Halving Global Cardiovascular Disease Mortality

Halving Global Cardiovascular Disease Mortality


>>Good morning. My name is Samira Asma, and I’m with the
Global Noncommunicable Disease Branch at CDC. And I’m very pleased to introduce Dr. Hamid
Jafari, Acting Director for the Division of Global Health Protection, to welcome us all. [ Applause ]>>Thank you, Samira. A warm welcome to all of you. This is a — it’s quite an honor for us to
welcome such a renowned group of experts from across the globe, including in particular
our keynote speaker, Sir Richard Peto. And this is really a symposium on halving
global cardiovascular disease mortality rates within a generation, quite an ambitious target. It is fitting for CDC to host this
symposium, as data, epidemiology, and science are the cornerstones of our agency
and are vital components of our decision making. In recent years, the epidemiological transition
has truly manifested as the global burden of disease shifts from infectious
diseases to noncommunicable diseases. More than ever before, people in the
prime years of their life are dying of cardiovascular diseases, heart
attacks, stroke, and cancer. And the evidence shows this
trend is only going to continue. [Inaudible] certainly health, economic, and political security of
countries around the world. Their increasing prevalence with high cost
implications undermine the ability of countries to provide quality healthcare
for all types of disease threats, including infectious diseases, of course. It is becoming more common to see people with
comorbidities, both noncommunicable diseases and chronic infectious diseases, requiring
a more integrated public health approach. No global health program is
complete without a focus on NCDs. We do need strong health systems that
are prepared to respond to any threat, whether it be NCDs, infectious disease
outbreaks, or natural disasters. A key component of CDC’s global entity strategy
is advancing innovative solutions based on evidence. The Oxford studies we’ll hear
about today uncover unique details about the world’s major causes
of death and disease. We are very supportive at CDC of this
work because this information is crucial to our own efforts, helping us to deploy
targeted public health interventions. Now I would like to introduce CDC’s
director, who has championed CDC’s involvement in addressing global NCDs, including
in particular cardiovascular diseases. Please welcome Dr. Tom Friedan. [ Applause ]>>Thank you, Hamid, and thanks
to all of you for being here. This is a really important topic, and
it’s one that sometimes is difficult to get attention to because
it’s not so apparent. But the world is well into
the epidemiologic shift. The disease burden is moving from
infectious to noninfectious diseases. NCDs now kill more people
than infectious diseases. Two-thirds of deaths worldwide, 80% of those
occurring in low and middle income countries, the rates of noncommunicable diseases are now
higher in lower and middle income countries than in high income countries, and this
is a major threat to not only health, but also social and economic development. We know that many NCD deaths occur young, and
this results in enormous healthcare costs, lost productivity, but most
importantly, human tragedy. Among NCD deaths, CVD is
the world’s biggest killer. Hypertension is the only thing that kills
more people than smoking in the world. About 18 million people a year die from
heart attack, stroke, and heart failure and will kill 25 million
globally per year by 2030. So things are bad and getting worse. We’re getting older as a world,
and we need to be able to prevent and treat noncommunicable disease
much more effectively than we have. I think for that to happen, some of the prevention strategies require a
political commitment and a willingness to expend not just financial
capital but political capital. For the treatment component, we need healthcare
systems to function effectively year round for chronic diseases for which there is no cure. This is very different from what we’ve
done so far with immunization programs or oral rehydration treatment for the infectious
diseases or even in treatment programs such as those for tuberculosis
which are time limited. High blood pressure kills about 10 million
people a year, almost as many people as all infectious diseases combined. About one in seven people around the world
with high blood pressure have it under control, and I think one thing we have to recognize is
that even if we do beyond on our wildest dreams in reducing sodium consumption, increasing
physical activity, preventing the continued rise of obesity, there’s still going to be
about a billion people who need treatment for hypertension and will
benefit from it substantially. We’ve known this for a while, but we’ve been
very slow to act, and that’s unacceptable. We need models that work for the mass
simple treatment of large numbers of people, and I don’t have a strong perspective on what
the model should be, only that it should work. In public health, we use data to drive
our actions, data to monitor and evaluate, data to have surveillance
and systematic research, data even to evaluate whether our
communications strategies are effective. This is a crucial alignment with all
of our approaches in global health, which is moving from kind of a general
approach of this is what we think will work, this is a general standard, to
what I call the 2.0 version of many of our international health programs where
it’s more of a surveillance and response. It’s more of a program implementation
based on looking at data and adjusting our programs based on that data. This has been critical to success
in public health for decades. We know, for example, that it took a
long time for the public health world to acknowledge the link between
tobacco and cancer. It would have been even longer if it hadn’t
been for the work of Sir Richard Doll. At the end of World War II, Doll and others
hypothesized that cancer and other — could have been caused by tobacco smoke
and other things and began a collaborative that continued for five decades, a collaborative
that Sir Richard Peto, who we have the honor of having with us here today, continued. For the past three decades, the University
of Oxford Clinical Trial Service Unit and the Epidemiological Studies
Unit has demonstrated that large population based
observational studies are feasible not only in high income countries, but also
in low and middle income countries. And I’d like to comment that one
issue that I think all of us have to really address is the importance
not just of looking at data, but of assessing every different
type of data feed for its validity, for its external validity, for its soundness. There has been I think an overemphasis
on randomized controlled trials. There’s no doubt that randomized
controlled trials are one of the great accomplishments
of the past century. There’s no doubt that to compare two
treatments using placebo controlled, double blinded trials is essential. But there’s also no doubt that there are
severe limitations to what RCTs can teach us. And other data sources such as population-based
observational studies and many other sources of epidemiologic data are crucially important. Now the work that the unit has done
there is a landmark collaboration. It’s one of a kind globally and is influencing
public health policy around the world. It addresses infectious diseases, including
AIDS, TB, and malaria, but also NCDs, cardiovascular disease, diabetes,
cancer, as well as the risk factors, such as tobacco, alcohol, and obesity. These are also able to project
progress or the lack of it when optimal policy action
is taken or is missing. For example, China has accelerated work on
tobacco control, and Russia has begun some work on alcohol control in significant part because of the data shared by
these very important studies. They’re drivers for action, but continuous
and rigorous evaluation of all forms of data is necessary and building national
capacity so that data is owned by country has to be a goal that all of us share. This morning, we’ll hear about large
population-based observational studies on risk factors for CVD in China, Cuba,
India, Mexico, Russia, and the US, and we’ll focus on how data from these studies
impacted the policy progress and debate, not only in these individual
countries, but globally as well. I thank the country representatives
for participating and sharing your findings and results. Your contributions are paving the way forward. There’s still a tremendous
amount that we need to do. And we’re honored to host this symposium
on having global cardiovascular disease. I would like to welcome now Sarah
