Saturday, May 31, 2014

HbA1c: Low glucose and acid (palmitic)

If anyone would like to combine a look at the pathology of low blood glucose levels with the psychology of diabetes research they might do worse than to look at this paper.

Hypoglycaemia is bad, very bad. At the vascular endothelium level and at the mitochondrial delta psi level. The sort of low levels of blood glucose frequently visited by people like Steve Cooksey and myself are a complete disaster. The group is looking at levels of blood glucose that I can easily achieve by skipping a couple of meals combined with a little mild aerobic exercise, walking or cutting the lawn. I’ve seen 2.7mmol/l on my (admittedly relatively inaccurate) Freestyle Lite. I felt fine at the time and I did nothing to adjust my blood glucose level before having supper a few hours later...

Perhaps I should be carb loading, just a little, to avoid hypoglycaemia induced damage? However this might drop my free fatty acid levels, which I would prefer to avoid.

The abstract of any paper is word limited and the role of FFAs in the findings don't get a mention, so you might assume FFAs have no significant bearing on hypoglycaemic injury.

The discussion section of the paper tends to confirm this:

“Our findings with respect to fatty acid and L-carnitine suggest mechanistically that free fatty acid utilization in vivo would not be sufficient to suppress LG induced NO suppression and excessive mitochondrial superoxide production…”

They seem pretty certain, but do they have data to support this? Flicking to the results section we see that the group looked at a palmitate with L-carnitine combination and found that it worked beautifully:

“We found that combined treatment of endothelial cells with 1% palmitate and L-carnitine reduced LG-induced mitochondrial superoxide production to normal levels (Supplemental Figure IIIa)”

Want the picture?

I think we will all agree that the combination of palmitate with L-carnitine completely, totally normalises superoxide production in the low glucose cells and was utterly harmless in the normal glucose cells. The graph speaks against the discussion.

This normalisation of superoxide generation appears to have been too much to bear, the concept that palmitic acid might be beneficial, so they separated the palmitate from the L-carnitine and tried them separately.

“mitochondrial superoxide production increased under LG conditions with the addition of 1% palmitate while L-carnitine alone returned mitochondrial superoxide levels to those similar to the NG condition (Supplemental Figure IIIb)”

Now, this sticks in my craw. This is wrong. Completely.

Here is Supplemental figure IIIb.

Any reading with an * above it is higher than the normal glucose (NG) reading. This includes the low glucose plus L-carnitine. The low glucose with carnitine value is greater than the normal glucose value, p < 0.05.

There is, however, a crucifix above the LG+LC. The superoxide production here is significantly less than that in the LG group with added neat palmitate, that’s the meaning of the crucifix. But because of the asterix we KNOW that the superoxide here is still higher than the normal glucose value, p < 0.05.

Stating that

“L-carnitine alone returned mitochondrial superoxide levels to those similar to the NG condition”

would require that you (could) remove the asterisk from the low glucose plus carnitine column.

You can't.

You might just as well say the normal glucose plus palmitate (NG+palm) is "similar to the NG condition". Which they don't. Asterisk.

Sigh. End rant.

There are many other things which are both good and bad about this paper. But, as a lipophile living in what sometimes feels like a sea of lipophobia, it really ticks me off to see frankly incorrect and unsupported statements like this. I still scratch my head about motive.

It’s quite clear that palmitate/L-carnitine normalised the excess superoxide production of low glucose completely. I have lots of palmitate in my blood stream, though perhaps not the 1% chemical grade palmitic acid used here! I've probably got reasonable amounts of L-carnitine in my cells too, being a moderate meat eater and highly fat adapted. But the big giggle would really have been to use ketones in the study. I have ketones. As Veech pointed out, many years ago now, ketones can completely replace the whole glucose/insulin metabolic pathway in an isolated rat myocardium. Very effectively.

The flip side is that the paper is looking for information about insulin overdose rather than simple low-normal glycaemia. My BG of 2.7mmol/l was in the presence of utterly basal insulin, markedly elevated free fatty acids and modestly elevated ketones. A diabetic on a sugar based diet having a little (or even quite big) accident with their insulin to starch balance will undoubtedly have suppressed FFAs waaaaay before they hypoed and will probably have a metabolism that can't remember what a ketone body is. The metabolic milieu of parenteral insulin overdose is radically different to that of ketogenic eating. You have to engage thought processes before deciding that any blood glucose below 5mmol/l produces graded superoxide overproduction and failure of nitric oxide signalling. It's understandable from a modern diabetologist point of view, just wrong to bury data which do not support your preconceptions.

If non-injecting (and non sulphonylurea popping) folks want to reduce their FFAs and eliminate ketones in order to avoid low-normal glucose levels by eating a few extra grams of carbs, that is absolutely fine. We can all ignore the biochemistry if we so wish, serious researchers do it all the time.

Personally, I still like to have a few ketones around. And rather a lot of free fatty acids.



Nigel Kinbrum said...

You wrote
"I’ve seen 2.7mmol/l on my (admittedly relatively inaccurate) Freestyle Lite. I felt fine at the time and I did nothing to adjust my blood glucose level before having supper a few hours later..."
You don't exercise at such a high intensity as Steve Cooksey. He momentarily blacked-out in Feb 2011, which resulted in me writing "Funny turns": What they aren't and what they might be after we had a "misunderstanding" (a.k.a. massive row) on Facebook because I mistakenly used the term "Somogyi reaction".

High-intensity (>85% max) exercise burns carbs at a rapid rate (>3.4g/min), no matter how fat-adapted you are. Having a significant amount of muscle glycogen has benefits, if you occasionally have to sprint or lift really heavy stuff.

Galina L. said...

I have the experience of very intense exercise in a fasted state (long 1.5 hour cardio with intervals >85max) , as I commented before on another blog post. After several months adaptation , it trained my body to produce glucose at a rather high rate of 121 mg/dl (always that particular number, much higher that upon waking up, I suspect my RelyOn is not very accurate and gives slightly higher numbers than a lab does) instead of relying on glycogen in my muscles. I doubt I have much of glycogen due to my LC diet which I follow for last 6 years. Before I got adapted, it felt pretty rough, may be led to the couple post-exercise migraines . I have no idea how healthy it is , but it is at least convenient.

Kindke said...

I do agree keto dieting and high intensity exercise dont mix well.

A few max intensity interval sprints on the elliptical machine at the gym and I feel very light-headed and have to lay down. a few times ive come close to passing out.

Same as when I go for high reps on squats.

Maybe you can argue im not "fat adapted" whatever that means, but my exercise performance is much better when im carb'ed up.

Nigel Kinbrum said...

Galina L. said...
"I doubt I have much of glycogen due to my LC diet which I follow for last 6 years."
Hi Galina. How many grams of carbs/day do you eat? I consider LC to mean ~100g/day and VLC to mean <50g/day.

Galina L. said...

Hi, Nige!
It was a while when I measured everything. When I did it for couple weeks I cycled between 30 to 65 grams of carbohydrates a day. It took me almost a year to be adapted to fasted exercises, and I had an unpleasant experience doing so.

OldTech said...

Very interesting post.

I have been wondering for some time what normal glucose would be for keto-adapted. This, of course, was not the focus of this post or the paper, but still this post and paper seems to suggest that LG may not be [as] damaging when keto-adapted despite the fact that the paper did not consider ketones.

John said...


I've been somewhat low carb for 4-5 years now and have never really noticed anything that would seem like glycogen shortage. However, only in this past year did I stop "carb refeeds," which I did just out of taste. I haven't had an appetite for sweet potatoes, etc in a long time for some reason. I probably eat <100 carbs per day and do weightlifting, climbing, and ballet. I did notice that when eating high carb my muscles got "pumped" more easily.

Maybe my protein intake helps, or maybe I get by with just enough glycogen. I do eat high kcal (4-5000).

There was a study in gymnasts, who seemed to do okay on ketogenic diets, though they were very high in protein (40-50% cal I think).

I've never measured my blood glucose personally though in fasting lab measurements it's been low 80s.

Puddleg said...

To be in the palmitate without carnitine state, you'd need to have scurvy.
While we're talking about ascorbate,

Mitochondria, Energy and Cancer: The Relationship with Ascorbic Acid

Stuart said...

Very interesting and wondering if most are not really low protein say <20% and extremely high fat?
I'm more in the <100g carbs and higher >30% protein but <40% and never seen my BG go below 3.8 and H1AC around 4.5 and never measured ketogenic but close and my gave up as no meter ketone meter worked for after using 7 of them, with support line so maybe I was just borderline and never fat adapted?

Nigel Kinbrum said...

Galina L. said...
"Hi, Nige!
It was a while when I measured everything. When I did it for couple weeks I cycled between 30 to 65 grams of carbohydrates a day. It took me almost a year to be adapted to fasted exercises, and I had an unpleasant experience doing so."
Thanks! How often do you do 1.5 hour cardio with intervals >85max?

It's interesting that your blood glucose level is on the high side on a VLC diet. This supports my idea that hepatic glucose production is not tightly-controlled.

