Antibiotics and Obesity – AMNH SciCafe

January 6, 2020 0 By Bertrand Dibbert


I’m really glad to be here to talk to you
about the human microbiome, especially in relationship to what happens to the microbiome,
the ways that we’re changing it. And I hope to show you that, as Susan mentioned, that
there may be some important consequences. Here’s a map of Greenland. Two maps of Greenland
in 1992 and 2002, and we all recognize this as evidence for Global Warming, for climate
change. Now, actually based on the election yesterday, there are some people who may not
recognize that, but here in New York we know what’s true. So we can think of this as a
change in our macro-ecology. And what I want to talk to you is about a change in our micro-ecology
that I believe is on the same scale. So let’s begin at the beginning. What we know is that
ever since there were animals on this planet, they have had microbes living in them, residential
microbes. This goes back five hundred million years, maybe a billion years. The microbes
don’t live just anywhere. They live in particular niches. That means that the microbe on your
arm is different, for example, than the microbe in your belly-button. We know that some of
the organisms are persistent; that means that we acquire them early in life, and they persist
weeks, months, years, decades, for our entire lifetime. And there are others that are transient,
short-termers. Some of the microbes are conserved. That means that there are microbes that every
person in this person shares, and there are microbes that are host-specific. What that
means is that, it is likely that even with this large number of people, every one of
you has some microbes that no one else in this room has. So that creates a very interesting
tension. Most of the cells in the human body are not this red leg, the human cells, but
they’re the microbial cells. And even more so, most of the genes in our body are microbial.
Human genome 23,000 genes. The human microbiome on average, about 2,000,000 unique genes.
Somewhere around 99%. So we can ask a series of biological questions about the microbiome.
What is the identity of the microbes that populate their hosts, who are they? What are
they doing? What’s their function? How is the host responding to them? Ideologically.
What are the forces that maintain equilibrium among the populations, why don’t they kill
us? Why don’t we kill them? How can we live together for so many years? We’re particularly
interested in that question. What are the unique characteristics of each individual,
how do we differ from one another microbiologically? And most importantly, how can we manipulate
this? How can we advance the science in this field so that we can understand the answers
to those previous questions so that we can prove how, and fight disease. When babies
are born, their microbiome is very un-adult-like, and then gradually, and progressively, it
becomes more and more adult-like until it is. And what’s important is that the age of
this transition is the age of three years. So this first three years of life is the time
when the microbiome is developing. It’s the most dynamic, it’s the most different from
adult, and our course this first three years is the time that babies are developing, also.
They’re developing their metabolism, they’re developing immunity, cognition. How long has
this been going on? This is a phylogeny, this a family tree of these wild hominids and you
can see there’s a certain pattern of branching, and this is what we accept is the ancestry
of the hominids. And here, is a family tree of the microbes of the fecal microbiota in
these same animals. And if you look carefully at the branchings, you see that they’re all
the same. So this property, this congruence between the two family trees means that the
transmission of the microbes, just like the host genes, the transmission is vertical,
just like their human or their monkey genes that they get their microbial genes. So this
tells us it’s inherited, it’s from Mom, and it’s also evidence for co-evolution. It’s
been going on, this family tree of hominids is about four million years of evolution.
It begins at the moment of birth. As the baby passes from the womb through the birth canal,
they acquire their mom’s microbes, we all, by in large, have acquired our moms’ microbes
where, as we go out the birth canal, we’re covered by them, we swallow the microbes,
then the baby comes into contact with the mom’s skin, and the mom’s breast, and the
mom is kissing, and licking the baby, and this is the inter-generational transfer of
microbes for us mammals it’s been going on for hundred million years. And it begins starting
at the moment of birth. That’s how it’s been. But now, in these modern times, moms aren’t
the same as they used to be. They live in an environment with antisepsis, antibiotics
that are selecting different organisms in them, their diets are different than they
used to be with different antibacterial materials in the diets, and other factors. And babies
aren’t the same as they used to be. In fact, today, in the United States, 32% of babies
are born by Cesarean section, that’s one in three. In some places, like Rio de Janiero
and urban China, it’s about 50%. So, this is a huge change in the beginning of the microbiota
and then the babies are washed extensively, very early, they are given formula which resembles
human milk in color but not in many other ways. And very importantly, babies are getting
a lot of antibiotics, and I’m going to talk about that. So the whole, this whole early
period of transition, the inter-generational hand-off, is very much impacted. And I want
to introduce a concept that we’ve been developing over the last, more than a decade, which I
call the “Disappearing Microbiota Hypothesis”, and it has four tenants. Beginning in the
late 19th century, changing human ecology has dramatically altered the transmission
and maintenance of our indigenous microbiota. These changes in ecology have affected its
composition, and we have examples of this. The altered composition affects human physiology,
and thus disease risk. We have some examples of this. And very importantly, loss of ancestral
bacteria usually acquired early in life is especially important because it affects a
developmentally-critical stage. So this is the hypothesis, about five years ago, Dr.