Lewington of Oxford University, who will speak about epidemiology and trials. Sarah is Associate Professor
of Statistical Epidemiology and a Scientific Director at Oxford. Please welcome Dr. Lewington, and join me. Thank you very much. [ Applause ]>>Well, thank you very much, and — for
inviting us, and it’s an absolute pleasure to be here, and thanks very much to Samira
and her team for making this symposium happen. So I’m going to talk for the next 15 to
20 minutes on epidemiology and trials and the meta-analyses of those trials. And this where the — I came to Oxford to do
in 1995 as a young, enthusiastic PhD student and have been working with Richard ever since. And I don’t think I need to tell this
audience that the major modifiable causes of vascular mortality, tobacco, blood pressure,
blood lipids, and adiposity and diabetes. And you’ll hear more about tobacco from Mike
Toon [assumed spelling] later on in the morning, and you’ll hear much more
about diabetes from Mexico. So I’m going to concentrate in my talk on the
blood pressure, blood lipids, and adiposity. And I’m going to talk about work that we’ve
done in the prospective studies collaboration. Now this is a collaboration of investigators
from 60 perspective cohort studies. There are a million adults in those studies, and
about 90% of those are from Europe and the US. So this is mainly results from high
income countries that I’ll set the scene. In those studies, there were 60,000
vascular deaths during an average of 13 years of follow up. And what I’d like to show is how it was with
careful attention to the statistical details that we got the answers right, and it’s those
statistical details that we’ve carried forward into our work in the low
and middle income countries. I arrived — I say I arrived in Oxford in 1995. I submitted my PhD thesis in 1999. So it took a long time to get the data straight. It took a long time to get
the analyses straight. But even then, we had a hunch that
we hadn’t got it quite right in 1999. And I don’t know if Richard will remember
this, but we had a eureka moment in a hotel bar in Beijing in 1999 where we realized what
we had to do to get these analyses correct. And this is the result from those studies. This is the effect of usual systolic
blood pressure on CHD mortality. And you see it’s published in The Lancet in
2002, but importantly what you see here is that at least up 69 — so for
premature death, about 20 millimeters of mercury lower systolic blood
pressure, approximately halves risk. And that’s throughout the range with
no evidence of a level below or above which lower blood pressure is no
longer associated with lower risk. And I think we know that now, but it’s — it — we must remember that this wasn’t
the received wisdom at the time. So what were the — what was that attention
that we gave to the statistical details? Importantly, we avoided reverse causality
whereby the blood pressure itself was a cause of the disease rather than the disease
being a cause of the blood pressure. And so, we did this by excluding prior disease where the disease itself was maybe
affecting the blood pressure. So where people had reported that they knew that
they had disease, but also we excluded deaths in the first few years of follow up
so that those people who had disease but didn’t realize it, we
also excluded those deaths. We also assessed the effects of the long-term
average, the so-called usual blood pressure and cholesterol over the period of follow up. So we weren’t relating risk just to
the baseline blood pressure measured on that one day 13 years ago, but
to some estimate of the long-term, the average blood pressure
over the period of follow up. And also, we avoided chance findings. A lot of the findings that were reported at
that time were based on very small numbers, particularly small numbers of events. And it’s just crucial that we get big numbers
if we’re going to avoid just chance findings. I’m going to illustrate this using a
random selection of participants chosen from the prospective studies collaboration. So again, we’re looking at CHD mortality
versus systolic blood pressure by age, and I’m going to look at it in 5,000,
50,000 and 500,000, in other words, looking at it in the size of a Framingham
Heart Study, the British Doctors’ Study, and finally the newer prospective
studies of 500,000, such as China Kadoorie Biobank, UK Biobank. So if we take a Framingham Heart Study
sized study and look at usual blood pressure against coronary heart disease mortality, you
see that the findings are all over the place. There’s really no evidence. There are some lines on here
showing what our best estimate of the strength of the relationship is. But you can see how in the days of
studies of 5,000 people at ages 80 to 89, we’ve observed an inverse association,
if you can see any association at all. It looks inverse. In other words, higher blood pressure
appears to be associated with lower risk. And we know this isn’t true. Now if we go to a study of 50,000,
so a British Doctors sized study, the picture’s beginning to come into focus. It’s becoming clearer that there’s a positive
association with increasing blood pressure. But there are still uncertainties there. If you look at the 40 to 49, does
that suggest that maybe again at lower blood pressure then the
risk begins to increase again? So we were still — a lot of
uncertainty about the true shape and strength of this relationship. And it’s only really when we get
to a study of the size of 500,000 that the pictures comes properly into focus. And we see that the strong positive associations
at every age group with no evidence above or below which there’s not a lower
risk with lower blood pressure. So other statistical details, now this is a
bit technical, but it’s really very important. It’s something called the
regression dilution bias, and it’s the fact that purely random
fluctuations in your risk factor and your blood pressure — measurement of
blood pressure and cholesterol don’t lead to random fluctuations in our
estimate of the risk of association. And they actually lead to a systematic
underestimation of that relationship. But due allowance for this bias using resurveys
of just a sample of the participants show that the true relationship is much steeper. So here’s why resurveys are so
important in observational studies. Again, it’s still CHD mortality versus usual
systolic blood pressure, but this time plotted against the baseline systolic blood pressure. So we divide people up into ten groups
according to their blood pressure measurement and plot risk against the mean of the
blood pressure in each of those ten groups. And you see here again a positive relationship,
but there’s maybe some sense that above about 180, although there’s still a positive
relationship, it may be somewhat weaker. But what we can say from here is maybe up
to about 180 there’s about a 25% lower risk with every 20 millimeters of mercury
lower baseline systolic blood pressure. But what happens in the resurvey
of those people? So we have 300,000 people who’ve been resurveyed
between two to 15 years after baseline, most of them within the first few years,
so on average three years after baseline. So now those blood pressure groups
have gone up onto the vertical axis. But you can see that between the
top and the bottom group there’s about 100 millimeters of mercury difference. And that was associated with about
a fourfold difference in risk. So 100 millimeters of mercury, fourfold
difference in risk, but just three years later, what we see is that the mean of the
top group has fallen considerably. The mean of the bottom group has risen. And the reason for this is that when you divide
people by a single measurement, at the top, you get people who not only have
genuinely high blood pressure, but you also get a disproportionate
number of people who were just on a random high on that day. For whatever reason, their blood pressure
was particularly high on that day. When you remeasure those people, the
people who were on a random high, that blood pressure’s going to be lower. And so, the mean of the group
falls, and similarly at the bottom. So the people that are on a random low for
whatever reason when you measured them, and so you’re overestimating the
difference in the exposure at baseline. And so, just three years later, actually there
isn’t 100 millimeters of mercury difference in the exposure between the
top and the bottom groups. There’s only 65 millimeters
of mercury difference. So that’s still associated
with the fourfold risk. It’s not 100 millimeters of mercury, but 65
that’s associated with that fourfold risk. And so, we’re underestimating the strength
of the relationship by about a third. And now if we plot risk not against the mean
