycemic" symptoms, my bG hasn't been below 70 and eating some salt has fixed me immediately (see electrolytes).

If I had symptoms that salt didn't fix or my bG was actually dropping, I'd consider that something to be corrected. But since my last insulin shot was Sept 22, 2019 and I'm not on any other meds that can cause hypoglycemia, that's highly unlikely to happen.

It seems very normal that folks in ketosis can get bG readings under 70, especially when fasting. It's a common experience in the ketogenic world.

On the other hand, I consider 140 to be WAY too high. This is weird because a year ago on 100 u TDD insulin/day, my bG was nearly always over 200, often spiking to over 300 and I'd have been thrilled with bG at 140.

But my new normal is 65-85, so I have an entirely different scale today. When I discovered that freaking cottage cheese spiked my bG over 100, I kicked that stuff to the curb; I can eat way better protein than that. Some foods give me a minor reaction, but the my bG runs 75-95 the following day.

I'm not yet "all better," pretty sure I'd still fail a GTT today. But the healing progresses with low bG and high ketones and that actually works without needing to know my GKI.