Stanley Falco at Stanford and I extended it and our view was back in the old days, whenever
that was, women had a normal ancestral microbiota. But if through the course of life she happened
to lose organisms, but it could not be restored horizontally from other people than the next
generation would be born with less of that microbial ancestry and so on, and so forth,
over generations. And so our view is that it in fact changes our cumulative across generations.
That it has been stepping down, that we humans are losing biodiversity, especially in developed
countries like the United States. Recently, I found a study in Japan, that supported this,
this was a study of an organism that used to be the major, dominant organism in the
human stomach called Helicobacter Pylori, and this is from a study three generations
in Japanese families, and here were the grandmothers, here are the mothers, and here are the children.
So we have evidence with this organism for this kind of projected step-down. So what
are the consequences of this? Well, I want to talk for most of the time, about obesity.
Because as all of you know, obesity has become an enormous problem in the United States and
in fact around the world. So this slide looks at obesity trends in U.S. adults. These maps
are from the Centers for Disease Control, and in 1989, there is no state in the United
States where more than 14% of adults are obese. By 2010, there is no state in which fewer
than 20% of adults are obese, and the national average about 30%. Unfortunately, this also
looks like the election map. Now what’s remarkable to me is that it’s happening everywhere, and
not just in the United States, it’s happening all over the world. And the other point is
that there’s only twenty-one years between these two maps. Something extremely powerful
must be going on to produce a change of this scale. Something very widespread. This is
the percent of obesity in kids of different ages, and the point is, it’s going up. Even
in the youngest kids, it’s going up. And this is representative of a large body of data
that the signals from obesity are beginning early in childhood, just when the microbiota
is developing. As indicated from the title of the talk, or of the invitation, I’m very
interested in antibiotics because antibiotics were discovered in the 1920s, the 1930s. They
first began to be used in the 1940s and have been used widely ever since. And so the timing
of the introduction of antibiotics fits with the development of obesity and some other
diseases, as well. So let’s look at some of the data on antibiotic use. This slide looks
at the top eight prescriptions for children in the United States in 2010. And of the top
eight, five of them are antibiotics. And if you add these numbers of courses of antibiotics,
it’s more than forty million. More than forty million courses of antibiotics in our children
basically every year, and that’s just for the top five. Now, in fact, antibiotics are
available everywhere. In the United States, you need a prescription to get antibiotics.
But in many countries of the world, antibiotics are available on demand. This slide is from
a small village in the Amazon, it’s a picture I took with my phone, in between the drinks
and the batteries, were the generic over-the-counter antibiotics. In China, it is estimated that
antibiotic use, which does not require a prescription or a physician, is two to four times as common
as ours. It’s higher, we believe, in India, as well, but we just don’t have the numbers.
