Sunday, March 29, 2015

HbA1c and Familial Hypercholesterolaemia

Ivor, over at thefatemperor, recently mentioned the lovely observational study, the Norfolk section of EPIC. That's where I live and it's HbA1c which correlates with CVD here. Cholesterol does not. At all. Tha' be 'ere in Norf'k, b'o'r. With apologies for the lapse in to the vernacular. My children are becoming experts. Computer is com-poo-er...

That nudged me to put this very brief observational post up on glucose dysregulation and CVD in the land where cholesterol is king, for people with heterozygous familial hypercholesterolaemia. There are many, many problems you could point out in this study, but those are intrinsic to a retrospective observational study. Take a group of hFH people who have survived a premature heart attack. Match a similar group of hFH people who haven't had a heart attack. How do you tell the difference between the groups? As a lipidologist perhaps you might suspect the LDL cholesterol level? That is exactly the problem in hFH after all... But:

"There was no difference in total and LDL cholesterol between the two groups. Patients with previous myocardial infarction had significantly higher levels of insulin, insulin resistance [and several other things I'm not interested in, which they would like to treat]..."

Insulin resistance is the problem, on a mixed diet. Do you think this might show in the HbA1c, just as it does here in Norfolk for us non hFH folks?

It probably does. In a similar observational study on hFH, using HbA1c itself rather than insulin/resistance parameters:

"Of special note is that HbA1c showed a significant correlation with average ATT [Achilles Tendon Thickness], independent of other parameters..."

Achilles Tendon Thickness is a marker of Badness in hFH. As your tendon thickens, so too does your carotid intima thicken.

In general, patients with hFH tend to have rather good glucose control compared to the general population. That might just be why they live as long as they do under the correct circumstances. But with hFH I suspect that, should you manage it, developing metabolic syndrome may be a very unforgiving problem. You have to wonder what side stepping the syndrome by low carb eating might do, giving chronic normoglycaemia without elevated post prandial insulin. Not holding my breath waiting for that one.

BTW, as statins worsen glycaemic control, could they actually make the CVD problems worse for hFH patients? Surely not. Surely we have stopped making booboos of this type. Of course we have. Of course.


Wednesday, March 25, 2015

Ketogenic vs moderate carbohydrate diets

I thought I might put up this graph:

It's from

Comparison of the Atkins, Zone, Ornish, and LEARN Diets for Change in Weight and Related Risk Factors Among Overweight Premenopausal Women: The A TO Z Weight Loss Study: A Randomized Trial

The lead author does not appear to be a LCer. In general his publications are rather pro plants and quite mainstream. To his credit he has published some negative studies amongst the pro plant stuff. Harvard, ultimately, is no hotbed of pro Atkins zealotry.

If we want to look at the macros we can check here:

A little arithmetic allows us to look at the carbohydrate intake on the first graph at differing time points for Atkins and Zone diets:

Of course the Atkins diet was an unrestricted calories diet, Zone has a caloric restriction applied.

I have the impression from the data of the A-Z study that low carb is good, slightly higher carb is acceptable, adding more carbs back in is a booboo and that the Zone is crap. Just an impression. From the graph.

As an aside, of course the unanswered question is what, exactly, would a sustained 54g carb intake have produced in terms of weight loss over 12 months? Or 20g/d over 12m?

It is very clear that carbohydrate restriction only works WHEN YOU RESTRICT CARBOHYDRATE. A low carb diet does not appear to be as effective as a low carb diet when it has morphed in to an ex low carb diet through added carbs. A similar pattern might apply to ultra low fat diets if anyone wants to go down that dark alley. They don't work when you add fat. Assuming you don't mind the biochemistry while you eliminate fat.

It is also very clear that when comparing an almost-ketogenic diet to a modestly restricted carbohydrate diet of around 133g/d carbs, something like the Zone diet, the modest carbohydrate diet is just as good, if not a little better, than a ketogenic diet. You know the graph:

My problem is trying to square the circle between these two studies. Obviously, no study is free of bias. I struggle somewhat with Dr Sears, of the Zone diet, being the group leader of the study which shows a diet with 133g/d of carbohydrate out performs a ketogenic diet. That is very strange and doesn't happen in Stanford.

People must make up their own minds.


Sunday, March 15, 2015

Insulin detemir (3)

I think it is quite clear how I view insulin detemir. Kindke was unable to resist finding the link to the abstract with the diametrically opposing view. I'll just stick both links in here to keep them together, so people can look at both research findings and draw their own conclusions.

Insulin detemir is not transported across the blood-brain barrier.


Insulin detemir is transported from blood to cerebrospinal fluid and has prolonged central anorectic action relative to NPH insulin.

I think it is reasonable to assume that at least one of these two papers is factually incorrect.

If you search on Begg and Woods as co-authors you will find papers redolent with words like "reward", "hedonic" and "dopamine".  That's Begg and Woods, if anyone can stomach it.

I was, in my normal confirmation biased way, much more interested in the sort of work produced by Banks, Morley and/or Mooradian. These folks appear to be scientists rather than psychiatrists and they have some great publications. They include major work on the blood brain barrier, leptin transport, insulin transport, leptin resistance, gerontology, diabetes, antioxidants, the list goes on and on.

Here are a few little gems I particularly enjoyed in abstract form which might be worth a mention.

I dislike antioxidants. This is quite interesting from Banks and Morley:

Effect of alpha-lipoic acid on memory, oxidation, and lifespan in SAMP8 mice.

Alpha lipoic acid is a mitochondrial component present in normal cells and is available in mega doses as a supplement. It's a serious and deeply mitochondrial penetrative antioxidant. It helps a lot with diabetic neuropathic pain. SAMP8 mice are oddities which have been bred for early onset senility and memory loss. They are used (probably totally inappropriately) for Alzheimers Disease research. Treating them with antioxidants improves their memory performance. You might think this is a good idea. The cost is measured by a shortening of their life as elderly SAMP8 mice from 34 weeks to 20 weeks after start of treatment (started at 11 months of age). This may or may not be a good thing if you are an SAMP8 mouse (death might be a release). How it applies to a person managing their diabetic neuropathy or trying to delay the progression of their Alzheimers Disease is fascinating and slightly worrisome. I'll stick to a life based around beta oxidation, normglycaemia and a little superoxide signalling, stuff the antioxidants. Last sentence of the abstract "These findings are similar to studies using other types of antioxidants". Sweet, provided you avoid sugar. And antioxidants.

The next snippet includes Morley and Mooradian as authors and relates to making your blood sweet,  literally this time, using intravenous glucose:

Mechanism of pain in diabetic peripheral neuropathy. Effect of glucose on pain perception in humans.

Simple hyperglycaemia in a normal person reduces the threshold for feeling pain. It reduces the severity of pain you can tolerate. This applies to a normal human being on a glucose infusion or a diabetic person on a diet designed by a diabetologist, no glucose infusion needed. If hyperglycaemia makes a tolerable stimulus in to a painful experience and makes just bearable pain become unbearable, how many chronic pain syndromes would go in to remission with sustained normoglycaemia? Fat phobia makes this question currently un-answerable. The paper was published in 1984. Does anyone fancy having a gangrenous foot amputated for diabetic complications and waking up on a dextrose saline infusion in recovery? And then being offered the "diabetes diet" on the post op ward?

Banks and Morley were also instrumental in the generation of data for the concept that trigycerides in plasma induce leptin resistance at the blood brain barrier, a few years old now but still quite a useful concept:

Triglycerides induce leptin resistance at the blood-brain barrier.

I find the cream bashing in this last paper a little distasteful and I have to admit that Banks appears to be unaware that high saturated fat low carbohydrate diets are THE way to reduce fasting trigycerides in real people. Can't have everything I suppose. But even if the cream effect applies to people, who cares if I am leptin resistant with a full stomach provided leptin will work perfectly well in the post absorptive (low triglyceride) period? I am a human, not a mouse. I ate a high fat meal without sugar last night, ergo I'm not hungry today. Low trigs equal leptin sensitivity...

I'll call a halt there. Life is full of interesting snippets which make sense. They usually come from the sort of people who say insulin detemir does not cross the blood brain barrier.


Wednesday, March 04, 2015

Insulin detemir (2)

Morphine is a rather odd opioid analgesic. It has a complex multi-ring structure with two rather prominent hydroxyl groups which render it rather more hydrophilic and significantly less lipid soluble than many of its relatives. If you bolus a patient with IV morphine there is a delay in its passage across the blood-brain barrier due to this relatively poor lipid solubility. Time to peak effect is significantly delayed to somewhere around 15 minutes because the brain concentration lags way behind the rapidly changing plasma concentration. The brain never "sees" the peak plasma concentration due to this delay.

Now, if you boil some morphine up with acetic acid you can form ester linkages joining acetate on to those two hydroxyl radicals to give you di-acetyl morphine, better known as diamorphine or heroin. Masking the hydroxyl radicals markedly increases the lipid solubility of the drug and so the brain concentration rapidly follows the plasma concentration. In general lipid soluble agents cross the blood brain barrier rather faster than more water soluble agents. Peak plasma concentration will give a rapid onset peak brain concentration, which appears to be associated with effects rarely seen with morphine itself. Giving the enhanced recreational potential. This is all basic anaesthesia pharmacology with excerpts from Trainspotting thrown in.

