looks like ketones might suppress endogenous
glucose production. higher fat, and particularly saturated fat, cause greater hepatic insulin
resistance. your chance of getting remission at one year is 86 percent. here's one thing you
could tweak in the diet that can help control your glucose without it being related to weight
loss. what are the best dietary approaches to type 2 diabetes? Our Guest today is Dr Nicola Guess,
who has an extensive background of both clinical and research experience in diabetes and nutrition.
she's a registered dietitian and she also received a PhD from Imperial College London for research
on diabetes prevention. she also trained at Johns Hopkins and at Tufts where she was a Fulbright
scholar.
She's currently research program manager at Oxford University where she studies diabetes
remission interventions. we discussed the effect of different diets on type 2 diabetes and why
information on this topic seems so confusing. people get a lot of mixed messages, they go
on social media and every Guru is promising that their diet is the one that cures diabetes and
conversely that every other diet is ridiculous and even criminal. why would you put people on this
diet or that diet? and the story sounds similar from all of these sources, and it is true that
you can find some evidence, we can talk about levels of reliability and data quality but it's
true that you can find some some arguments for all these different dietary patterns in terms of
putting diabetes in remission and things along those lines.
You must get these same questions
from your patients all the time, I wonder how you explain this this confusion to them. yeah,
I mean generally what we know definitively in terms of remission, so cure I don't, I prefer to
use remission because I think it describes what's happening most accurately, it's that right now you
don't have diabetes according to the diagnostic criteria but it doesn't mean that it's not going
to come back, so I think that's the right term to use, and the best evidence we have now is that
weight loss, however you achieve it, probably, is the primary determinant of remission, so
there's that wonderful graph in the direct trial where if you lose 15 kilograms or more your chance
of getting remission at one year is 86%, and it goes down if it's just 10 kilograms or more, and
down further, so you can see there's kind of a dose response effect of weight loss on remission
and that's a pretty strong bit of evidence.
Right now that trial is the best one done so far, it
was with a meal replacement product, do I think it means that you need to use that diet to get to the
weight loss? no I don't. everything that we know about physiology and short-term trials suggests
calorie restriction and weight loss, however you achieve it, are going to help you get to your
goal. so that's what I tell my patients, and there are a million ways to lose weight. there are two
things that I think are interesting independent of weight loss, and the first one is what my research
is actively looking at and that's dietary protein, and the basis for this is that what we think from
the direct trial and accompanying physiological studies is, what´s getting people remission is
they get back the first phase insulin response and that is, let's call it a spike, like a massive
spike of insulin that basically happens after you have a glucose load or a meal, so it goes
up, it brings down glucose straight away, and it helps you control your postprandial
glucose concentrations in particular, and what the DIRECT study showed is,
with weight loss, people who did get remission and didn't get remission both
had improvements in insulin sensitivity, they both had improvements in liver fat, they
both had improvement in pancreatic fat, but the distinguishing factor was the first phase insulin
response.
If you got back the first phase insulin response, you got remission, if you didn't get
the first phase insulin response back, you didn't, so this pulse of insulin after a meal seems to
be really key for controlling blood glucose, and so what's interesting to me is that protein,
specifically the amino acids and protein, nudge the pancreas really powerfully
to produce more insulin after a meal, so that's one thing that I think is intriguing.
even more intriguing is that the pancreas or the beta cells remain responsive to amino acids
even in late stage diabetes, so what we know with regard to beta cell functional failure in type
2 is they don't respond to glucose very well, so even if your glucose goes up after a meal,
you have two slices of bread whatever it might be the sensor on your beta cells no longer
detects the glucose as efficiently as before so it can't send the signal through to produce
more insulin but guess what, amino acids do, so the signaling Pathways of amino acids on the
beta cell seem to remain intact even if you've had type 2 diabetes for 10 years, and so that's really
exciting because it probably means, I think, my hypothesis, that higher protein can lower your
blood glucose independent of weight loss, there's a lot of evidence that it does, but secondly, it
might be an effective approach for people who've had type 2 diabetes for a long time.
So that's my
basic description of where I think the evidence is. there is some some earlier data on plant-based
diets, super high fiber and high carbohydrate as well, on achieving normal glucose concentrations
in type 2. I don't think that evidence is as strong for everybody else but it's a way of losing
weight and many of my patients go that route. so then does that mean that in addition to the
general advice of whatever sustainable diet helps you lose weight, you also advise
a substantial intake of protein together with that? yeah I do, with with most of my type
2 patients absolutely yes I do, I mean not only do I think it is physiologically a powerful tool
for glucose homeostasis per se, there's moderate maybe, let's call it semi-poor quality evidence,
that it might help with weight management, and I say that just because the effect size is pretty
small, like on average if you look at weight loss maintenance trials it looks like higher protein
diets do a bit better but it's not a huge effect, so I don't really hinge my treatment on that
because I don't think it's very strong data but you know, if you have a patient with overweight
or obesity and type 2 and NAFLD, because lots of patients with type 2 diabetes have high liver
fat, protein helps you fix those three things, but one thing I would add, it works pretty
well because I deal with a Western population, people are happy to have more protein, it's
quite acceptable, but I always consider the cardiovascular side of things, so I really try and
promote half animal-derived, half plant-derived protein because you get a good mix in there, if
you're going for the plant derived proteins it's fiber and all of the other goodness as well, but
it's frankly easier to meet your protein targets with animal-derived products for most patients.
yeah, molecularly do we know how the glucose, presumably the insulin signaling Pathway is
blocked in the beta cells but the protein response seems to be preserved, do we have any idea how
the selective preservation of the signaling works? no we don't, but I have a paper
hopefully coming out soon, it's basically a review, and the whole purpose of
it is I think and my co-authors believe, we have ignored the potential of amino acids for promoting
an insulin secretory response, and actually even more interesting as well and scientifically
intriguing in the sense that maybe this is where nutraceuticals could go is that glucose is,
there's only glucose, glucose-stimulated insulin secretion is just one thing, it's the glucose
and it's the coupling to produce a release of insulin.
There's 20 amino acids, and we think they
all work slightly differently so… I was going to say the stimulus secretion coupling of glucose is
pretty well characterized and it's easier because there's just glucose, with amino acids there's 20
of them, there are shared and distinct pathways, so they're shared with glucose-stimulated
insulin secretion but some pathways are distinct, the different amino acids probably work through
slightly different pathways and that probably also explains the potentiating effect that, because
they work differently, adding a couple of amino acids together, maybe just three of them, you
could get the maximal insulin secretory response compared to whole protein, and we think that's
where that could be a really exciting path for nutrition research. in terms of quantifying, if
we were to compare the effect of weight loss to the effect of manipulating protein intake, what
are we talking? whether it's rate of remission or however we would quantify that to put the two in
relative terms. so I can't answer that question because we simply don't have the data, and the
studies that have been done are not comparable, the thing about weight loss that is worth
acknowledging is it does improve hepatic insulin sensitivity, improves muscle insulin sensitivity
and it seems to fix beta cells if you've had type 2 diabetes for short duration, you know, it
improves a bunch of cardiovascular risk factors, and the level of certainty there is much much
stronger.
What amino acids do, I think, the most exciting part of it is just almost like an added
benefit, almost an adjunct, you know, so I don't think you would ever in many cases intentionally
design a clinical program where you wanted to maintain someone's weight loss, hopefully
you'd still get some weight loss, but for me this adjunct of adding amino acids might mean that
instead of having to lose at least 10 kilograms to get remission, maybe if you could just get to five
to seven kilograms, the added benefit of amino acids on that insulin response might help you
get there, and I think if we could achieve that, we would increase the number of people at a
population level who we could help get remission. in terms of an increase of protein, that would
be a swap with some of the carbohydrates, some of the fat, or does it not matter? I
think it probably has to…
Hopefully, one ongoing study I'm doing is trying to tease this
out a bit, what the ideal proportion should be, and I've got a few other studies planned. based
on the data available, I think you have to be about 30% kcals from protein and about 30% kcals
from carbohydrate, probably as a maximum, and the rest fat, and again let me reiterate, I always
emphasize predominantly unsaturated and minimize saturated. it's worth mentioning, we don't
do dose-response studies enough in nutrition, and the reason I say 30 is because that's what
the studies that looked at this looked at, and the comparative group was 15 or 17. so maybe
25% of calories from protein and 30% calories from carb would do the same, we don't know if
there's a cut off so the higher or the closer my patients can get to 30%, great, but if it's a
bit of a struggle I don't worry too much about it it can be tough to get enough so I give out
diet sheets and I call things primary proteins and secondary proteins and I emphasize that they
should try to get both in every meal particularly lunch and dinner, breakfast is the hardest so
for breakfast I might even encourage a protein powder shot or having protein powder in a cereal,
whatever that might be, but for lunch and dinner what I would be looking at, the primary protein
would be something where the vast majority of the calories in it are just protein, so you're
talking chicken breast, tofu, mycoprotein, that's like a corn vegetarian alternative, where there´s
not much fat in it, there's not much carb in it, it's just protein.
I call that primary protein
and that's the easiest way of getting enough, and then in addition I have secondary proteins,
so those might be things like pulses where yes, they've got a decent amount of carb in but they've
also got enough protein that if you combine the two you can get closer to your 30% of calorie
mark. we've talked about what the best diet would be and it's not really a a name diet as
people like to throw around on social media but technically any of these name diets would
work. low carb, low fat, low calorie, intermittent fasting, all these approaches
could work and I think we see evidence that all of them do work at least for a subset of
people, and by the way I've always wondered if the fact that we only see them, all of these
pretty much work for a percentage of people, if that means that it's the percentage of people
that are attuned to that intervention or is it the percentage of people that would do well on any
intervention because they are the most disciplined ones or the ones that stick with whatever you
give them, do you have an opinion on that? I mean, I think based on the type of studies that have
been done, and they are typically free-living, very rarely do studies provide food so the onus
is on the participant in the study to stick with the dietary intervention and so forth, so for
me the evidence is on the side, based on the the type of trials that have been done, that it's
people that would probably do well on most diets, but nonetheless one of the things that I always
find striking is the amount of extra support that you can have in any weight loss trial and so I
mean I've worked on lots of weight loss research and even when you're putting out an advert,
so let's say, I did my PhD, I did my postdoc at Imperial College, it was well known for doing
really exciting work in appetite hormones so they develop, they do first in man studies of hormones
that reduce appetite so it had a good reputation for research, and whenever you put out an advert
say in a newspaper to recruit participants, people are so desperate for something that works and they
see that it's, you know, whether it's Imperial College, whether it's Cornell, whether it's, you
know, a big name, they think you finally got the answer, so the first thing is, the response that
you get are people who are absolutely desperate for weight loss, they are so determined to make
this work, and then you consider that they come into an environment where they get way more
input than anyone will ever probably be able to give them in Primary Care, and you have what
should be a state-of-the-art diet to the extent that that exists, and yet, dropouts are about 30
to 50% from trials even with that, and you know we see the weight loss is generally pretty modest,
until DIRECT, I mean 10 kilograms at one year was incredible, and by the way I don't want to be
biased, like VIRTA, it wasn't a randomized trial but it was a single arm open label trial, they
got 10 kilograms in one year so you can do it but that's rare, most weight loss trials it might be
five, seven kilograms at one year and from then on it's straight up again, so I guess what I'm saying
is weight loss is really really hard and yes, to your point, I think the people that do well
in these trials are people that would do well on probably any dietary intervention, at least
over the short term.
So in terms of the clinical approach, when you get a new a new patient coming
in, do you explain that end goal is going to be to lose x amount of weight, do you give them
a repertoire of possible dietary approaches or do you have a default that you suggest, how
does it work? yeah, a really great question, it's really a lot of listening and very often,
often the first thing is you have to understand how much a patient knows, your average patient
knows more than me so you don't want to be as the clinician like telling a patient "oh by the
way this has this many calories" because your patient's going to be super knowledgeable so the
first thing to do is gauge how much they know.
The second thing is I try to understand their goals
and this is important because remission might be the gold standard as far as the NHS is concerned.
Why do we want that? because it saves money, gets people off medications, lowers medical
costs and so forth but that might not be what a patient wants so you have to you know, what
are their goals, what do they want to achieve? and then you also want to try and understand
what are they willing to do, to give up, because many interventions are tough, and so you
want to gauge all of this but in general yes, what I would do is I would explain,
if remission is what they're after, I would explain the research in as lay fashion
as I can and I would say, here's what we know, what do you think about that, you know, let the
patient lead that consultation, and then we would try to tweak things.