So, wherever our civilization has gone, so have antibiotics, because they’re important,
life-saving drugs when they are used for life-threatening illnesses. Now recently, the Center for Disease
Control did a survey of antibiotic use in the United States and what they found is that
there were 258 million courses of antibiotics used in the United States. That’s 833 courses
per thousand population, or five courses for every six people. And according to the CDC
investigators, this has been going on year after year. Now the antibiotic use isn’t uniform
by age. Kids in the first two years of life have the highest rate of use. It goes down,
then it comes up as people get older. The average child in the United States in the
first two years of life, is getting almost three courses of antibiotics, and by the time
they’re ten, they’ve had ten courses, and by the time they’re twenty they’ve had seventeen
courses. Again, this is across everybody, some kids have more, some have less. But these
are big numbers. And in fact, these numbers are consistent with many other surveys, albeit
those are much smaller surveys than this huge nation-wide survey. Now, recently a colleague
of mine, Dr. Jeff Gerber at the Children’s Hospital of Philadelphia shared with me data
that is soon to be published. Looking at variation in use of antibiotics, prescribing of antibiotics
by doctors associated with CHoP, which is one of the best children’s hospitals in the
United States. They’ve looked at twenty-nine pediatric practices associated with CHoP and
350,000 child visits and they found between the least-prescribing and the top-prescribing
practice is 100% different. There’s a two-fold difference between top and bottom in the rates
of prescribing antibiotics. And when they look at prescribing at broad spectrum antibiotics,
there is a four-fold difference. So all doctors aren’t the same. There’s a lot of variation
in antibiotic prescribing. Now, if we go back to the CDC data and ask about the geography
of antibiotic use, we find some interesting trends. The national rate is 833 per thousand.
In the northeast it’s 830. It the Mid-West it’s 868. In the West it’s 638, much lower
than the national average, and in the South it’s 936, it’s much higher than than the national
average and in fact, between the South and the West, there is a 50% difference in the
per capita rate of antibiotic use. And as a specialist in infectious diseases, I can
tell you that there is not a 50% difference in the rate of series bacterial infections.
This difference reflects practice and culture. Now, what I found especially interesting was
to compare the two maps. The CDC map in 2010 of obesity, and the map of antibiotic use.
You will see that there is a lot of resemblance here. Enormous resemblance. Strikingly non-random.
Now these are observational data. They don’t tell us whether A causes B, B causes A or
C causes A and B. But they certainly are consistent with the idea. So because these were observational
data, we decided that we were going to try to do some experiments to better understand
the relationship between antibiotics and obesity. And we decided to study animals. For the last
seventy years, farmers have been feeding antibiotics to cows, to their livestock, as growth promoters,
to make them bigger, to bring them to market earlier. The farmers use very low doses of
antibiotics, what we call sub-theraputic antibiotic treatments, or we abbreviate as STAT. And
they use them for growth promotion of their livestock. They have shown that they can fatten
up their chickens, cows, pigs, sheep, just about all the livestock that they try. And
this is a big swath of vertebrate evolution, it’s not just mammals, it includes birds.
Over the years the farm scientists have investigated a broad class of agents and they find that
virtually every antibacterial agent they try works regardless of the chemical structure
of the antibacterial. Its class, its target, or its bacterial spectrum. Antivirals don’t
work. Antifungals don’t work. Specifically, antibacterials. The earlier antibiotics are
started, the bigger the effect of growth rate, and on feed efficiency, this is the ability
to convert food calories into body mass. And you want that if you’re a farmer. But maybe
not if you’re a parent. So we began a series of studies in mice in which we gave mice STAT
levels of antibiotics or not, we examine their characteristics, which we call phenotypes,
we analyze their microbiome, and we look for relationships. So we’ve been working at this
for a while and I want to show you the results of about five studies. The first study was
led by Dr. Il-Sung Cho when he was a fellow in my lab, he’s now a member of the faculty
at NYU, and Il-Sung gave in their drinking water mice four different courses of antibiotics
at levels that are used on the farm, that the FDA approves. And here you can see these
big increases in body fat percentage and we can illustrate this by what’s called a dexa
scan, where the fat is colorized in yellow, you can see this as an example. We postulate
that the antibiotics are mediating differential selection of the microbes in the colon. And
that’s when antibiotics, they select, they kill some more organisms, they suppress others,
and that allows certain bacteria to thrive. That’s natural selection. We showed in Il-Sung’s
study that the antibiotic dosing changed the composition, the population structure of the
bacteria. It altered the representation of genes involved in the synthesis of short-chain
fatty acids, which the bacteria and we use for energy. We show that there was increase
synthesis, or production of the short-chain fatty acids and this goes by the portal blood
circulation to the liver and we show that induced genes in the liver to create fat.