Insulin detemir was developed to give an insulin with a very flat glycaemia controlling effect for use as a basal or background insulin. The clever people at Novo Nordisk deleted the terminal threonine from the B chain and attached a medium chain fatty acid to the now terminal lysine at position B29. The rather nice 14 carbon saturated fat, myristic acid, sticks out from the insulin molecule and neatly binds to the fatty acid binding site of albumin. It does this very rapidly and keeps the insulin bound and ineffective. Over the hours which follow there is a slow dissociation of the insulin from albumin which allows a very shallow dose response rate for glucose control. Ideal for a basal insulin.

There is a suggestion that this tagging of insulin might facilitate its transport in to the brain, a sort of heroin-insulin tweak. The idea is that myristic acid might facilitate the transport of insulin in to the brain and lead to a massive suppression of eating and subsequent weight loss. Assuming you are a true believer in the central anorectic effect of insulin. Which, sadly, I'm not.

Years ago, when insulin determir was first paraded as the living proof of the central anorectic effect of insulin, I looked up its structure and thought, as you do, that FFAs in general have very limited access to the brain. Insulin is not morphine and the myristic acid is not acetic acid. That big, long side chain of detemir is directly related to the sorts of free fatty acids which are specifically excluded from the brain. My own prediction would be that insulin detemir would have a significantly REDUCED effect within the brain.

It turns out that, at least in some labs, that my idea was slightly correct. But my idea was limited compared to the actual effect. Insulin detemir not only fails to cross the blood brain barrier itself but it also blocks the ability of ordinary human insulin to pass from plasma in to the brain. There is probably a specific insulin transporter which is nicely blockaded by an insulin molecule with the fatty acid tail of detemir sticking out. This paper says it all:

Insulin Detemir is Not Transported Across the Blood-Brain Barrier

Not a lot of mincing of words there.

If we go to labs with an outlook on life which I find comprehensible we can clearly see that physiological doses of insulin, within the brain, augment lipid uptake in to adipocytes, enhance adipocyte sensitivity to insulin, increase lipogenesis and augment fat gain. Largely through the sympathetic nervous system. I can't see how anyone would be surprised by this. Quite why anyone would expect central insulin to do the opposite of what peripheral insulin does at a comparable concentration is beyond me. I enjoyed this paper:

Central insulin action regulates peripheral glucose and fat metabolism in mice

"Moreover, chronic intracerebroventricular insulin treatment of control mice increased fat mass, fat cell size, and adipose tissue lipoprotein lipase expression, indicating that CNS insulin action promotes lipogenesis. These studies demonstrate that central insulin action plays an important role in regulating WAT mass and glucose metabolism via hepatic Stat3 activation".

How clearly does it need to be spelled out? This one is fun too:

Brain insulin controls adipose tissue lipolysis and lipogenesis.

"Here, we show that insulin infused into the mediobasal hypothalamus (MBH) of Sprague-Dawley rats increases WAT lipogenic protein expression, inactivates hormone-sensitive lipase (Hsl), and suppresses lipolysis. Conversely, mice that lack the neuronal insulin receptor exhibit unrestrained lipolysis and decreased de novo lipogenesis in WAT".

If you go looking you can find papers from Oz and Cincinatti which show that insulin detemir DOES cross the blood brain barrier and DOES suppress food intake, far better than neutral insulin does. In their own labs of course.

But I cannot forget that if you transport a researcher out of a Cincinatti psychiatry department and put her in to an industrial insulin lab she cannot get any effect of centrally infused insulin detemir or neutral insulin for that matter. Novo Nordisk cannot demonstrate this marvellous effect of insulin, even their own special insulin, in their own lab. We all know that much of the mindset of obesity research is not particularly effective at producing results which work. How they get the results derived from their ideas in their labs is what fascinates me! You couldn't make stuff up this counter intuitive. Maybe in another post.

Back in the real world we have this:

Insulin detemir results in less weight gain than NPH insulin when used in basal-bolus therapy for type 2 diabetes mellitus, and this advantage increases with baseline body mass index

Insulin detemir causes a small weight loss in morbidly obese patients, those with BMI >35kg/m2. Why? Because it blocks the brain entry of the chronically (and markedly) elevated levels of insulin so common in the morbidly obese. It has limited or zero effect within the brain in its own right. The brain simply loses awareness of the systemic pathologically elevated insulin. If plasma insulin is high enough this sudden loss of insulin's access to the brain can result in a decrease in brain driven, neurologically mediated, forced lipid storage in adipocytes, i.e. a little weight loss.