at baseline, but against the mean at survey, we see that the true relationship
is really very much steeper. And so, it was this attention to
detail that took another few years. And you see that the paper
was finally published in 2002. So it took seven years to
get these data together, to get the data to straight,
to get the analyses correct. But I think it was worth it. This paper has now been cited
more than 7,000 times, and I don’t think that would have
been the case if we published it — if we’d thrown the data together
and thrown an analysis out quickly. It’s also allowed us to get the analyses
right for the next ten to 15 years after that. So just a reminder, observationally
we see that 20 millimeters of mercury approximately halves
the risk, but 20 millimeters of mercury is a big drop in blood pressure. And the trials really only show about
a ten millimeters of mercury fall — change in blood pressure if you’re lucky and if
you work very hard, and that would be associated with about a 30% difference in risk. So work done by Stephen MacMahon’s group
at the Georgia Institute for Global Health, and this work was actually done by an
MSc student of ours in the Department of Population Health, looked at all
the blood pressure lowering trials, the randomized controlled trials of blood
pressure lowering agents, 300,000 randomized, 25,000 vascular — cardiovascular events. They found importantly that the
risk reduction was proportional to the blood pressure reduction. So they were seeing the same strength
of relationship for a given change in blood pressure and that the ten millimeters of mercury lower systolic blood pressure reduced
the cardiovascular disease events by about 20%, so not quite the 30% that we observed with
naturally occurring lower blood pressure, but in lowering blood pressure
within just five years, you achieve most of the epidemiologically
expected reduction in risk. But as I’ve said, importantly,
most of these trials achieve less than ten millimeters of mercury
lower blood pressure. One or two of them with really aggressive
blood pressure lowering achieved 15 to 20, a few of them just over ten, but most
of these trials were really looking at blood pressure reductions of
five to six millimeters of mercury. So moving onto blood cholesterol, and
here’s data on usual total cholesterol within the prospective studies
collaboration, so the same style of figure. You’ve got risk on the vertical
axis against the usual cholesterol. And you see here again for cholesterol
the same proportional reduction in risk for a given change in cholesterol and
about a two millimole per liter fall in usual cholesterol, again,
approximately halves risk. I’m looking at trials. So how do the — what’s the trial
evidence on lowering cholesterol? Well, as you all know, we have very
effective cholesterol lowering, LDL lowering treatments with the statins. And a group in Oxford, the Cholesterol
Treatment Trialists, led by Colin Baigent, analyses by our colleague Jonathan
Emberson, they’ve analyzed data on 100,000 patients randomized to — in statin
trials, 25,000 cardiovascular disease events, and again see that the risk reduction
was proportional to the LDL reduction. Each millimole per liter lower
LDL reduced CVD risk by about 25%. That’s after the first year of treatment, so
slightly lower in the first year of treatment, but after the first year of
treatment within about five years, each millimole per liter lower LDL reduced
the cardiovascular disease risk by about 25%. And I think importantly for LDL cholesterol,
these reductions of two millimoles per liter, 77 milligrams per deciliter, are possible
with effective — cheap, effective drugs. Moving on to overweight and obesity, and I’m
now going to look at data recently published from the Global BMI Collaboration. So this is work that was coordinated by John
Danesh in Cambridge and Frank Hu in Boston. And Richard and I worked
very closely with them again on getting the details of the analysis right. So the analysis based on the principles
from the prospective studies collaboration, but we worked very hard with them to get the
story straight, again really important for BMI, even more important for BMI
that we avoid reverse causality. And so, the analyses were restricted to
never smokers who reported no prior disease and excluded the first five years of follow up. And this left data on 55,000 coronary heart
disease deaths among 4 million adults. And what we see here is that for people who are
obese, ten kilogram per meter squared change in BMI again approximately halves the risk,
but ten kilogram per meter squared is huge. I did a rough calculation before I came. I think it would be me reducing from
being 17 stone to the weight I am now. So it’s a huge change in weight, but that’s
what the data showed, that at least down to 25, so going from obese to not obese,
you get about a halving in risk. So to conclude, I think halving global
premature mortality is achievable, but I think it will require lifestyles
changes, particularly in smoking, and importantly wide use of cost effective
generic drugs, particularly the statins and the blood pressure lowering drugs. Thank you. [ Applause ] I’m happy to take questions. No? In which case, I’m pleased
to invite Richard. I cheated. I was supposed to be talking about primary and secondary prevention,
and I thought wait a minute. I want to talk about overall mortality as well. And so, in the night I stayed up into the
small hours and set the alarm for five o’clock and tried to squeeze some extra slides in here. So I’ll try and get through them. There was a guy in Oxford who was a
famous preacher, and after he died, they found that he had got all these
sermons that he’d given all written. And in the margin of one of the sermons,
it say, you know, speak quickly here. Uncertain of facts. [ Laughter ] So you’ll find there’ll be some
cases where I’ll speak quickly. Okay. I — I’m supposed to be
talking about halving CVD mortality, because that’s what this symposium’s about. But what I really want to talk
about is halving premature death. You know, CVD, you know,
okay, it’s very important. Yep, it’s the most important. But it — you know, I’d like
to halve overall mortality. So I’ll talk a little bit about
that while nobody’s looking. This is Richard Doll’s slogan. Death in old age is inevitable,
but death before old age is not. So everybody who reaches old age
is going to die of old age, but — die in old age, but the deaths before old
age are largely from avoidable causes, as are quite a lot of the deaths in old age and
a lot of the disability and a lot of the misery. I’m not saying that old people don’t matter. I’m 73 myself. Actually, I quite like talking
about, you know, the — you know, how do we actually avoid death before
70 because it means that I’m completely safe. You know, it’s all you lot that
are going to die before 70. So — but if you’re going to do that, you’ve got
to look at the deaths in middle age are from — mostly from noncommunicable disease,
especially if you go to say 50 to 69. It’s nearly all noncommunicable disease,
apart from pneumonia and TB and some accidents and injuries, whereas the deaths before middle
age, before age 35, are mostly accidents, injuries, and infection, rather
than noncommunicable disease. And you’ve got to think separately
about these two age ranges. You got to say separately each
can be halved in each age range. And the first question is how many
deaths have we got at the moment? I’ll give crude numbers of deaths rather than
probabilities of death because they’re a bit — they seem to be a bit more newsworthy. And so, at the moment we’ve
got 53 million deaths a year, and 30 million of them are before 70, 12 million
before middle age, 18 million in middle age. I actually looked at the definition
of middle age in the dictionary, and it’s — you can define it how you like. The dictionary, it was quite nicely written. It said middle age is defined as that
period between youth and old age. [ Laughter ] No, no. I haven’t finished yet. It’s that period between youth and old age
variously reckoned to suit the reckoner, which is lovely dictionary writing. [ Laughter ] Variously reckoned, so I wonder how it’s — that old enough that you’re starting to get
some noncommunicable disease and not so old that people think you’re better dead anyway
because if you try and go to the global health where people are concerned still with, you
know, a lot of the problems of the poor, and you start arguing for the importance
of doing things for people in their 70s, you’re going to get a glazed look coming across
the person’s eyes that you’re talking to. Okay. So these are the numbers, 12 million,
18 million, 23 million, so total 53 million. And then how many of these