I do always try to educate on
protein because I just think, I'm a pragmatist and as much as weight loss definitely helps, and 10
kilograms does and so does 15 kilograms, people do struggle to lose that amount of weight and
then they very often fight against regaining it, so as a pragmatic approach, kind of saying, well
here's one thing you could tweak in the diet that can help control your glucose without it being
related to weight loss and that can be empowering for them to see their glucose go down and their
A1C go down. so you try to gauge which foods are very important to them and which foods they could
take it or leave it. and then maybe focus on those to phase out. yeah exactly, and I'm always really
honest, I mean sometimes I might just have like meal plans or pictures of foods and talk about
what the approaches involve because for example, like the meal replacement approach, you're talking
two to four months of meal replacements and people differ immensely and they tend to be quite black
and white about this, whether they're going to do that or not, I mean some people just go hell
no, I'm not eating milkshakes for three months, are you crazy? and it's a non-starter so you might
say, well what about doing two meal replacements in a meal, but you would work with them in
that way.
If someone says, oh I love carbs, I'm really into bread, really into potatoes,
or culturally it's really important to me, it's how my family eats, there are certain things
that just aren't going to work, because you're running uphill as it were. yeah that makes
total sense. going back to this idea of susceptibility to different dietary patterns,
do you ever get patients who don't do well on maybe the first thing they try, maybe they try
a low fat diet first and it just doesn't work for them and then try a low carb diet and it
works really well, or the other way around? I mean in general, patients are often fighting
with their weight as a lifelong battle and I can't remember the last time I met a patient that hadn't
at least tried those two. so very rarely is it trying something as straightforward as low fat
or low carb.
I like the tweaking approach and kind of food exchanges, so whether it's replacing
some saturated fat with polyunsaturated, usually with cardiovascular risk that can be useful for
patients, whether it's replacing carbohydrate with protein, that's a switch that's not related
to cutting calories, whether it's increasing the amount of dietary fiber they have, these
are things that fit into any dietary pattern, so even if they've tried something and it hasn't
worked, or maybe they might have done keto and it helped them lose weight but their cholesterol went
up and their doctor said ´forget about it´, like trying to work within a dietary pattern that
a patient likes and then using those tweaks generally works better than low fat over here and
you go to low carb, so that's generally how I try to work it. yeah so not being bound by the labels
but rather flexibilizing these things to whatever works for the individual. for sure, yeah. in terms
of people on different stages of the progression, let's say someone who is healthy and wants to
prevent pre-diabetes or diabetes versus someone who might have pre-diabetes versus somebody who
has established diabetes, is there a fundamental difference in terms of the dietary approaches
or is it all the same? I mean that is such a great question, so certainly with type 2 diabetes
because you have this beta cell functional failure and that's the primary reason
the glucose is all over the place and we now have a growing amount of evidence in
that population, so to me they're quite different, if you go back to say someone with normal
glycemia, do I think they'd have benefit from doubling the proportion of calories from
protein? no I don't, and the reason being is that beta cells work just fine, they produce
enough insulin and so there is some evidence, I'm not totally convinced by it, suggesting that
high protein causes peripheral insulin resistance, if it does happen the effect size is probably
small, so I kind of weigh it up like this, if I think about someone with type 2 diabetes,
the benefit of a high protein approach is that you're going to fix the beta cell which is the
huge problem and probably simultaneously not have much of an effect on insulin sensitivity, but
if it does, it's like, say you worsen insulin sensitivity by this much but you're improving
beta cell function by this much.
