Monday, April 20, 2020

Double Bond Index and longevity in humans

Preamble: I've had this post written for some time (there are a fair few in this category) but this tweet from the World Health Organisation has prompted me to hit the publish button. In particular this piece of advice begs the question of incompetence vs malicious intent (I doubt the latter):

Here's the post:

I thought I would revisit the idea of trans fatty acids because the late Fred Kummerow got an honourable mention on twitter recently. He is largely responsible for the removal of industrial trans fatty acids from the food chain. No one would argue that that was not a Good Thing.

Back in the 1970s a study was completed which applied a diet from which saturated fats were largely removed and linoleic acid, mostly from corn oil, was increased to about 13% of calories. In a "control" diet saturated fats were as unchanged as practical and linoleic acid limited to just under 5% of calories. That should be a pretty good test of the miraculous benefits of dietary PUFA for blood cholesterol lowering.

However the "control" diet just happened to be specifically increased in industrial trans fatty acids from commercial margarine (1960s style USA margarine), though no one knows by how much, ie there was no genuine "control" diet.

This is what the diets looked like

Cholesterol lowering diet:

"Liquid corn oil was used in place of the usual hospital cooking fats (including hydrogenated oils) and was also added to numerous food items (for example, salad dressings, filled beef (lean ground beef with added oil), filled milk, and filled cheeses). Soft corn oil polyunsaturated margarine was used in place of butter. This intervention produced a mean reduction in dietary saturated fat by about 50% (from 18.5% to 9.2% of calories) and increased linoleic acid intake by more than 280% (from about 3.4% to 13.2% of calories)".

"Control" diet:

"It was designed to appear similar to the experimental diet. Notably, free surplus USDA food commodities including common margarines and shortenings were key components of the control diet, making the daily per participant allocation from the state of Minnesota adequate to cover the full costs. As common margarines and shortenings of this period were rich sources of industrially produced trans fatty acids, the control diet contained substantial quantities of trans fat. Compared with the pre-randomization hospital diet, the control diet did not change saturated fat intake but did substantially increase linoleic acid intake (by about 38%, from 3.4% to 4.7% of calories)".

You have to wonder about the inclusion of trans fats in the control diet. Did Ancel Keys (co-principal non-author) realise, even as long ago as the mid 1960s, that trans fats were bad? A little stacking of the deck has never been been considered an issue when it might help support the lipid hypothesis.

The experiment, designed to confirm the benefits of PUFA, failed completely. There was zero benefit from cholesterol lowering using dietary linoleic acid. Keys never published the results, hence my use of the term "co-principal non-author" because non published study results cannot have any authors. Happily enough data were excavated by Ramsden et al 40 years later to be published in 2016 as

Re-evaluation of the traditional diet-heart hypothesis: analysis of recovered data from Minnesota Coronary Experiment (1968-73)

There was, overall, no effect on total mortality when comparing the two interventions. To rephrase that: Increasing dietary linoleic acid was no worse than increasing trans fats, overall.

OK. So we could stop there with nothing more insightful than an observation of the moral and scientific bankruptcy of the architects of the lipid hypothesis. Nothing new there.

But what we actually have here is a study comparing two diets, one with a marked increase in the double bond index (DBI) of the lipids vs one with a modest increase in DBI, if we ignore the problems of trans fats.

We have something resembling the CRON mouse study in which lard as the lipid source gave a greater longevity benefit compared to fish oil or soya oil. We can view the present study as an intervention which altered mitochondrial membrane lipid composition in a direction of enhanced ageing based on increased DBI of those membrane lipids. But this time in humans, and with no calorie restriction.

Rather than looking at specific diseases we can ask whether increasing the DBI of your mitochondrial lipids might simply make you biologically older than your chronological age. This should show in the all cause mortality data, irrespective of the cause of death (ignoring "One flew over the cuckoo's nest" scenarios, even though this was a mental hospital study). However this would be hard to isolate in younger people because they are far enough, chronologically, away from death that ageing them by 10 years (a totally fictitious value, merely used for illustrative purposes) wouldn't show up much in all cause mortality. Assuming the risk of death at 40 years of age is similar to that at 50 years of age, nothing will show.

But, if you are 65 years of age and eating 13% of your calories as corn oil derived polyunsaturated fats makes you behave biologically as if you are 75 years of age, this just might show as increased all cause mortality.