Nigel Kinbrum said...

john said...
I've been somewhat low carb for 4-5 years now and have never really noticed anything that would seem like glycogen shortage. However, only in this past year did I stop "carb refeeds," which I did just out of taste. I haven't had an appetite for sweet potatoes, etc in a long time for some reason. I probably eat <100 carbs per day and do weightlifting, climbing, and ballet. I did notice that when eating high carb my muscles got "pumped" more easily."
Cheers, John. As I think that muscle glycogen stores are "trickle-charged" by blood glucose while relatively sedentary, whether or not someone on a VLC diet runs out of muscle glycogen depends on the frequency and duration of high-intensity exercise.

Galina L. said...

Nigel, that my blood sugar is higher on a VLC diet as a response to an intense physical activity (over-wise it is very slowly dropping during the day if I fast and perform only regular day-time activities) makes me think than it is the adaptation of my body to the low level of glycogen.

karl said...

First - 2.7mmol/l is about 46mg/dl - I don't believe your meter. The truth about those meters is they are only slightly accurate +/-15% or +/-20% over a small range. Finding the complete specifications for a particular meter is often impossible (they simply claim they are FDA approved).

Next - keto adaption takes time that seems to vary between people and simply eating excess protein will push you out of it. Testing with keto sticks or looking at post-prandial BG will let you know where you are. I wonder if we lose the adaption faster than we gain it?

I have a question - not sure if there is a clear answer. I know low-carb people that have excellent fasting BG but their HbA1c didn't look so good. This could be because of Different reasons
1 -, they are getting high post-parndial BG from protein and don't know it

2 - the HbA1c test is not as accurate for people eating low carb.

3 - The accuracy of HbA1c has been over-sold.

See -

I am quite disappointed that most MDs don't look at postprandial BG - A lot of the time fasting BG looks good until after a lot of damage is done. Early detection could greatly improve the lives of people willing to avoid carbs.

There is also the question of peak BG - would the average claimed by HbA1c show the actual damage done or just the area under the curve? I doubt that the damaging effects of BG are linear.

Peter said...

Nigel, I have an elderly Raleigh Clubman 12 speed racing bike, circa 1980. One day I pedalled it up Applepie Hill hill out of Compton. I used first gear. A few days later I did it in second gear. Next in third gear. After a couple of weeks I could get up in 12th gear. The steepest bend is close to 1 in 5. I did it one more time in 12th gear then lost interest. As you note, I’m not much in to exercise.

Streatley Hill climbs out of the Thames Valley at Goring Gap. I can only manage this hill in 1st gear but it is steep enough and sustained enough that it feels very much like I work at VO2 max throughout the climb.

None of the climbs elicited anything suggesting a hypo. What was very noticeable was that, on cresting either hill, there was no lactic acidosis, no payback time for oxygen debt, respiratory minute volume just dropped to normal cycling demands and there was no need to pause at the top of the rise. Deep ketosis, no idea what carb level I was on, but rather low.


Peter said...


Mel published this at this while she was at the AHT, we've more data since. The meters are not perfect but the Freestyle Lite was never 20% out compared to the commercial lab at the Trust or to our bench top machine in Berkshire.

No one expects 2.7mmol/l to be accurate any more than the 1.9mmol/l was accurate when Mel checked her BG on some random glucometer at the Trust at the end of a 13hour shift w/o eating in deep ketosis. The figures may not be accurate but they don't represent 4.0mmol/l either. There is a rumour the meters are designed to under read at low readings but there are zero data published data to back this up, we really looked. We also generated low glucose samples by letting blood stand before paired readings and they seemed fairly good. We we using discard feline samples, not human. Hence the study to see if they were OK.


Michael Frederik said...

I have a raleigh 20 torpedo duomatic. It goes up any hill. It runs on elevated ketones only. No probs.

Zachary said...

I know you're a busy guy but i'd be curious to know your thoughts about this study being done at my local university:

I find it outrageous they're putting average joes on a high-fat diet for 5 days and testing for endotoxin, intestinal permeability, and skeletal muscle metabolism. I sent them an email to volunteer, mostly as a joke. :)

Unknown said...

I'm not sure why everyone is debating that Peter could have dropped his BG that low and not experienced hypoglycemia.

> After fasting 2 months, administration of weight-adjusted doses of insulin produced identical maximum insulin concentrations and disappearance curves. However, no insulin reactions nor significant rises in catecholamine excretion occurred despite equal extent and rate of glucose fall. Glucose concentrations as low as 0.5 mmoles/liter (9 mg/100 ml) failed to precipitate hypoglycemic reactions.

As low as 9 mg/dl, yet no hypoglycemic reactions! In a published study.