are cardiovascular disease? Well, it’s — it’s only — of
the 12 million before age 35, few would be from cardiovascular
disease, very few. Of those 35 to 69, about 5 million would be. So in terms of death before age 70,
it’s 5 million out of 30 million. It’s one in six. Then you get into old age when
cardiovascular disease does predominate. And that’s when you get your next 12 million. And so, in total, yeah, we’ve got 17 million. These are 2010 numbers, 17 million deaths
a year from cardiovascular disease. It’s probably 18 million a year now, out of
a world total of about 53 million deaths. So it’s a third of all the deaths. But it’s only a sixth of the deaths
before age 70, and that seems to me to be a more appropriate metric if we’re
trying to think about how important is this in comparison with various other things. Obviously, it varies from population
to population, but for the world, the number of deaths before 70 seems
to be quite a useful thing for talking about the relative importance
of different problems. I mean formally you can do it by doing
complicated calculations about years of life lost or dallies or
quality adjusted life years. But I think there’s still something
to be said for simple millions. But if you’re doing simple millions, then
I don’t feel that one can take, you know, deaths after age 80 and count them in the
same sort of breath as malaria deaths. So I find this cut at age 70 works quite well in giving me statistics that
I feel comfortable with. Okay. So what — what’s happening
to these trends? Well, the first thing is if we take crude
numbers, that number 30 million is not going to change over the next ten, 15 years. You know, round about 2030 it’ll still
be about 30 million deaths a year because death rates are going down. Communicable disease death
rates are going down worldwide, but so actually are death rates
from noncommunicable disease. They’re going down, not up,
taking the world as whole. And in many populations,
they’re going down, not up. So the death rates from people
of a given age are going down, but the actual population is going up,
you know, because people didn’t die. So look at me. I’m about 73 now. When I was born in the 1940s, well,
there were a lot more people born in the 1990s than in the 1940s. So as the world moves on another
half century, the number of people in middle age is going to
go up just automatically. So the number of deaths will go up. And this combination of decreasing
death rates and increasing population, it’s — it balances out at the moment. And so, that number 30 million
will stay constant. So you learn that number by heart because
it’ll still be true about 15 years from now. There’ll probably be fewer of them
before 35, more of them after 35. But it’s — it is true, as we said earlier,
that cardiovascular disease is accounting for a bigger proportion, an
increasing proportion of all deaths, including an increasing proportion
of the deaths in middle age. But this is really because the cardiovascular
disease, the noncommunicable disease, is going down slowly, and other
causes are going down even faster. So it’s not that among people of a given
age that either cardiovascular death rates or noncommunicable disease
death rates are going up. They’re actually drifting down. I’ll illustrate this. I’ll apologize to those who in my talk
here last time because I’m now going to show three slides that I’ve shown before. I want to show what happened
over the last hundred years in England, strictly actually seen in Wales. So I want to show you the patterns of
mortality in England in 1910, 1960, and 2010. So we’ll start off 100 years
ago when my mother was born. Actually, she was born in 1909. So this is the survival pattern,
and you can see one in six of the kids are dying in the
first five years of life. Two-thirds die before age 70. Then as we go from 1910 to 1960,
we get clean water, vaccinations, better nutrition, and medicines that work. And the death rate goes down. And so, instead of 16% dying in
early childhood, it’s down to 4%. So you go to the 1960 births, and it’s 4%. And if we go to 2010 births, it’s about 0.4%. So death in childhood just disappeared during
the 20th century, mostly in the first half of the 20th century in rich
countries like Britain. But then we’re left — I’ve chosen
males, and I’ve chosen Britain in particular for a particular reason. It’s because in 1960, we have the worst tobacco
death rates in the world, and that ten-year gain from 1960 mortality patterns to 2010
mortality patterns, more than half is — of that is because so many people
either stopped smoking or never started. We’re now getting further along, so it’s
actually the never started is kicking in. And the next picture I’ll show looks
like a little cigarette, but it isn’t. So it’s actually a bar diagram. Tom Friedan, if you look at your emails
while I’m talking, I shall kill you. [ Laughter ] Turn that off, or turn it over. [ Laughter ] Okay. Anyway, the next slide’s
especially for you. So this is roughly speaking the
probability of dying in middle age. So what’s the probability a 35-year-old’s
going to die before they’re 70? And back in 1960 and 1970 death
rates, it was about 42, 41%. And now it’s gone down at 2010 death rates. It’s only 19%. And the shaded bit is the tobacco deaths. And even this is a bit of an
undressed in the tobacco deaths. It’s seeming that not all of the
vascular mortality associated with tobacco is caused by tobacco. So we’re going down from somewhere like
about 19% killed by tobacco down to about 4%. And if we take all of the vascular
disease associated with smoking, then we’d be going down from 23% down to 4%. So we’ve got the main thing that’s driven
UK death rates down is the great decrease in tobacco deaths due to the great
campaigners and the great epidemiologists. Okay. Well, we think we’re good,
but the Americans are even better. So here’s US cigarettes per adult, and I expect
there’s people around you who are familiar with nearly all these events
over the last half century. But anyway, the — I like this graph with the
red lines on it saying ten a day, three a day. And as you can see, back in the 1960s, this is
US cigarettes per adult over the last century, actually more than a century, 1900 to 2012. And you were running at about
12 a day back in the 1960s, and now you’re down to about three a day. So you’ve had a decrease of about
three-quarters, while we’ve had a decrease of only I’m afraid about two-thirds. So we’ve gone down from about
9.6 a day down to three a day. Anyway, it’s pretty good in both
countries, but still despite that decrease, because those who smoke have now
smoked all their adult lives, smoking is still despite this decrease the
most important cause of premature death in the United States and in the United Kingdom. So you wouldn’t believe it how bad smoking
is, even though we’re down to three a day. It’s still a quarter of all
US deaths at ages 35 to 69. The next graph is the estimates which Mike
Toon and I and various others are working on. The left hand is absolute numbers of deaths
from smoking, top male, bottom female, in the United States over the last 60 years. And the right hand side is saying what fraction
of all of the US deaths at 35 to 69 was caused by smoking, again top male, bottom female. But over on the right, you can see the
male and the female have come together. It’s now a quarter of all male deaths, a
quarter of all female deaths, and it’s falling, but it’s still an even bigger
number than the number of deaths due to obesity, adiposity, and so on. So still actually smoking is the number
one external cause in the United States. And if we look at cancer, the same is true. So it’s a quarter of all deaths in middle age. When it comes to cancer, smoking is still in
the United States 33% of all cancer deaths in middle age, 33% of all
cancer deaths at any age. It’s one-third of all cancer deaths. Yes, it’s falling. But it’s still a third, bigger than all the
other causes of cancer we know put together. So the next picture again is estimates that
Mike Toon and various others are putting out, Adam Lopez, showing what US cancer
death rates should have been like over the last 60 years
and what actually went wrong. So the approximately horizontal line in the
males on the left that finishes at 5% goes down from about, you know, 6, 7% down to 5%. That’s what US male cancer deaths should
have been like if nobody had ever smoked. And the big hump is what got
added to that by smoking. So back in the 1980s and
still in the early 1990s, smoking was half of all US male
cancer deaths in middle age. In females over on the right, the top line
is what female cancer death rates should have looked like if no woman had ever smoked. You got decreasing cervical cancer