And then we showed that the mice were fatter. So, on the basis of these studies, Laurie
Cox, who is a graduate student in the lab, did a series of experiments, I’m going to
show you a few of them. And the first experiment I’m going to show you is an experiment that
we called “Fat STAT”, where we asked, “What’s the effect of combining a diet that’s high
in fat, it’s high in calories, what we call a high-fat diet, or HFD, and STAT on body
composition?” So Laurie began her experiment with two groups of mice. Either getting penicillin
in their drinking water, that’s the STAT group or not, that’s the control group. Everybody
got normal mouse food for the first seventeen weeks of life and then half of the mice were
put on a high-fat diet and half of the mice remained on a normal diet. We looked at male
mice and female mice. We looked at the total mass, how big they got, their fat mass, and
their lean mass, how much muscle. Here this black line is the group that got normal chow
and no antibiotics, that’s the control group. And the group that got antibiotics were bigger,
just like happens on the farm. And if they were put on a higher-fat diet, they got bigger
still, and if they were on a high-fat diet plus antibiotics, they were the biggest. So
there’s an additive effect. When we look at muscle mass, we see that they put on more
muscle, just like the animals on the farm. And when we look at fat mass, we see the big
accumulation in fat happens in the mice starting on the high-fat diet here, there’s an additive
effect. So up to this point, the mice had been getting antibiotics for all their life.
But people generally don’t get antibiotics for the whole life, they get shorter courses
of antibiotics. So next, Laurie did an experiment that we called “DuraSTAT”. She asked, “Is
the adiposity, is the fatness durable with only a limited antibiotic exposure?” So in
addition to the no-antibiotic group and continuous life-long antibiotics, she had a group of
mice that only got eight weeks of antibiotics or only four weeks. All three groups of mice
receiving antibiotics were bigger in total, lean and fat mass. So four weeks was enough
for the full effect that didn’t need to be on it for their whole life. And in other studies,
Laurie showed that first four weeks of life was much more important, the effect is much
stronger than if they got it later. So now, what about the microbiome? In this form of
research, microbiome research our holy grail is a fecal transplant, or some kind of transplant
where we can test what happens if we give the microbes to a germ-free animal, an animal
that doesn’t have its own microbes. So Laurie did an experiment that we called “TranSTAT”.
She asked, “Is the growth phenotype, is the characteristic of enhanced growth transferable
by the mircobiota alone?” Forgetting about the antibiotic. So, for this experiment, her
donor mice were mice that either received the antibiotic or not. She picked mice from
the middle of the group, so not to bias things. She sacrificed those mice and preserved the
microbes in their secum in a special transport medium that preserved the viability of the
organisms and then she gave them food by an oral gavage, or she administered it into the
stomach of recipient mice and these were germ-free mice, here they are, in their bubble, received
from Taconic Farms, they were Webster mice, they were three weeks old, and now these mice
are conventional. And they were housed and kept for the next five weeks. So the mice
that received the microbiota from the STAT mice were increased in total mass, no change
in lean, but an increase in fat. So from this study, we have evidence that the microbiota
alone are sufficient to transfer the characteristic. Now I want to finish the experiments just
by telling you one other line of work, and that has to do with diabetes. Type 1 Diabetes.
Some of you know, and that’s juvenile diabetes, some of you know that this disease, which
is a terrible disease, which forces children to start taking insulin and is associated
with long-term disease, there’s no cure for juvenile diabetes, and people on average do
not have a full lifespan. Type 1 Diabetes is increasing around the world. On average,
in developed countries, it’s doubling about every twenty years. And so Alexandra Levanos,
another student in the lab, decided that she was going to see if there could be any relationship
between this and antibiotic use. So she worked in another model of mice, a model called the
“NOD Mouse” and these are mice that spontaneously develop something that looks just like Type
1 Diabetes. And Allie’s hypothesis was that early-in-life antibiotics would change the
gut microbiome and lead to Type 1 Diabetes accelerated in NOD mice by affecting the expression
of genes in the small intestine and by changing the T cells in the intestine and to make a
long story short, Allie has shown all of these and recently she successfully defended her
Ph.D, so she’s now Dr. Levanos and unfortunately, for me, she’s graduated, which is good news
for her, but we hate to lose her. So, her work has shown that even in this system, not
just obesity, but Type 1 Diabetes and we think it’s because of the changes in the T cell
have accelerated. So, I’ve been telling you about studies in mice, what about humans?