In the absence of marked hyperinsulinamia, i.e. in less obese type 2 diabetics, insulin detemir causes weight gain because there is less tonically elevated plasma insulin for the central uptake blockade to neutralise. There is no weight loss effect, although gain is undoubtedly blunted.

Insulin detemir is the best indicator I have seen that the central role of physiological concentrations of insulin within the brain is to augment fat storage. This makes sense to me.

I wouldn't ask a psychiatrist to develop an anaesthetic protocol. Or a weight loss protocol!


Monday, March 02, 2015

Insulin detemir (1)

There is a certain belief structure within obesity research which maintains that the central action of insulin is to limit appetite. Obviously, not everyone agrees with this. I would like to do some wild speculating (any resemblance to real life events is purely accidental) about this paper:

Evaluation of the lack of anorectic effect of intracerebroventricular insulin in rats

The paper is very interesting, partly for what they failed to reproduce but mostly for the affiliations of the authors.

The first thing to say is that, if you work in the pharmaceutical industry, you want drugs which work. Hardcore. It’s no good fudging your results when working in pharmaceutical R&D because you’re going to get caught out as soon as anyone tries to actually use your drug. Which is guaranteed to happen. The drug has GOT to work. Industry has no fudge factor. You might have to lie, evade, obfuscate, misplace computer files and massage data to hide the serious adverse effects of your functional patented drug, but you wouldn’t want to have to do this for a molecule which is ineffective in the first place. Statins are very, very effective. At lowering cholesterol. The fudge factor comes from whether this does any good for any person and what multiple adverse effects the drugs might generate.

So I have respect for the integrity, within certain defined limits, of a drug company R&D team. The managers and PR crowd are another matter altogether. Think Dilbert.

Let’s look at the authors of this paper:

There are three. Jessen is the first author, so probably did the bulk of the work and wrote much of the paper. She works in the department of insulin pharmacology at Novo Nordisk, the company which makes insulin detemir. Bouman is last author so is possibly Jessen's line manager and also works for Novo Nordisk. Insulin detemir is interesting because it is the only insulin ever to have been shown to cause weight loss in any patient group. OK, this is limited to morbidly obese (BMI>35) type two diabetics and the weight loss is very small. But it does happen. Quite amazing really and quite different to any other insulin formulation on the market, all of which reliably cause weight gain. Hence I suspect the project at Novo Nordisk was to find out the hows and whys of this strange effect.

Jessen and Bouman will have started with generous (by academic standards) funding, because the drug industry will work at a potentially rewarding idea in a rather more motivated manner than an academic department. Neither author has any track record of publishing on the central anoretic effect of exogenous insulin. Their job is to get reliable and repeatable results about how insulin detemir is special. In this project they failed to achieve any sort of anorectic effect of insulin detemir, or of any other sort of insulin, within the brain. Mucho problemo.

Clegg is middle author and works in the Department of Psychiatry, University of Cincinnati. She has a vast number of publications, several of which feature the successful anorectic effect of insulin when administered directly in to the brain. In at least one such study she is the lead author.

Jessen has co published with Clegg back in 2001 on a non insulin related subject, presumably before Jessen moved to work for Novo Nordisk. They know each other and have worked together before.

I have this image of two industrial pharmacologists setting out to investigate the CNS effects of their rather promising systemic drug, insulin detemir, comparing it to routine and more obesogenic neutral insulin. They fully expect central insulin to be anorectic because they've read all of the papers. That's their job. They expect insulin detemir to be extra effective. In the first run of experiments using intra cerebral administration they failed to get any effect, of any type of insulin, on food intake. None.

This is big. And bad. EVERYONE in obesity research KNOWS that insulin, within the brain, suppresses appetite (excepting the few people who think this idea is bollocks of course, there are always a few people who think logically).

Jessen and Bouman probably think they have made a mistake somewhere along the line. They know that Clegg can, in academia at least, deliver results that show a suppression of appetite in rats following centrally administered insulin. They call her over from of Cincinatti to trouble shoot their problems.

In a hard nosed, financially driven situation, she can't do it. From the abstract of the study:

“Although we varied rat strain, stereotactic coordinates, formulations of insulin and vehicle, dose, volume, and time of injection, the anorectic effect of intracerebroventricular insulin could not be replicated”.

It seems to me that there are differences between academia and industry. It’s the difference between holding a religious belief in the central anorectic effect of insulin and looking for an effect which might suggest a marketable drug which will actually work to assist weight loss. I would call the latter "The Real World".

Time to discard the idea that centrally acting insulin is an anorectic agent? Kudos to the researchers for publishing.