death rates, stomach cancer death rates, and then in the last 25 years, you’ve got decreasing breast cancer
death rates due to better treatment. Breast cancer death rates are now down
to half of what they were in middle age because treatments work and they get given. But then down at the bottom you’ve
got what smoking has added to that, and had there not been all these efforts
against that in the United States, that female attributed to smoking would’ve
by now been up to the background death rate. You know, you’d have had half of all the female
cancer deaths in middle would’ve been caused by smoking, but people actually took it
seriously, including a number of people in this audience, and the
results have been very good. Okay. A little bit more on
tobacco, then I’m going to get back to what I’m supposed to talk about. But I am a tobacco addict, and
it just seemed too good a chance. You know, I’m never going to be made
a hero again any time in my life. So the — what can you do about tobacco? Well, obviously, you want
awareness and campaigns and so on, but the key thing internationally is price. Well, and in America, in
Britain, in Europe, it’s price. And the people who have done price best
have been the French and the South Africans. Over the period of 1990 to 2005, they
tripled the real price of cigarettes in South Africa and France, tripled the price. What happened? Consumption went down by half, even
after you allow for a little bit of extra smuggling in from
neighboring countries. And what happened to the government? Did the government get rich or poor? The government doubled the amount
of real money they were making. So it was triple the price, half
consumption, and double the tax rate. Triple, half, double. Now what if China did that? We’ve got a million tobacco deaths
in China, now heading for 2 million. You know, triple the price. The government would make more money. They’d get rich. So real cigarette price triple, smoking
half, government tax yield doubled. Well, you know, that’s pretty good. Two countries did it. We’re not going to get changes as good as that
worldwide, but could we get the price doubled? Could we get the real price doubled? Well, yeah, we could. Governments want money. They want money. They want it. You know, well, take it. So, you know, we’ve got a WHO target that
we’ve got to have smoking down by a third. You know, it varies whether
it’s by 2025 or 2030. I think they’ve pushed it back to 2030 now. They want cigarette consumption
down by a third worldwide. How are you going to do it? Well, we could try telling people
to stop and see if that works. And I’ll tell you it won’t. It’ll help, but it won’t. And at the moment, the government — the
governments of the world make $300 billion from actually making cigarettes
and taxing cigarettes. WHO target is a third less
smoking by 2025 or 2030. If the real price stays constant
and smoking goes down by a third, then the governments are going to lose
about $100 billion a year in real money. Now that’s not big money for
governments, but it’s not small. I mean it’s you know, it — you know, it — you
know, sure, it’s not up in the trillions level. But it’s, you know, $100 billion a year. I mean some people would quite like that. [Cough]. But if they go the other way
and if they take the opposite strategy, if they double the real price, that alone
will make smoking go down by a third, and the governments will gain $100 billion. Now I find that plausible. I find the idea of them losing
$100 billion a year implausible. Okay. Now I’ve got to speak up a bit now. Trends in mortality, okay, what’s happening
to world mortality in early childhood, early adult life, later adult
— late — in later middle age? And here is the answer for the world as a whole. The bottom in each of these graphs
starts at 1970 and finishes at 2010. The blue one is under five mortality, down from
17% down to 5% — it’s from 14% down to 5%. So over the last 40 years, under five
mortality has gone down from 14% to 5%. It’s gone down by more than half. The red one is mortality ages five to 49,
and that got hit by the AIDS epidemic. Even with the AIDS epidemic causing that
slowdown, it’s gone down from 17% down to 10%. Now the top one is mortality
in later middle age. This is predominantly noncommunicable
disease mortality, although it does include some TB and pneumonia. But there’s — you know, a very
late [inaudible] or anything like that by the time you get beyond 50. And the probability that a 50-year-old will
die before 70 has gone down from 36% to 24%. It’s gone down by a third. We’re doing better. We’re improving in every age range. And if it hadn’t been for the HIV epidemic,
we’d have been improving even more. You can say, “Well, okay,
that’s the world as a whole. What about the rich and poor?” So here we’ve got the poor on the
left and the rich on the right. In every, every — in every region of the
world, every World Bank region of the world, the poorest one-sixth, the richest
one-sixth, or the middle two-thirds, you’ve got decreasing death rates, except
for the sort of horrendous effects of HIV in Sub-Saharan African, which you can see in the low income countries’
mortality at 50 to — five to 50. But the blue, in the biggest
decrease, we want equity. Well, okay, who’s had the best
decrease in child mortality? Kids die on an absolute scale,
not on a relative scale. And it doesn’t matter if
you take 0.1% down to 0.01%. What matters is you take 20% down to 10%. And in the poorest kids, look, they’ve
gone down from 23% down to 10%. The — and if you look at the lower
middle income, the absolute gain is great. I mean it’s the rich who have — aren’t
getting a big decrease in child mortality because you can’t go down below zero. But when you look up at the
mortality at say 50 to 69, it’s similar everywhere,
everywhere, nearly everywhere. There are a few exceptions, which
I’ll show you, particularly Russia. It’s getting better. So if we take the 25 richest countries, then
here’s the child mortality, and I’ve put arrows against the countries who are
going to be discussed today. And you can see progress pretty well
everywhere, except where you’ve got war or the HIV epidemic in Africa, little kinks. But overall, things are getting better. In the US and in Russia,
China and Mexico, India — see China and Mexico both over the last 40
years largely got rid of under five mortality. Five to 14, you’ve got some hideous exceptions. Iran, Saddam Hussein invaded Iran
and killed a million Iranians. That did a bit of a spike in the
death rate, but they were all males. If you look just at females,
you don’t see that spike. Over in Africa, you’ve got places where
chronic war and things don’t get better. You’ve got South Africa, the
world’s worst HIV epidemic. You’ve got Russia where they
drink vodka, and that — you know, when the Soviet Union collapsed, they
turned to vodka, and the death rates went up. But overall, things are getting better. And when we look at later middle
age, 35 to 69, except in Russia, you can see that overall there is improvement,
not enough improvement perhaps in some places. You know, the — there’s Egypt somewhere here. You’re not seeing very much. That’s perhaps because they’re putting on
too much weight, and it probably is actually. But overall, China, US, Mexico,
India, and the world, improvements. So we’re living in a time where
things are getting better. Actually, one more thing on trends,
and I’ll get to the main subject. This is the probability of death. If you actually took the world 2010 death rates, apply it to the world 2030 population,
how many deaths would we get? Okay. This is the number. You’d get about 40 million deaths before age 70. That’d be about 20 million before
age 50 and 20 million at 50 to 70. But we aren’t going to have 2010 death rates. We’ll have falling death rates. What’s been happening to death rates
in the first decade of the century? Well, that’s what’s been happening. Those red things say what was the change in
mortality in the first decade of the century? Perinatal, maternal, communicable, it’s
gone down by a third in the under 50. It’s down by a quarter in the 50 to 69. Noncommunicable, down 15% per
decade throughout the age range, injuries down, you know, 10, 15% per decade. So total, down a quarter in the
under 50, down 15% 50 to 69. So things are getting better. We won’t have 40 million. We’ll have 30 million. We’re in a time where things are getting better. So unless there’s something dreadful worldwide,
tobacco, HIV, alcohol, adiposity, and war, being too fat and war, these
are the only big cause of death that have fluctuated substantially
since 1990 in some large populations after you allow for population growth. So when we come back to the