So there have been a number of epidemiologic studies that have looked at the relationship
of antibiotics in other early-life purturbations and now I’m going to tell you about two studies
that we did with colleagues in England. In the early 1990s, a big cohort of more than
ten thousand children was enrolled in England in Avon, it’s called the Avon Longitudinal
Study of Parents and Children, there were lots of data to obtain, they were followed
for the next fifteen years, and Leo Tresandi in the team asked, “Is there a relationship
between infant antibiotic exposures and early-life mass?” and the main result of the study showed
that children who were exposed to antibiotics during the first six months of life were more
likely to have markers of obesity. And now there have been four new studies that have
shown the same phenomena. In Colorado, here in New York, Noah Mueller has done work, and
also in Manatoba. And in another study from AL’s pack, we looked at the association of
Cesarean delivery with child adiposity age from six weeks to fifteen years led this study,
led by Jan Bunstein, the results were the same.The kids who were born by c-section were
more likely to have markers of increased when you control all other known factors. Also,
there have been four studies that have shown the same phenomenon. And so how do we put
it all together? So, when babies are born, they aren’t fully formed. And we know in the
uterus, babies have stem cells. And stem cells are cells that are pluripotential, they have
to decide how many times to go around and if they’re a mesenchymal stem cell they have
to decide, should they become fat, or bone, or muscle? So we’ve been paying a lot of attention
to diet and calories, but our view is that we need to start paying attention to the early-life
microbes and the cells they’re talking to because the hypothesis is that these co-mingle
populations are sending signals out that are affecting these developmental decisions. And
here’s a cartoon of the new days, in which kids have perturbed early-life microbes due
to antibiotics and other factors, they have different microbes, they have different cells
that are signaling, creating a different context. So what do we need to do? In the scientific
community, we need to do research about antibiotic consequences to find out, “Are antibiotics
as benign as we think, are they as free as we think?” And if we find there are risks,
we have to be better-educated. As scientists, we have to develop more specifications instead
of broad-spectrum antibiotics, we have to develop narrow-spectrum so we can treat infections
more precisely. That’s gonna mean better diagnostics. We need to re-mediate, we need after somebody
takes antibiotics we need to replace some of the organisms that are lost. But we don’t
fully know what they are yet. And that will come probiotics, not necessarily the ones
you’re buying today, but scientifically developed probiotics. We need to enhance some of the
organisms that are depleted, they’re present but depleted with prebiotics. And maybe we
have to do a reversal, we have to start giving back some of those vanished organisms. Replace
them, maybe with islets from the Amazon, or from Africa. And ultimately, we have to monitor.
Now, I wanna go back to the antibiotic use in the United States from that CDC study showing
this cumulative use I showed you before, because shortly after this was published in the U.S.,
a group of scientists from Sweden said, “Let’s look at our data.” And they found something
quite remarkable. By the age of three, the average Swedish child has about 1.4 courses
of antibiotics whereas our child has about 4. And when our child has 10 antibiotics by
ten, they only have 4. At every age, the Swedes are using only 40% of the antibiotics we’re
using. So it shows, and the Swedes, by the way, are at least as healthy as we are. So
it shows at least by this metric that about 60% of the antibiotics we’re using are at
least unnecessary. So there’s lots of room for improvement. So what’s the doctor of the
future going to be doing? I think that the doctors, especially for kids, in addition
to the usual well-baby check-up, they’re going to analyze their poop. And they’re going to
figure out which of the globally necessary microbes they need and which of the personal
microbes they need and they’re going to administer them. And then we’re going to asses, and we’re
going to get the kids on the best path for life-long health. It’s going to take some
science to get there. And I want to thank all my colleagues at NYU, including Isabel
Teitler, who is here tonight. So thank you very much for your attention.