Ramsden provides us with these graphs. In people under 65 years of age nothing shows:

but in people over 65 years of age there is visibly increased all cause mortality in the corn oil subjects:

Because the raw data for these graphs could not be recovered it is impossible to perform any sort of statistical analysis but it looks to me like there might be some indication that basing your diet around corn oil PUFA might be worse than eating trans fats, late in life. Given the raw data I suspect it might be possible to calculate how much linoleic acid might shorten your lifespan and by implication I would expect it might also shorten your healthspan, which could actually be worse.

Trans fats come out unexpectedly well. You have to wonder how much more benefit removing linoleic acid might provide, especially if you are an elderly person trying to avoid ARDS in the ITU.



Rich said...

Nice post. I think you mean Fred Kummerow.

Peter said...

Oops, thank you, fixed now!


DLS said...

Just According to Keikaku

Eric said...


first of all a gem. Apparently, high cholesterol is among the co-morbidities that lead to CV deaths in NYC. I wouldn't call it a morbidity at all. Not sure if this was a statistical fluke or cognitive bias at work.

I would also like to hear your opinon on a real mystery.

ICU physicians all over the world have been finding that patients will have 40-50% oxygen saturation according to the optical monitor clip yet are alert and that their lungs are still pliant. The explanations I have seen are:

- Christian Drosten (virologist at Charité in interview by Stern magazine):
My colleagues from the ICU wards have explained that CoVID patients have to be ventilated differently from e.g. those after heart surgery. This viral pneumonia damages lung tissue at the periphery. The capillaries become congested by infected cells. The body hence does not get enough oxygen into the blood. The blood pressure in the capillaries goes up. This increases the resistance the heart sees from the lung, which it cannot keep working against for an extended time. This is why patients with preexisting heart problems often die in ICU. What is needed is a fine balance between overpressure ventilation and regulation of the blood circulation. This is the high art of internistic ICU care.
--> doesn't seem to explain functioning at low oxygen saturation

- Shirley (poster at Malcolm's blog): She might be an experienced nurse and says with a high fever, the body shuts down peripheral circulation. She goes on: This may explain the paradox. The oxygen saturation of the blood supplying the major organs in this situation is far nearer to normal than that measured by the monitor, which is usually clipped to a finger. The peripheral circulation has a poor blood flow which gets depleted of oxygen quickly hence the very low reading. An accurate central reading can only be obtained by sampling blood from a central arterial line. The brain must be getting enough oxygen if the patient is fully conscious.
The advice when working with seriously ill patients was always: ‘watch the patient, not the monitor.’
I can only explain the response of the medical staff by the fact that they are not used to caring with patients with very high fevers with peripheral shutdown in the Intensive care ward.

- Then there there was this in todays NYT.
This doc is actually calling for everybody with as much as a cough to get a pulse oxymeter. His explanation is that the virus destroys the cells that make surfactant, so the bubbles in the lung collapse while the whole lung remains pliant for a while. In this situation, the body still manages to get rid of carbon monoxide while not getting enough oxygen, which explains why patients are alert at low monitor readings.

This seems to be incompatible with Shirley’s explanation. I am not sure if it is compatible with the explanation given by the Charité ICU physicians about capillaries in the lung shutting down first.

Eric said...

oops, here's the link to the cholesterol gem:

Eric said...

This is a letter to the editor of a diving magazine by an Austrian emergency care physician who is a diver, does diving medicals and has treated a good number of patients from one of the skiing hotspots there. In contrast to early oxygen supplementation that the doctor from NYC recommends, he has noted that this seems to accelerate deteriration.

Other (young) patients exhibit the same picture and come (walk?) into the ER with a lowered oxygen saturation. Except for a higher breathing frequency that they themselves hardly notice they are doing fine in spite of having a blood gas composition that would indicate immediate intubation according to established textbooks. You give them 2 liters of oxygen and they seem to be doing better and a few ours later they end up itubated in ICU with massive lung failure. Many colleagues harbor the feeling that oxygen initiates a cascade.

Andere Patienten haben das gleiche Bild und kommen mit einer erniedrigten Sauerstoffsättigung in die Ambulanz. Bis auf eine erhöhte Atemfrequenz, die die Patienten kaum selber bemerken, geht es ihnen den Umständen entsprechend gut, obwohl sie derart schlechte Blutgase haben, dass man laut Lehrbuch unverzüglich an eine Intubation denken sollte. Gibt man jenen dann 2 Liter Sauerstoff, wird zwar die Sauerstoffsättigung etwas besser, aber ein paar Stunden später liegen viele von ihnen auf der Intensivstation mit Intubation und schwerem Lungenversagen. Das Gefühl, als ob der Sauerstoff irgendeine Kaskade auslöst, haben ganz viele Kollegen.