CVD, which is what I’m supposed to be talking about, here we’ve got the numbers. What can we do about the 5 million CVD
deaths in middle age or about CVD in old age? Can we halve CVD mortality in middle age? Well, in Britain, we’ve halved
it twice in the last 30 years. CVD death rates in 2010 were
one-quarter of CVD death rates in 1980. So here’s UK cardiovascular
mortality death rates. The top line is males. It’s not fair. Males always do worse than women. And the bottom line is women. And the red cross in the middle is the
average of male and female in 1980. Sixteen percent would have died
of CVD at 1980 death rates. At 2010 death rates, 4% would have died. If I did the same thing for the US,
you’d have started off higher up than us. You’d have started off right at
the top of this graph back in 1970. And instead of going down to
4%, you’d have gone down to 6%. But still, there’s huge decreases
are in progress. So can that happen elsewhere? Yes. So what can modify it? Well, these are the big four,
tobacco, blood pressure, blood lipids, adiposity, diabetes, same as Sarah showed. How could — where have the
twofold differences come from? Well, light smoking, smoke five or ten a day. That doubles your risk. Twenty millimeters systolic,
that’ll double your risk. Two millimeters LDL cholesterol,
that’ll double your risk. Ten units of BMI, that’ll double your risk. Okay. So these — well, okay, I’ll
show one slide on tobacco there, one slide on blood pressure there. Sarah linked it. One slide on blood lipids, they are certainly
— it just says that the red stuff’s true. And then what about adiposity? Well, we’ve got the global BMI
consortium, and here it is. This is the global BMI consortium. Red is female. Blue is male. And the two lines are 20 to 25. The normal range should not go beyond 20. The WHO is wrong on that. The normal range should stop at 20
or maybe even a little above 20. But anyway, in the range of 20 to 25, it’s flat, and then as soon as you get the
slightest bit above 25, the risks go up. Minus 25.5, or you get down about to 23. And actually, because males have higher death
rates than females, the absolute risk for males, the absolute risk of obesity
for males for a given level of obesity is three times as big as women. The relative risks are bigger,
and the absolute risks are bigger. But this claim — the claims that
there was, you know, overweight and grade 1 obesity are not
dangerous are mistaken claims, and they were due to exactly the
kinds of epidemiological mistakes that Sarah Lewington was describing. You know, there’s grade 1 obesity. The claim that that is not
dangerous is completely loopy. It was made. It’s been widely circulated all
over the US, and it’s total rubbish. The last two slides, secondary
prevention of vascular death, this I hope will lead into discussion. I think that I would stress the drugs
work best for those at high risk. One thing that puts you at high
risk is high blood pressure. Yep, if you’ve got high blood pressure,
definitely then go ahead and treat. But if we were going to try and get
something happening in China, in India, in anywhere where you’ve got doctors, and
actually in a lot of developed countries as well, high income countries,
then I would like to see for those with disease who’ve still got a good quality of
life, and they’ve got something serious to lose from having a disaster in the near future
because they’re not yet 99 years old, then we need to ensure affordable availability
and widespread use of generic statins, blood pressure lowering drugs,
aspirin, things like this. How do we actually make it happen? What social things do we — I mean Mexico is a
disaster, as the Mexicans are going to describe to you, because they’ve got, you know, all
sorts of healthcare, but the treatments that should be given aren’t being given. How do you actually change
things so that they are? This would be practicable in high and middle,
and you don’t have to have a screening program. So my last slide is this, that the aim should be
to treat high risk, not just high blood pressure or high cholesterol, although that
might be what determines high risk. But one of the key things is
have you actually got disease? You get greater absolute benefit than you
would generally do in primary prevention. Diabetes could very well be a thing
determining high risk in Mexico. You don’t need a screening
program to find those to treat. You don’t need to medicalize [assumed
spelling] apparently well individuals. I think we should start here,
get the drug supply sorted out. Why are generics in China ten times
as expensive as generics in India? And then get things actually used. And, yes, certainly it’s then going to extend
to high risk defined in various other ways, but I think would start with the emphasis
on secondary prevention among those who have got a reasonably good quality of life and they’re young enough
that it actually matters. You know, you’ve got a reasonable amount of good
quality life expectancy, not just a good quality of life now, but good quality
of life expectancy. Those are the ones where the protection
would be of greatest absolute value. So I’d like to emphasize that. And, Tom, I thought I’d put that up as
background for the discussion really. That’s meant to lead into
discussion, if that’s okay.>>Absolutely.>>Okay. Thank you very much indeed, and — [ Applause ]>>Well, as you’ve heard, Sir Richard is a
big thinker and not afraid to say anything. [ Laughter ]>>Your email [inaudible] year round. [ Laughter ]>>I was checking your facts, Sir Richard.>>Oh, okay. Sorry. [ Laughter ] Okay. No, that’s fair enough. Okay.>>Let’s start with the issue. You know, Farr said that death rate is a fact. Everything else is an inference, and I
think all of us like to focus on death. But one thing that might motivate more action
on NCDs is if we can document if it’s true that uncontrolled blood pressure
in middle age increases the risk of cognitive decline later in life. And there’s growing evidence –>>Oh, yes. It does.>>That it does that. So I wanted to first ask you about –>>Yes.>>That particular issue.>>Yeah. I think although actually the
prevalence of cognitive decline in later life, because vascular death rates are
going down, also the prevalence of vascular dementia is going down in Britain. I don’t know the statistics for here. So Alzheimer’s isn’t changing. The prevalence isn’t changing. But vascular dementia is. And again, it’s a nice example of stuff
that’s coming out of prospective studies because there have been all these claims about
physical activity protecting against dementia. And it turns out that all it is,
is if you’re losing your memory and if you — then you generally slow down. You stop being active. And so, if you actually ask
people about physical activity, then it predicts very strongly
the onset of dementia over the next few years over the first decade. But it really doesn’t predict
the incidence of dementia in the second decade because it’s not causative. It’s reverse causality. So it’s again the point that
Sarah Lewington was making. You — if you’re doing these prospective
studies, you’ve got to actually be careful to avoid reverse causality and
inferring what’s causal and what isn’t. But, yeah, I think that is. But also, I mean there’s just strokes. I mean you know, currently,
what is vascular dementia? Vascular damage. I mean stroke is just devastating. I mean it can be absolutely devastating. And that’s the — that’s such a definite
one, and the treatment is actually — you know, we’re giving how big
the effects of treatment is. I mean Sarah said that ten millimeters
systolic, which you could achieve with very, very widely practical regimens, would knock about 20% off your vascular
incidence rate, vascular death rate. But it’ll actually knock a little bit more than
20 off the stroke rate and a little bit less than 20 off various other things. So, yeah, definitely I think the
emphasis on stroke being left — I mean people are probably
more afraid of a nasty, really horrible stroke than they are of dying.>>I think that’s quite –>>And probably with good reason.>>I think that’s quite true. In fact, in our tobacco education work,
we found that what motivates the people to quit is not the risk of dying, but the
risk of being disfigured and disabled.>>Um-hmm.>>Let’s talk about tight control for a minute. Ten millimeters is a — is a support,
but the sprint trial and others suggest that lower is better in terms of blood pressure.>>Yes, that’s true.>>However, not all routes to getting
low are necessarily the same in terms of medication versus activity or other things. With blood pressure, if you’ve really
got an elevated blood pressure, you’re going to be [inaudible]. But what’s your sense of how low we should go? Let’s start with blood pressure. You mentioned cholesterol. There’s a lot of debate in this country about
standard treatment versus treatment to a number.>>Yes.>>I would distinguish hypertension, where the
number is probably more reliably predictive of the outcome than cholesterol, where we
may have a much more complex phenomenon that we haven’t fully elucidated
in terms of LDL.>>I think they’re probably comparably simple. If — you know, when you get all the — all the evidence rates, I mean I know
that you’ve got all the complications and the other cholesterol fractions. But I think that LDL cholesterol and
blood pressure are comparably simple because there may be an extra
complication on blood pressure. A fluctuating blood pressure with a
given average, may be more dangerous than a less fluctuating blood
pressure with the same average. But there are details. But overall, basically getting blood pressure
down will really reduce the incidence of stroke, heart attack, and various other vascular
problems, and including vascular dementia. But what can you achieve? The sprint trial you mentioned,
that’s pushing it. And they’ve got a difference of 15 millimeters. No trial has ever managed to get a
difference in treatment and control of more than 20 millimeters, not of all the
trials that have ever been done. There’ve been about 60 different
trials of blood pressure lowering. None has managed to get a
difference bigger than 20. So, you know, when Sarah’s got her graphs
up there with these notches, you know, where a doubling of risk goes
with each 20 millimeters, that’s a more extreme difference
than has ever been achieved. And if we were trying to talk about what can
you do in the general population, particular in, you know, the cities of — the cities of India,
China, you know, where you’ve got doctors who can diagnose this and you’ve got
good doctors, but you’re not going to have closely supervised treatment, then I think probably ten millimeters would
be pretty optimistic as an achievement as what you might actually achieve
by trying to put people into things.>>Well, many of those trials
are looking at standard of care versus a more intensive regimen, whereas globally what we’re looking
at is potentially no treatment.>>Yes, yes.>>At all, versus treatment. And you should be able to do
a lot better than ten there.>>Well, you’d be doing well
to do it and maintain it. I mean you’ve also got to — you got to
keep people on these things and to work out how to actually get things given. I mean it’s –>>Exactly. The question is –>>I mean I have slightly got double
standards here because I’m saying, look, if we got rid of all smoking, then
we could achieve magnificent things. And you could say, well, you’re not
going to get rid of all smoking. What are you going to get rid
of by some change in smoking, and — which would be a fair analogy. But I think — I think that as to what you would
achieve in a population, you’d be doing well in a population if you get a few millimeters.>>Yes.>>If you’re talking about
treatment of an individual, yeah, you get a really compliant
individual, you could get more than ten. But probably if you had a
program of just trying to treat — just trying to make sure that people
are routinely treated, you’d get — I would have thought you’d got
something around about ten millimeters, which would knock in the trials
something like 20% off vascular morbidity. And I think what you then need — the key
need I think is probably polypharmacy, and the question is how do we
make polypharmacy practicable? Because the statins are safe, they’re
effective, and there’s a 30-page paper in The Lancet last week giving
all the details on — that all these claims of outside
effects of statins are just rubbish. It’s revealed to be the longest
paper The Lancet’s ever published. And it’s clear that statins
are extremely effective, that the side effects are relatively negligible
in comparison to benefits for anybody who’s at an appreciable risk of
occlusive vascular disease. And one should be basically giving things
in combination if people have diabetes. Well, you know, when the British did their
trial of trying to give more intense treatment of diabetes, they just factored
in blood pressure lowering. That was the thing that really
was the most striking result. You know, it — you need blood
pressure lowering statins for diabetics as well as glycemic control. You need blood pressure lowering
statins and low dose aspirin for people with a history of occlusive vascular disease. I think we’ve got to work out how do we get
polypharmacy to be taken with cheap generics? How do we ensure that they actually stay cheap and that doctors don’t have
perverse incentives not to provide — not to prescribe cheap things,
which they do in China? You know, you’re never going to
make it big in a Chinese hospital if you prescribe cheap, effective drugs. So how can — how can we actually
get things to happen politically? But I think that’s actually the fastest
way, apart from tobacco control. It’s the fastest way of actually
getting CVD mortality down. If we want to hit this 2030 goal, we’ve got to
actually have medication, and we’re not going to achieve — we’re not going to achieve
these goals with one type of medication only. We need polypharmacy, and we need
to work out the politics of supply. I mean one simple thing, if
somebody at the moment wants to actually produce some combined pill
and say, This is a wonderful thing. It combines the benefits of blood
pressure lowering aspirin and statin,” then they’re not allowed to make that claim
because they would have to do a trial. It would cost them, you know, maybe $100
million to actually make this claim. If — so what I think we should
do is just change the rules that if a generic manufacturer can demonstrate
three things, then they’re allowed to — say you want to put together two or
three or four drugs into one thing. If they can demonstrate good manufacturing
practice, shelf life, and bioavailability, then they should be allowed
to make the claim that this is like giving all the things together. And that would be a changing medication.>>You said that to me — you said that to
me several times, and I’ve asked the Food and Drug Administration of the US who says yes,
we can do that under our current regulations. So I don’t know if that’s true.>>But they don’t do it.>>But that’s what they said.>>Manufacturers don’t believe, and they –>>Well, manufacturers aren’t producing it. But let’s get to that in a minute. I want to ask one last question before
addressing the question you raised at the end here. In the US, we’ve seen a slowing of the decline in cardiovascular disease in
recent years, and areas –>>Yes, since back in 1910 — since 2010,
it’s going down pretty well till 2010.>>Yeah.>>And then it’s sort of slowed since 2010.>>Are you seeing that elsewhere? Do you think it’s primarily
obesity or something else?>>I suspect in the case of the US,
it probably is primarily obesity. We haven’t seen it in Britain, and of
course, Mexico’s done the opposite. Since they’ve introduced their healthcare
— health insurance system in 2004 to 2012, it looks as though that which had stalled,
partly because Mexico’s the fattest country in the Americas, may have
started to go down again. But I’m sort of looking at that
a bit with the eye of faith. You can’t redo trends since 2008. But if I — if I trust the data on a
trend, maybe Mexico’s actually, you know, getting better treatment, and
therefore starting to decrease. Yeah, it probably is that. You need to — and also the decrease in tobacco,
well, I mean there’s only so far it can go. You got rid of three-quarters of it. And so, although you need to
concentrate on the last quarter. I mean, yes, certainly, it would
— tobacco deaths are going down. Obesity deaths are going up. So they are going to cross over at some state. They haven’t yet, but they’re going to.>>So let’s take your last point here,
and this is something that Sir Richard and I have had debated at various times. I think that we come to the conclusion
that what’s right is what’s right for the community and the country. You want to get as many events
and deaths prevented as possible.>>Community, and the country,
and the world, because what’s true in one country isn’t true in another. And if you’re going to try and get the places
where the bulk of the CV deaths are happening, of course, in Russia you’re doing
something about bloody alcohol as well, which we’ll come to later.>>So have you calculated what proportion of
the preventable early deaths can be prevented by taking the high risk approach
versus the general population approach? Of course, you’ll have to — the number