This isn't debatable, hypoglycemia in the presence of ketones isn't an issue. If people have other studies showing a fat/ketone adapted person or group of people experiencing hypoglycemia through any mechanism, please share. Otherwise, you're making things up.

Zachary said...

In regard to the Virginia Tech high-fat study.. After emailing the team inquiring about it they sent me a response:

" The high fat diet is 55% fat and is isocaloric for five days. Isocaloric means you will not gain or lose weight, but the total study time is 29 days. This includes baseline testing, a 2 week period of diet standardization (where we tightly control your calories) and the 5 days of high fat diet a post testing. You would need to remain physically inactive for the remainder of the study if you qualify. The diet is a mix between saturated, unsaturated and polyunsaturated, but no trans fat. The diet is foods such as TV dinners, meat and cheese. We provide all food and compensation is $400 for completion of the study. Please fill out this survey to see if you meet the subject criteria."

TV Dinners?

OldTech said...

@Dustin, @Peter

This is the first time I have heard of no hypoglycemia when in keto. It makes sense to me if the brain is fat adapted. Still why is this not more widely known? I got beat up by some type I's for suggesting very low carb because of their fear of hypoglycemia. Or am I not understanding what is implying.

Peter said...


High fat and endotoxin is important, it's on the "to do" list...


Peter said...

Lovely find Dustin! Ta.


raphi said...

My 2 cents:

My fasting BG level is ~80mg/dL (~ 4.5mmol/L) or lower.

Following a 16hr fast (or >), my glycolytically demanding workout (~ 15min) elevates my BG to ~160mg/dL (~9mmol/L).

My blood glucose then drops precipitously (below 100mg/dL) and steadily reaches fasting levels thereafter. So GNG does...what it's supposed to do without ‘overstepping’. This walks & talks like metabolic flexibility.
In fact, my breath ketone analyzer, Precisions Xtra ketone meter (& the Bayer KetoStix) all tell the same story: my body is still providing detectable & likely impactful levels of beta-oxidation, despite everything else.

Whether I then go on to eat a keto or mixed meal post-workout, my post-prandial BG levels tend to drop to to ~60mg/dL (~3.3mmol/L). This is likely due to increased post-exercise musculoskeletal glucose uptake & a capacity to rely on FA substrates. I’m not light-headed, moody or any of that. Actually, I tend to feel particularly energised (whodda thunk it!?).

- "Effects of a short-term carbohydrate-restricted diet on strength and power performance"
- "Ketogenic diet does not affect strength performance in elite artistic gymnasts"

“keto is incompatible with glycolytic activity” is a straw-man argument at this point. It’s obviously compatible (to say the least). The question is how does it fare compared to non-keto diets in terms of specific performance, health & longevity measures. What is the interindivdual variability? How can it be tweaked?

For heavens sake, “It doesn't matter how beautiful your theory is, it doesn't matter how smart you are. If it doesn't agree with experiment, it's wrong!” - Richard P. Feynman

raphi said...

I'm pretty sure Jesus & his Dad were high-fiving each other when they figured out GNG & ketones were advantageous systems to bring together for us meat-machines...especially considering how they loved to see us wonder hungry for days while fighting off other animals!

P1 said...

I was looking at the NHANES III data, which showed that A1C < 5 increased mortality. BUT, what was really interesting was that in the groups with AIC in the 4s the standard deviation of the population exploded. In other words, some people who had A1C of 4.5 lived much longer and some people with A1C had much higher mortality.

That made me think these were actually different populations entirely, and they were improperly grouped together. Maybe the higher mortality groups were injecting insulin, or had other disease conditions. I've always wanted to understand the reasons for that data.

Unknown said...

P1 excellent point. Glucose changes the optics in the brain. You lose the ability of harness polarized light carried on your electrons that get released to your now glycated brain tissue. Non native EMF exposure raises your perceived need for glucose as laid out in my EMF 4 blog post. There grouping in these charts underscores that these people actually represent different populations entirely. One has optimal optic transmission in their brain and the other does not. Here you see the effect of chromophores in the brain. No one seems to see what I see in these charts. This explains how HbA1c ties to Vitamin A and D cycles in the brain. Vitamin A and D cycle link to RXR receptors which directly tie to DHA. It is clear with my view point that they were improperly grouped together in the data and the most important link to longevity was buried.

js290 said...

Liphophobia is a faith based proposition. With fatty acids being a better source of ATP than glucose, Nature as a liphophiliac should at least be the null hypothesis. This is what natural selection tells us.

Not sure how they can decouple the actions of L-carnitine from palmitate. Be wary of any "scientist" trying to decouple coupled systems.