Frunobulax said...

@Eric High cholesterol can be completely innocent (a sign of active fat metabolism from ketogenic diets) or dangerous (if oxLDL is high). So you can't say that high cholesterol is a risk factor. It's a bit like saying high blood pressure is a problem: It isn't if you measure it during a workout. You have to make additional qualifications. Assuming that atherosclerotic patients are at risk, high oxLDL is a problem, while high "normal" LDL would be completely safe.

IMHO, the issue with intubation could simply be that intubed patients (often in coma) will move a lot less, therefore making it more likely to see pulmonary embolisms.

For those who speak German, I blogged on this yesterday:

Eric said...

@Thomas Yes, I am aware of that. That's why I ironically called the finding a gem.

Very interesting article about the Swiss embolism study. Are you aware of an English version of the Tagesspiegel article? This might be right up Malcolm's alley.



PS: Warum hat Deine Homepage einen Trauerrand? :)

Jay said...

@Eric I can't read the NYT article because it wants me to register. But it sounds like they misread or misinterpreted the data. From what I saw of the actual data (I found it in a tweet somewhere but on comorbidities and hospitalization/death risk, hyperlipidemia actually had reduced RR (0.61) of hospitalization (stat. sig. ) and made no significant difference to risk of critical illness.
Paper here:

Eric said...

@ Jay They are providing coverage on Corona for free but apparently still require you to register for free. It used to be possible to read one or two articles after deleting all cookies or using incognito mode. Also, links to their articles posted elsewhere seem to display fine. Go to and search for cholesterol. I can't be really sure if this works for everybody because I have a NYT subscription.

Last not least: thanks for the link. That's what I thought would happen based on FH research.

karl said...

Does anyone have a contact at WHO? You might send them a link to Petros ARDS post:

Ask if they really want to parrot/promote an ungrounded narrative that might kill people.

Outlaw said...

I wonder how much they are being given... enough to induce proper IR? How will MCTs act? Shouldn't viral replication deplete glycogen?

Atleast they are ignoring the WHO.

BigWhiskey said...

Guillermo Fernandez said...


This could be a plausible mechanism assuming the saturation readings are actually showing arterial values rather than peripherally altered ones (which seems more likely to me).

Peter said...


That's a good article. I can't be arsed with many of the hypotheses regarding the hypoxaemia of the current coronavirus pandemic. Respiratory physiology pretty well explains it. For years John Nunn's "Applied Respiratory Physiology" was bed time reading. Great book.


Eric said...

Are you saying none of the experienced MDs worldwide now expressing their astonishment at functional patients with low ox sat hat appropriate bedside literature?

Peter said...

Eric, absolutely. Anaesthesia is a black art to the medics. ITU consultants of course have some idea of lung function but in the past they have usually been anaesthetists, as many still are, and these people have lung function understanding as second nature. Interestingly some of the best fake doctors sneaked in to anaesthesia. It happens because the senior anaesthetists in charge of training them expect junior doctors to know nothing about the subject, hence a knowledgable hospital porter can pass muster for a recently graduated doctor when entering anaesthesia. Watch Paper Mask. . Re hypoxia: CO2 drives respiratory urgency. Given some functional lung tissue the body can maintain normal CO2 with modest effort and the patient feels fine. With an SpO2 of 80% the patient would not even be aware of the hypoxia. Athletes train at these levels of hypoxia to generate extra red blood cells without resorting to erythropoietic doping. As lung function progressively crashes both CO2 rises and SpO2 dramatically falls, and the patient becomes distressed. It occurs suddenly. Correcting minor falls in SpO2 with inspired oxygen won't stop the terminal changes in SpO2 as the degree of ventilation perfusion (V/Q) mismatch kills the patient. V/Q mismatch does not respond to increased inspired oxygen. John Nunn explains it well but few non anaesthetists will have read his masterpiece.


cavenewt said...

"Researchers from Wenzhou, China looked at clinical laboratory features including lipid levels of patients with COVID 19. They found dramatic reductions in the cholesterol levels of patients infected with COVID 19, compared with healthy controls (1)

"The study provides data to suggest that cholesterol levels decline quite rapidly during the early stages of infection and increase as the patient starts to recover. Therefore, indicating that cholesterol may have an important role to play in defending the body against such infections."

There's a link to the preliminary report (the actual paper isn't published yet) at the end of the article.

Cholesterol and COVID-19: