Carbosis and the "Seed Oil" Theory: Part I
(Contains "end-to-end" citations (where I give a link not just to a source, but to a note that quotes the relevant part of the source) marked like this: link.)
An enduring mystery known to the low carb and Paleo spheres
is how a very low fat, very high carb diet
such as the Kempner Rice Diet or the Potato Diet
can result in improvements in diabetes and obesity (Metabolic Syndrome diseases)
despite triggering mechanisms in opposition to a high fat, low carb diet,
which has been found in numerous studies to be beneficial for these same conditions.
Denise Minger wrote about this in her usual thorough and thought-provoking way
several years ago now,
dubbing the phenomenon "carbosis" and provocatively suggesting
that "magic happens" at both ends of the fat-carb spectrum.
(Do not fail to notice that the weight loss results for Kempner's diet in particular
did require caloric restriction, and that therefore they must not have been due to purely to
increased satiety, as some patients apparently required consensual whipping to stay in line.
At least that's what she said.
Other results may be independent of weight loss.)
I won't go over the evidence for and considerable nuance about those diets here (see the link above).
In this post, I'll just briefly go over a couple of potential explanations
for why carbosis could work, even though at first blush
the resulting lipogenesis would seem to be the worst possible move
for someone with metabolic syndrome.
Perhaps surprisingly, this connects to
what is now best known as the "seed oil" theory.
Finally, I'll touch on how this might relate to
the other currently popular diet trend, "P:E"
the bastard child of the "Protein-Sparing Modified Fast" (PSMF)
and the Protein Leverage Hypothesis (PLH).
The "seed oil" theory in brief
The idea that increased consumption of seed oils is the cause of the obesity epidemic
is extremely popular right now, and has some good arguments behind it.
The common crux of the arguments is that high linoleic acid (LA) intake is the cause.
Seed oils are high in LA, an omega-6 polyunsatured fatty acid (PUFA).
LA is found in grains as well as seeds,
and so a grain-based diet alone increases LA consumption over a diet wthout grain.
This means that the seed oil theory provides an alternative or complementary
explanation to the carbohydrate explanation
and the reduced nutrient explanation
as a potential causal link between
the advent of grain agriculture and the rise of metabolic syndrome diseases.
But the more recent extraction of oils high in LA for human consumption
has resulted in much higher intake of LA than grains alone could accomplish.
There is some evidence that the adoption of these oils into the food system
parallels increases in disease,
and so from a correlative point of view, the theory is consistent.
But what about mechanism?
How could LA cause obesity?
There are two (maybe) complementary mechanistic theories for why LA might cause harm.
The "Proton Theory" (or ROS theory), put forth first by Peter Dobromylskyj,
is based on energetics.
(He's been developing the idea with over 70 blog posts spanning a decade. Start here.
Essentially, when PUFAs are used for energy,
they create fewer reactive oxygen species (ROS), compared to saturated fats.
I realise that may sound superficially like a good thing,
because we associate ROS with oxidative damage,
but it turns out that like a lot of things in the body,
ROS is not actually harmful unless it goes on unopposed.
ROS has very important signalling roles,
including signalling the response of the cell to insulin — how readily it takes up carbohydrates.
So as the theory goes, with less ROS, fat cells will take up more glucose and store it as fat.
And that means more glucose is going to storage in the fat cell
instead of circulating longer as fuel for muscles. So it's fattening.
In addition to Peter Dobromylsky's blog, two excellent resources for understanding this theory
are Brad Marshall's "ROS Theory of Obesity" series,
and Michael Eades' presentation, "A New Theory of Obesity".
A second, independent mechanistic explanation is that LA and its oxidised metabolites cause direct damage.
The more LA you consume, the more it is incorporated into tissue phospholipids,
and this can cause vulnerabilities by virtue of being susceptible to oxidative damage,
imparing the function of the tissue involved.
For example, cardiolipin is part of the mitochondrial membrane,
and according to this theory, if it is overly structured with LA,
the resulting tissue changes may be implicated in metabolic diseases.
Tucker Goodrich argues for this direct damage side of the equation.
How does carbosis connect to the LA theory?
Lipogenesis depletes linoleic acid
When you eat a very low fat diet,
that obviously means you are not contributing much LA to your metabolism from your diet.
But what about LA you already have?
Brad Marshall, for example, has discussed the possibility
that having stored a lot of LA as fat, a fat person may be at a distinct weight loss disadvantage
because even when they rely solely on their fat stores for energy,
they could well have a higher proportion of LA in their fat stores.
But a very low fat diet, in particular a non-ketogenic one,
not only doesn't contribute LA,
it can actually also deplete it!
The reason is that when lipogenesis is stimulated,
it upregulates the conversion of LA to arachidonic acid (ARA).
See this study, for example.
Interestingly, in this study, rats were first fasted —
in other words, their biochemistry was steered far to the ketogenic side
where lipolysis is maximised and lipogenesis minimised —
and then afterwards put on a high carb, very low fat diet.
This caused rapid conversion in the liver of LA to ARA,
as well as other fatty acid conversions you might expect:
palmitic to palmitoleic and stearic to oleic.
Linoleic acid was essentially completely depleted from the liver in the span of two days.
Other tissues such as adipose appeared to be undergoing a similar process, but much more slowly,
such that there was only partial depletion.
An important result of the study was that these conversions
were much more rapid than when the high carb, low fat diet
was started without the fasting first.
This suggests first of all that
a mechanism in the benefit of a "carbosis" diet could be LA depletion.
It also suggests that moving rapidly from ketosis
to a high carb, low fat diet might maximise the LA depletion effect.
So an enterprising person might try interspersing a few days of, say, just starch or fruit
into their otherwise ketogenic diet as a way to attempt to deplete LA from fat stores.
A benefit to "Carb Cycling"?
The above strategy, if optimised,
would look almost identical to Lyle McDonald's description of the Cyclic Ketogenic Diet (CKD).
As you might know, I'm a long-time fan of Lyle McDonald.
His book, The Ketogenic Diet, was one of the first published books on the topic,
and I devoured it when I was finally able to get my hands on a copy in 2002 (Thank you, Zooko).
In the book, he not only summarises much of
the science on ketogenic diets available at the time of its writing,
but he specifically applies this knowledge to fat loss and body building.
He gives targeted recommendations for implementing a CKD
as well as a PSMF, which I'll return to below.
The basic idea of the CKD in this bodybuilding context is to
periodically diverge from the ketogenic diet in order to refill muscle glycogen
and maximise potential muscle gains from weight lifting.
Then you get back into ketosis as efficiently as possible to maintain leanness.
In practice it looks like one day once a week of very low fat, very high carb food
placed before a glycogen depletion workout
and then followed the next day by a return to ketosis sped up by an aerobic workout
to induce the "ketosis of exercise" (a well known effect).
While the merits of this approach to muscle gains may be debated,
it's possible it has additional body composition benefits via LA depletion,
provided the low fat part is adhered to.
I'm presuming the lipogenesis upregulation
relies in part on the suddenly lack of availability of fat from the body due to insulin
and fat from the diet.
Therefore it must be generated with high priority.
What about "high" protein?
Another possible way to do a low fat diet could be to combine low carb and low fat
leaving only protein as a macronutrient making any significant contribution to the diet.
But here things get sketchy.
We definitely need some protein.
Total fasting has risks. The problem is that there is always a base level need for protein
to take care of tissue repair in the body, making up for unavoidable losses.
If you don't eat at least enough protein to take care of those needs,
you could die from organ failure, as the body steals muscle indiscriminately to make up for it.
This has happened in recent memory.
You might think this is a bad strategy evolutionarily,
but the fact is that most animals do not have high fat stores.
So the ketogenic state in almost all non-human animals only lasts for a short time —
too short to compromise lean mass to the same degree as if you are getting a good supply of fat.
When their body fat runs out, metabolism changes again
from a fat-based one to a protein-based one,
which kicks in hunger and food seeking behaviour via cortisol.
It's an emergency situation, because if you're down to using protein for energy,
death is imminent.
This is called "Phase III" starvation; a phase humans rarely reach,
exactly because lack of adeqaute protein will kill us before we run out of fat.
(Phase I runs on glycogen stores. Phase II runs on fat.)
So the Protein Sparing Modified Fast PSMF was created to emulate fasting but without this terrible medical risk.
The more recent trend of eating high protein and low fat,
sometimes called "P:E", meaning (high) protein to energy ratio,
has in some other cases borrowed the name PSMF.
But it is not a PSMF,
as a PSMF is by definition a low protein diet.
The ketogenic diet takes the PSMF base and enhances it with more fat for energy.
This might seem like a bad idea for weight loss.
Wouldn't adding fat directly compete with losing fat from your body?
Perhaps surprisingly, not necessarily!
There are rate limits to how much fat you can release from your fat tissue
on the one hand, and how fast you can burn it on the other.
If you are very metabolically healthy, but for some other reason carry a lot of fat,
you could possibly get all the energy you need this way.
But if you have any impairment in releasing or oxidizing fat
(for example, if you have high insulin levels, or a fat disorder, such as the suprisingly common condition, lipedema),
or if your body fat is low enough, or some combination of these,
the rate just won't be enough to feel well.
You may actually get better results by adding as much fat as your appetite dictates.
After all, your appetite should be a direct reflection of how much energy intake you require,
based on what you have going.
But what happens if you add protein on top of adequate protein instead of fat?
Humans are unique in being able to stay ketogenic while protein intake is ample
at least up to a point — but that point is far above just nitrogen balance.
This creates a strange relationship to protein for humans.
Unlike the wild carnivores that have been carnivorous from many millenia
humans are new on that scene, having diverged from an herbivore lineage much more recently.
We cannot use protein for energy as efficiently.
As I have argued in my talk about the human "Lipivore",
we are specialized to use fat for energy.
Humans that try to get energy from protein develop "rabbit starvation".
So if you are eating a PSMF and are lacking energy as described above,
but you add protein instead of fat, you might actually feel worse
and even more hungry.
Instead of emulating the phase of fasting in which fat is the primary source of energy,
you are emulating Phase III fasting, the state of emergency;
the state that makes you ravenous.
This is why claims that high protein diets are uniquely satiating
are highly suspect in my view.
Protein is in some situations more satiating than carbohydrates,
but it is not more satiating than fat,
and even the "Protein Leverage Hypothesis",
which posits that higher protein foods prevent obesity
because animals overeat low protein foods until protein needs are met,
can only explain an effect when moving from protein inadequacy
to protein adequacy.
In fact, at high ends of protein intake,
fat leverages satiation better than protein,
exactly as rabbit starvation anecdotes reflect.
(I go into this in great detail in
my talk on "Goodhart's Law" as applied to diet.)
In truth, a higher protein lower fat diet does work really well for some people.
But for other people it is a disaster, causing hunger and little-to-no weight loss.
The two approaches have never been clinically tested head to head!
It's quite possible that there will be conditions and characteristics
that make one approach more likely to work than the other.
Based on anecdote and consistent with the above reasoning,
people who have fasting intolerance — feeling tired and cold and irritable
instead of energized and zen-like in response to fasting —
seem to do much better on high fat, lower protein approaches.
Potential Follow-Up depending on my time and your interest:
What happens when you add linoleic acid to a high carb, otherwise low fat diet,
vs to a low carb, high fat diet?
Index to my Free2Z posts here -- ambimorph