Sorry, trying to capture that. your net result is a
win for glucose homeostasis. conversely, if you take someone who's got normal glucose tolerance,
you're not going to get any improvement in beta cell function that means anything because like
I said the beta cells work fine, but even if you do in a healthy person increase insulin resistance
this much, well the net result is you´re possibly, possibly worsening their prognosis, and it's just
almost not necessary unless you are exercising a lot, over the age of 60, 65, I don't see
any reason for having any more protein. okay, so the protein would be kind of a more
aggressive step that you would only go to if the person already has established diabetes? yes,
and then to the middle point, my answer is going to be non-committal, the reason being we have
such a poor amount of data in that population, all we know, the diabetes prevention trials
have shown us weight loss of five to seven percent however you pretty much achieve it
seems to prevent the development of type 2, there have been some smaller studies trying
to isolate, well is it the fiber? is it the saturated or polyunsaturated fat? none of it's
convincing, it just seems to be weight loss, I'm interested in the idea of protein helping to
keep glucose under control because what we do know is you do have some deficit of beta-cell function
in prediabetes but some of the things that we don't know is what the long-term effect of high
protein might be on beta cells that work okay, and I think we should be cautious as scientists
and not jump ahead and say, well it works fine in type two so let's try it in pre-diabetes.
Right.
yeah so the focus would still be on the general body weight and the amount of body fat that the
person carries, and for prevention it would be try not to put it on and for pre-diabetes/diabetes
it would be trying to lose some. yes exactly, and I think the intensity of the
intervention differs or increases as you go along the pathway to type 2, so actually prevention of
weight gain is relatively easy compared to weight loss, prevention of pre-diabetes is easier than
prevention of type 2 once you've got pre-diabetes, so the interventions can be far less intensive and
probably for most people general healthy eating, I know that sounds so vague and unspecific but
that's basically the truth, exercising, not gaining weight, eating a reasonably healthy diet
within a wide boundary of choices is fine for most of the population to prevent the development of
pre-diabetes. yeah, if we were to insert another category in the middle, which is people who might
have insulin resistance but still normoglycemic, so the pancreas is still able to compensate and
maintain normal glycemia, there would it still be an issue of essentially a little bit excess body
fat but just not enough to be pre-diabetic? for me when I think about insulin resistance, to me
that is just synonymous with overweight, I think that the proportion of the population who have
overweight and obesity is pretty similar to the proportion who have insulin resistance, and it's
worth mentioning there's no actual, that I know of, consensus definition of insulin resistance.
It
might be a fasting insulin above a certain range or the insulin:glucose ratio for example, but
in general those two things track so closely, there hasn't been a whole bunch of decent
research on nutrition independent of weight loss on insulin resistance, and part of the
reason is like I said there's no definition, you could measure fasting insulin, and I think that's
perfectly good, from all of the studies I've seen, weight loss has such a huge effect on fasting
insulin that nothing else matters that much, so again, it's a question of lose weight however
you can.
We also talked about effects of specific diets aside from weight loss, like for example
low carb diets or low-fat high carb diets, which viewers often ask about. now there's
less evidence on that aside from weight loss but there are some studies so we discussed
possible effects of those dietary patterns there is some evidence in healthy people that
going on a keto-style diet, and to my knowledge most of these have been like high poly low
saturated fat, kind of healthy keto diets, do lower insulin independent of weight loss
so that's something that's interesting because it seems to do so both in the fasting and the
postprandial states but I don't know if there's long-term data, not that I necessarily think it
would stop happening over time.
I don't think that lowering carbohydrate in the diet independent of
weight loss, unless you get to ketogenic amounts, lowers glucose independent of weight, so I think
where we have these trials where it looks like low carb does really well and lowers glucose
independent of weight, I think it's the protein. where I think low carb lowers glucose under normal
protein intakes is where it's ketogenic, because it looks like ketones might suppress endogenous
glucose production, so I should just say that, so when I'm talking about low carb diets with you I
basically mean a high protein low carb diet, and I think by the way that's what most low carb people
are doing, we just call it low carb and people forget that it's actually high protein and I think
it's the high protein that's the active ingredient for the crowd that loses weight and who puts their
diabetes in remission by eating a boatload of, like low-fat high carb diets, a lot
of times they eat a lot of fruit, a lot of starch, but they cut out the fat, is
that just because it's an elimination diet so they're losing weight and that's the mechanism
by for the remission of diabetes in those cases? I mean, I don't know for sure but whenever
you see the large amount of weight loss and we know of the dose-dependent impact of weight
on remission, it's really hard to attribute that normalization or lowering of glucose to anything
else with any degree of certainty, and the same is true for any plant-based approach where you
see this marked weight loss, because it's like, how do I distinguish these two factors, you
can't, it's just speculation, I mean I wouldn't be surprised at all if all of those beneficial
compounds in a plant-based diet are improving some insulin signaling mechanisms in the muscle,
do something good to the beta cells, I mean that wouldn't surprise me, I would think the impact
of that is small, what we know from for example clamp studies is if you feed more carbohydrate you
have greater uptake of glucose into the periphery and so I would have thought if you switch
someone to a high carb diet you improve their peripheral insulin sensitivity so that could be
another mechanism but before someone has, in the overweight or obese phenotype of type 2 diabetes,
a weight neutral study, high carb High plant, I can't speculate further that it does
anything much beyond or added to weight loss you can measure what degree of insulin you
need to shut down hepatic glucose production and so if the amount of insulin you need to do
that goes up, we call that insulin resistance specifically of the liver, and from memory,
some of the studies that look at higher fat, and particularly saturated fat, cause greater
hepatic insulin resistance, so effectively you need more insulin to shut down glucose production
from the liver and in that sense what we might be looking at with those studies isn't necessarily
a positive effect of excess carbohydrates on insulin sensitivity, it's the negative effect of
a higher fat diet on hepatic glucose production for prevention of type 2 diabetes and cardiometric
disease, maintaining a healthy weight and doing a good amount of physical activity are so
above and beyond in terms of effect size of your risk of type 2 diabetes and everything else,
nothing else really matters, I mean does dietary fiber probably help a bit? probably.
Does having
more polyunsaturated or mono than saturated fat probably help a bit with insulin sensitivity? yes.
but the effect of not gaining weight and keeping pretty active, they just slaughter everything
else in terms of risk of type 2 diabetes. okay quick summary of all that. the strongest evidence
points to weight loss, however you get there, so that explains why we hear all these different
anecdotes for completely different diets, there's no contradiction, one guy lost weight on a
low carb diet, another on a low-fat diet, another by fasting and another by counting calories.
They
all got to the same destination using slightly different roads. in addition to that, there
might be specific effects for some nutrients independent from calories and weight loss,
for example protein and amino acids stimulate insulin production by the pancreas, ketones
May inhibit glucose production by the liver, unsaturated fat seems to be a little
better than saturated fat for liver insulin resistance and carbohydrate may improve
insulin sensitivity peripherally, in the muscle. several viewers have asked previously about these
classical studies looking at very high carb diets improving glucose tolerance even in the absence
of weight loss, and it's not completely understood but this might help explain those findings.
But
again, the effect of these specific nutrients is relatively small compared to weight loss, so the
role of protein is an adjunct, as Dr guess called it, knowing about this may help some people reach
their goals, others may not need to increase their protein and go on these very high protein diets
if they can lose weight and get to where they want with other dietary patterns. and we looked at
several examples of specific dietary patterns, we looked at a high protein one with 25 to 30 percent
of calories coming from protein and another 25 to 30 coming from carbohydrate, and Dr guess uses
that routinely, we also touched on ketogenic diets and high carbohydrate diets, so usually low-fat
high carb which some people do well on. all three of those dietary patterns can be designed to check
all the main boxes: low in Ultra processed foods, decent amount of fiber and mainly unsaturated
fats.
Most of the time diets can be tailored to support all aspects of health at the same time.
these examples are not exhaustive and they're just tools in your tool belt, they're not shackles
so don't let your hands get tied by these labels, if something's not working for you, change it.
just tweak it to fit your goals. so we heard from Dr Roy Taylor a while back also on type 2
diabetes remission and now we heard from Dr guess, and you see that there are some patterns that
emerge, right? so this is why it's really cool when we get the opportunity to talk to different
people all with a very high level of expertise in a field.
and we also see some other details pop up
around the edges, like these roles of the specific nutrients, so we're working towards a unifying
model of type 2 diabetes. really cool, right? and for people working on prevention, so completely
healthy right now and want to stay that way, maintaining healthy body weight is number
one as far as prevention of type 2 diabetes, and again, that can be done with a variety
of dietary approaches. I know you guys have questions about diabetes in thin people, type
1 diabetes, continuous glucose monitors, CGMs, especially in healthy people, and a lot of
other questions, and we have videos on all of that in the pipeline, so stay tuned, thanks
for watching, I'll catch you next week, bye.
