Experiment: dropping Metformin

[Eddy Edson]

I'm experimenting with dropping Metformin. I wanted to keep taking it, for all the possible anti-inflammatory etc etc etc benefits, even though my BG has been "normal" for many months. But then I came across research indicating that it can sometimes interfere with exercise benefits, which led me to also think about whether it might be increasing lactic acid build up when I walk a lot, an issue with my peripheral artery disease. So let's see if I notice anything different when I drop it; I can always go back if my BG spirals out of orbit.

I've really been tracking only my waking BG for the last few months. It's always in the range 4.6 - 5.2. What to expect without the Met?

Met on average seems to deliver an HbA1c reduction of ~10 mmol/mol, which should correspond to an avg BG reduction of ~1.5 mmol/L. But there's a lot of variability and it becomes less effective as baseline BG reduces, though I don't find much research on that. So .... no idea!

I dropped the Met a couple of days ago and it should have mostly cleared out of my system by now. So today is Day 1 of the experiment, and waking BG = 4.8. No difference there. Let's see how that develops & whether I notice any exercise differences.

 

[Northerner]

Good luck Eddy 🙂 I think metformin takes a couple of weeks rather than a couple of days to 'clear' the system so the consequences may not be apparent yet.

 

[Eddy Edson]

Good luck Eddy 🙂 I think metformin takes a couple of weeks rather than a couple of days to 'clear' the system so the consequences may not be apparent yet.

I thought that too, but looking into it, I was surprised to see this:

Elimination

Renal clearance (see Table 4) is approximately 3.5 times greater than creatinine clearance, which indicates that tubular secretion is the major route of metformin elimination. Following oral administration, approximately 90% of the absorbed drug is eliminated via the renal route within the first 24 hours, with a plasma elimination half-life of approximately 6.2 hours. In blood, the elimination half-life is approximately 17.6 hours, suggesting that the erythrocyte mass may be a compartment of distribution.

https://www.drugs.com/pro/metformin-extended-release-tablets.html

 

[Docb]

Keep us posted Eddy. I've never been convinced that Metformin did anything to my blood glucose levels although I have never been able to do an experiment unconfunded by other factors.

After a week without gliclazide I am convinced that that medication had had a significant effect.

 

[Eddy Edson]

Keep us posted Eddy. I've never been convinced that Metformin did anything to my blood glucose levels although I have never been able to do an experiment unconfunded by other factors.

After a week without gliclazide I am convinced that that medication had had a significant effect.

It does seem to vary a lot between individuals (doesn't everything?). Interesting re gliclazide. Would it be fair to say that the main mode of action is to increase insulin production? While I guess for metformin it's increasing peripheral insulin sensitivity. I would say that my T2 variety has more to do with the latter than the former, so I am expecting to see some impact from cutting metformin.

And one of the possible interactions of metformin with exercise according to the research I referred to is interference with the peripheral insulin sensitivity improvement you get from exercise - the mechanisms cancel each other out to some extent and you might not get the synergies you'd otherwise expect.

So I'm looking for impacts after eating and after exercise, not so much on waking. From my one day's worth of experiment:

- After morning soy coffee+chia seeds, a handful of nuts and a 40 min walk, I scored a 5.3. That's a smidgen less than I would have expected. So weak support for the interference scenario.

- After berries+pumpernickel+avocado+soy coffee+chia seeds for lunch/brunch, with no exercise, it was 7.0. Maybe half a point higher than I would have expected before cutting the met. So weak support for Met still being useful for post-prandial IS for me, at least in the absence of exercise.

- In the evening after a further ~90 min walking over a couple of sessions and dinner of salmon+beans+a potato+yet more soy coffee+chia it was 6.3. Ummm ... maybe a bit high? Dunno. But a confounding factor was being obliged to sit through The Lion King 🙂 (Great CGI work; wish the software used to generate story, script and dialog was as good.)

In terms of non-BG-related exercise impact: at the end of the day, not as zapped from 2 hours walking as I might have been before. But then again it was an extraordinary day for mid-Winter here: 20 degrees, cloudless, gentle breezes; and that probably helped.

The experiment proceeds. A luta continua!

PS: And then because I was feeling energetic and the evening is mild, another 30 min walk, followed by a 4.9 reading. Kind of in line with what I would have expected before. But my first 13km+ walking day since around Easter, and not zapped at all & no leg pain to speak of.

 

[Docb]

As far as I am aware gliclazide works by increasing insulin production so this would suggest my problem is related to that rather than insulin resistance. Give me a couple of weeks to collect some data and I will put the numbers up for criticism.

One of the things that gets me about diabetes is that the only thing that is measured is the outcome - that is bg level - and that treatment is dependent on either the prejudices of, or guesswork by, the HCP. Not much better than random chance, depending on how good a guesser your HCP is. In the days when I did production support I would have been fired for taking that approach to a machine that was malfunctioning because it is the most inefficient and expensive way of getting things working better. If there were two possible broad reasons for a malfunction ( like insulin deficiency or insulin resistance) the first thing I would do is to figure out which it was so that I could get appropriate remedial action in place as quickly as possible. The last thing you do is to start pulling levers at random hoping that you would hit on the best way forward.

 

[DaveB]

As far as I am aware gliclazide works by increasing insulin production so this would suggest my problem is related to that rather than insulin resistance. Give me a couple of weeks to collect some data and I will put the numbers up for criticism.

One of the things that gets me about diabetes is that the only thing that is measured is the outcome - that is bg level - and that treatment is dependent on either the prejudices of, or guesswork by, the HCP. Not much better than random chance, depending on how good a guesser your HCP is. In the days when I did production support I would have been fired for taking that approach to a machine that was malfunctioning because it is the most inefficient and expensive way of getting things working better. If there were two possible broad reasons for a malfunction ( like insulin deficiency or insulin resistance) the first thing I would do is to figure out which it was so that I could get appropriate remedial action in place as quickly as possible. The last thing you do is to start pulling levers at random hoping that you would hit on the best way forward.

Yes, there is a lot of guesswork by HCPs with regard to diabetes medication e.g. start with Metformin. If that doesn't work add Gliclazide and so on. These two meds have completely different functions with the former more relevant to T2 and the latter really only T1 (or those T1s mis-diagnosed as T2). HCPs hold Metformin in high regard but it only ever has a small effect on BS. Gliclazide can have a big effect unless all your beta cells are dead. The eternal problem is that there are no good tests to distinguish between the different diabetes types hence the guesswork. GAD 'proves' T1 (but not exclusively) and that's about it. Body weight can be a good differentiator.

 

[leonS]

Doctors are not Engineers!

 

[Docb]

Doctors are not Engineers!

🙂

I rather like DaveB's idea about the difference between T1 and T2; T2 being insulin resistance and treatable by drugs that reduce resistance and T1 being impaired production of insulin. Sort of makes sense in that those we currently call T1 are those with near 100% impairment and blatantly need insulin treatment whereas those who still have a degree of home made insulin production can get things under control by a combination of drugs to boost insulin production and reducing carb intake so as not to overload whatever insulin production they have. Must be some complications with that simple idea which I am ignorant about but wouldn't it be good if there was some way of getting a handle on pancreatic function. I'm sure those well versed in the medical black arts will tell me it can't be done, but that's what my technical brain is telling me anyway.

 

[Eddy Edson]

The more I dig into T2, the less I seem to understand it. And what the mechanism of Metformin is supposed to be. Maybe more on that later ...

Day 2 of experiment:

I had a rough night, waking at 3am with a stomach upset (I blame the salmon), dozing, waking again at 5am ravenous. Scoffed nuts and snowpeas, slept again at about 6am, woke at 10am. So not a good day for data but FWIW consistent with yesterday. So far since waking thru mid-afternoon I've tried to do the same as yesterday, food and exercise wise - boring but anything for science.

Final waking BG: 5.0.
After coffee+chia+40 min walk: 5.6 (noting that the extra-large flat white has Bonsoy milk, which is sweetened with tapioca syrup for 20+ grams of carbs, so I'd expect it to have an impact, plus I'm generally in a bit of pain from the PAD after a 40 min walk early on, which maybe does also. But the last couple of days it seems to be a bit better).
After berries etc lunch/brunch, no exercise: 7.3

Each of those data points is 2-3 ticks higher than yesterday's equivalents, maybe down to disturbed sleep, stomach probs, nuts/snowpeas scoffing early on. Or measurement error, or squidginess, or whatever.

So far the data seems to support:

- Maybe minor short-term improvement in insulin sensitivity from exercise in the absence of much food.
- But maybe minor reduction in post-prandial IS.
- Maybe slightly better walking, maybe from reduction in lactic acid build up.
- No change to waking BG.

But obviously way too early to really say anything.

Update pre-bed: Blah - my stomach probs came back - almost like Metformin side effects! I guess I've got a bug or I've eaten something.

Anyway, I'm provisionally blaming that for a high pre-bed reading, 5.7. Haven't eaten anything for hours apart from a few nuts & would have been in the 4's previously. If I'm wrong about the bug etc then it looks like cutting the Met may indeed have a significant impact on my underlying BG, which I'd expect to see reflected in tomorrow's waking read, I suppose.

More importantly, really, walking does seem to be better at least for now, despite the gammy tummy.

 

[Docb]

The more I dig into T2, the less I seem to understand it. And what the mechanism of Metformin is supposed to be. Maybe more on that later ...

Amen to that Eddy.

 

[Northerner]

The last thing you do is to start pulling levers at random hoping that you would hit on the best way forward.

Surely you'd only start doing that after you'd switched it off, then switched it back on again? 😱 😉

I think there are a lot more variables at play with human beings, as diabetes has so many factors and influences it can't be boiled down to a 'this' or 'that', so HCPs follow treatment paths that suggest courses of action and medication. You can have insulin insufficiency or impairment, or both, plus there may be many more individual influencing factors. I don't think we're quite at the stage yet where Bones McCoy can just sweep a red light over us and pinpoint a solution, although things are progressing 🙂 Probably just too expensive at the moment to perform the suite of tests that could narrow things down, far cheaper to just use relatively informed guesswork.

 

[Sharron1]

I'm experimenting with dropping Metformin. I wanted to keep taking it, for all the possible anti-inflammatory etc etc etc benefits, even though my BG has been "normal" for many months. But then I came across research indicating that it can sometimes interfere with exercise benefits, which led me to also think about whether it might be increasing lactic acid build up when I walk a lot, an issue with my peripheral artery disease. So let's see if I notice anything different when I drop it; I can always go back if my BG spirals out of orbit.

I've really been tracking only my waking BG for the last few months. It's always in the range 4.6 - 5.2. What to expect without the Met?

Met on average seems to deliver an HbA1c reduction of ~10 mmol/mol, which should correspond to an avg BG reduction of ~1.5 mmol/L. But there's a lot of variability and it becomes less effective as baseline BG reduces, though I don't find much research on that. So .... no idea!

I dropped the Met a couple of days ago and it should have mostly cleared out of my system by now. So today is Day 1 of the experiment, and waking BG = 4.8. No difference there. Let's see how that develops & whether I notice any exercise differences.

Really interesting. It is an experiment that I have often thought about. I have also wondered about the effectiveness of metformin. From my understanding I thought greater benefits are diet and exerciseAs I do loads of exercise, wat low carb and take Metformin, my numbers are consistently in non diabetic levels, I am not sure which of the above are most beneficial or if they all work together. Good luck.

 

[Eddy Edson]

Amen to that Eddy.

So as far as I can make out from clueless biochemistry-moron-level digging in the research, T2D is generally thought to be associated with insulin resistance in a variety of different ways, with different BG effects.

The most important IR effect is often said to be interference with the regulation of gluconeogenesis in the liver. That's not glucose production from carbs, as I understand it, but from other nutrients via a number of different mechanisms. Normally that gets suppressed when there's plenty of BG, but with IR, the suppression doesn't happen and the liver keeps gluconeogenesis-ing away, leading to elevated BG. T2D's typically have much higher rates of gluconeogenesis.

From the patient's point of view, the impact is supposed to be on fasting levels, mainly. And the major mechanism of Metformin is supposed to be reduction in gluconeogenesis. From which it is inferred that Met's main benefit is to bring fasting levels down. It's not a mechanism you would expect to have much impact on post-prandial levels.

On the other hand, peripheral (ie skeletal muscle) IR is also a factor for T2D's, and Met is also supposed to help with that. This is a post-prandial kind of thing, but it's supposed to be a secondary factor.

On the other hand, some well-credentialed researchers (ie not Yahoo gurus etc) seem to take issue with this view. They say that gluconeogenesis problems are only really a big thing when BG is chronically quite highly elevated, say > 10. In better controlled subjects, peripheral IR is the main factor, and the benefits of taking Metformin are seen mainly post-prandially.

I've probably screwed that little summary up one way or another. But it does illustrate a surprising lack of absolute expert consensus about fundamental things to do with T2D and Metformin ...

 

[Eddy Edson]

Really interesting. It is an experiment that I have often thought about. I have also wondered about the effectiveness of metformin. From my understanding I thought greater benefits are diet and exerciseAs I do loads of exercise, wat low carb and take Metformin, my numbers are consistently in non diabetic levels, I am not sure which of the above are most beneficial or if they all work together. Good luck.

Hey, happy to be the guinea pig!

Seriously, I'm really only doing it because maybe Met interacts with things arising from my peripheral artery disease when I walk a lot. Kind of like symptoms of over-exercise, when really you haven't done that much. If you don't find anything like that then it's probably not worth fiddling around with the Met/exercise/diet formula, I'd say.

 

[Docb]

Surely you'd only start doing that after you'd switched it off, then switched it back on again? 😱 😉

I think there are a lot more variables at play with human beings, as diabetes has so many factors and influences it can't be boiled down to a 'this' or 'that', so HCPs follow treatment paths that suggest courses of action and medication. You can have insulin insufficiency or impairment, or both, plus there may be many more individual influencing factors. I don't think we're quite at the stage yet where Bones McCoy can just sweep a red light over us and pinpoint a solution, although things are progressing 🙂 Probably just too expensive at the moment to perform the suite of tests that could narrow things down, far cheaper to just use relatively informed guesswork.

Understand that logic Northerner, but is the complete absence of attention to the issue, at least as far as I can find, that I find a bit grating. I suppose the current method keeps lots of people in a job. Its the great benefit of doing something inefficiently!

See where you have got to Eddy Edson. The simple idea that metformin makes your insulin work better, which you tend to see in basic info, is not reflected in the technical literature. It is a hell of a lot more complicated and there is no consensus about what is what. It does work, and works well for some, but I suspect there are a lot for whom it doesn't and who are putting it away because the book says so and not because it is efficacious in bringing down blood glucose.

 

[Eddy Edson]

The simple idea that metformin makes your insulin work better, which you tend to see in basic info, is not reflected in the technical literature

Well, I'm not sure that's true. Everything I've read points to Met reducing IR, it's just that the question of what that actually means is complicated, so that it'll be different strokes for different folks. Which isn't surprising and I would expect general advice to present a simplified picture - what else could it do? Particularly since it's not completely understood by experts.

I want the Dr McCoy scanner! Medical researchers should get off their butts and invent it already 🙂

 

[Sharron1]

Hey, happy to be the guinea pig!

Seriously, I'm really only doing it because maybe Met interacts with things arising from my peripheral artery disease when I walk a lot. Kind of like symptoms of over-exercise, when really you haven't done that much. If you don't find anything like that then it's probably not worth fiddling around with the Met/exercise/diet formula, I'd say.

Believe me I am the seriously over cautious type no fiddling for me. Loads of thinking and even over thinking Good luck.

 

[Eddy Edson]

Day 3 ....

Waking BG: 5.0. Stomach still a bit iffy. Metformin withdrawal symptoms, no doubt 🙂

After carby b'fast (fruit+melon+big soy flat white+chia) and 30 min walking plus bout of brain-work & a meeting: 6.5.

So far today, all consistent with pre-experiment levels. Walking continues to be easier.

Update: After late lunch (pumpernickel, avocado, raspberries, strawberries, lemon juice, soy flat white, chia) and no exercise: 7.4

A bit high, but maybe to the point, stomach went squiffy again so ...? I hope I haven't developed a food allergy? If I had to give up almonds, walnuts, soy or seeds I'd be at a loss - have to start eating beef or dairy or something 🙂 Eating enough is hard enough as it is: weighed in at 62.7kg this morning, so failed to hold at 63kg & getting close to BMI = 19, which is just too low. No matter how much I enjoy being skinny (my efforts to get people to call me "Slim Eddy" haven't been at all successful but I live in hope.)

 

[Eddy Edson]

For reference, this piece from last month is worth a look: https://blogs.sciencemag.org/pipeline/archives/2019/06/24/metformin-and-exercise

Talking about recent research showing that Met and exercise can interfere with each other, for some people.

Metformin looks far more plausible as almost anything else – a fertilizer additive, a component of marine paint to keep barnacles from attaching to ship hulls, a veterinary deworming agent. Anything other than a multibillion dollar diabetes drug.
...
The authors (a multicenter team in Illinois, Colorado, and Oklahoma) studied patients in their early 60s who had no chronic disease but had at least one risk factor for Type 2 diabetes, and who had never taken metformin. For twelve weeks, they engaged in a program of aerobic exercise, and took either metformin or a placebo along with that.

Exercise comes out looking good, as it generally does in such studies. Participants in the placebo group (exercise alone, in other words) lost fat mass, improved their oxygen handling, and decreased their fasting insulin levels. The metformin-plus-exercise group was a bit different, though: some of these metabolic measures improved, but in some cases by not as much, and there was a lot more scatter in the data. Indeed, when they gave participants an oral glucose tolerance test at the end of the study, the exercise group improved across the board, whereas the metformin-plus exercise group had about half the participants improve but the others actually get worse compared to the start of the trial.

Looking down at the cellular level (muscle biopsies, ouch), the differences became even more apparent. Exercise caused an increase in mitochondrial respiration in skeletal muscle, as advertised, but the metformin treatment definitely seemed to interfere with that process (as determined by a number of measures, and especially apparent in ADP titration experiments). The effect seems to be via some intrinsic mitochondrial function(s), rather than on protein synthesis.