needed to treat will be vastly higher –>>Yes.>>For someone who hasn’t had an event.>>Well, you miss the people who
get wiped out by the first disaster. I mean although I think we should actually
put diabetes in as a disease and count that as secondary prevention in some sense
when you’ve got a diagnosis of diabetes. So, yeah, you’d miss all of the things that
just caused permanent, massive disability or death when they first strike you. So — and if somebody is actually at high
risk, you know, got high blood pressure, and you found it, you checked
and it was consistently high, you put them on treatment,
you were doing them a favor. So where you’ve got resources, yes,
that’s definitely a good thing to do. At the moment, we don’t have the
availability of convenient formulations. You know, if you were a Mexican doctor and you
wanted to actually put somebody on say triple or quadruple therapy, you can’t just prescribe,
you know, a simple sort of generic pill and sort of make the arrangements for the family to make
sure the person actually keeps on taking it. It just isn’t conveniently available. I think we need a bit more convenience
that doctors can do this if they want to. I’d concentrate on that. And that, of course, would help
both treating the hypertensives because if you’re hypertensive, then sure, get your blood pressure down,
but also take a statin.>>There’s a big difference between some of
the push for a poly pill that would be given to the general population and the fixed drug
combination that would include a statin, perhaps aspirin, antihypertenstives. Incidentally, I may have
mentioned it to you before, Kaiser in the US has a protocol called ALL. All diabetics over the age of a
certain age have to receive aspirin, A, lisinopril, and lovastatin, ALL. And with that really simple protocol, they’ve
been able to drive down cardiovascular mortality in the diabetic population enormously. And exactly as you say, for
a patient with diabetes, blood pressure is a thing
that will kill them first. Cholesterol is a thing that
will be easiest to control. And sugar will be the thing that
will be most difficult to control and for which you’ll have the most
difficulty showing an impact of control.>>Yeah. Yeah, I’d love to know what to do
about sugar intake because you’ve got the guy in the audience here who first proposed
the Mexican sugar tax and then managed to persuade the Mexican first Minister
of Health and then the President to actually go ahead and endorse it. So — and he — he’ll tell you about diabetes
in Mexico and what Mexicans are doing. Now we’ve always done — he’s got the price
model [inaudible], and now he’s got to have it so that the price of the sugary soft drinks
is triple that of the non-sugary ones. And maybe Trump could persuade the
Mexico president to be cooperative. [ Laughter ]>>I don’t know how we got there from
primary versus secondary prevention. [ Laughter ] But we’ll have time for one or
two questions from the audience. I think that the — we are
on the same page, and –>>Yes, yes.>>And we need to –>>This is a detail. This is a detail. We’re on the — we want to make
sure that things that happen — that the things that work get done,
whether it’s tobacco, diet, or medication. Drugs don’t work if they’re not used. That’s obviously. And there are obstacles.>>And for tobacco control, we haven’t
come nearly as far as we could globally. A lot more is needed, and tax absolutely
is the single most important initiative to reduce consumption rates, smoke-free
spaces, targeting ads, limiting advertising of the tobacco industry,
marketing of the tobacco industry. Yesterday there was an article in the Wall
Street Journal that Newports are being given out with $1 coupons all over the US. They get another generation hooked. So we have huge challenges
still with tobacco control.>>Yes.>>I think with obesity,
we don’t have a good sense of what can be done that will make a difference.>>No, no.>>The sugar tax — soda tax
— is without a doubt the –>>It’s doable. It has the advantage of being doable. Doesn’t it?>>And area –>>It’s not sufficient, but it’s
doable, and it’s worthwhile.>>Yeah. But I think, you know, honestly, we
don’t know what’s driving the obesity epidemic. We have a lot of theories. But we don’t know. And unlike tobacco, where we can
say, “If you’re a government, and you haven’t done these four
or five things, shame on you.”>>Yeah, do it. Yeah.>>”Because your people are dying
from diseases that they’re getting that you could prevent for
no money essentially.”>>Yeah.>>”You could gain money and save lives,”
whereas with obesity we have to try things and see what works and analyze it vigorously.>>Yes. Yes. I mean, yeah, it’s very difficult. I don’t know what would be the right thing. I mean you’ve tried imposing
portion controls in New York.>>Not me.>>I thought it was while you were — [ Laughter ]>>After my time.>>It was after the time you were in New York?>>We did menu labeling. We did –>>Sorry?>>We did menu labeling, and
we did government procurement. So what the government procured, we
required to be healthier food in schools, in childcares, in senior centers, in jails.>>Yeah.>>That seems to have made a difference.>>Yeah. Yes, I think it just needs to be
an experience of trying various things. Isn’t it?>>Yes.>>And one striking change when you come from
Europe is the difference in portion size. It really is quite striking. You know, whenever people come, they
say, wow, this — is this one portion? And it did — you know, that much — I don’t
know how everyone changes traditions like that. Anyway, it’s — but we are — we do
have this background thing though that overall death rates are going down. So we should be — yeah, we’d
like them to go down faster. Yeah, we’ve got 30 million
deaths before 70 every year. You know, but we are actually
winning, I mean in a way. If you look at basically from, you know,
1850 to 2050, then in that 200 year period, that was the time when we got
rid of premature death roughly. I mean it’s just — I mean premature death was
— you know, very, very few people would make it to 70 back worldwide back in the
death rates of the mid-19th century. Actually, the previous century was even worse. It actually started already by 1850. There had been some progress. But the change from then to when premature
death in childhood and early adult life and in middle age was just absolutely
a norm to the situation we have now where in many parts of the world it isn’t. I mean that’s an enormous change. And we’re still moving in that direction. So overall, we’re winning. But it’d be nice to be winning a bit faster.>>Do you know –>>If you take a look at
sort of a decades long view.>>Do you know of examples from low and middle
income countries where there has been treatment that has worked on a population-wide basis?>>Well, Cuba is the obvious example. I mean Cuba in the 1960s, they came in, and the
leaders were, you know, fanatical about trying to give healthcare because the
poor didn’t have any healthcare. And they actually gave things
that were cheap and worked. And the child death rate really dropped. And back in 1970 when WHO was trying to
classify the countries of the Americas, they got these countries with
low mortality in childhood and, you know, early adult and middle age. And the only three were Canada,
United States, and Cuba. Now lots of the others have caught up. I mean so Cuba now isn’t special. But it was early on, and
they did things cheaply.>>But in terms of treatment of
hypertension, cholesterol in middle age, and on a wide enough basis
to get a population impact?>>Well, I think we’re — I think that when we
look at the vascular death rates that I showed in Britain, that’s — you’ve got big
decreases because of the decrease in smoking. That’s a wonderful contributor to
the vascular mortality reduction. You’ve got treatment of acute events,
acute heart attacks, and so on. I mean that’s a big improvement. And we’ve got drugs that work. But I think we’re — I think we’re
actually getting to the point where these ordinary drugs
are being used at a level that is appreciably affecting
national death rates. And so, I would have put diabetics
in there as — along with the ALL. Now Samira’s walking on now. You know what that means.>>For hyper — for diabetics? Anyway, the — before Samira interrupts us — [ Laughter ] I think one thing that was notable in your
slides was the smaller decline in injury.>>Yes.>>And we know that as CVD comes down, injury is going up proportionality
as a cause of death globally.>>Yes.>>It’s not just war. It’s traffic fatality, but
it’s also suicide and homicide.>>Yes.>>Thoughts about that? Obviously, in Russia, if you can control
alcohol, you can do a lot for that. Globally, if we can improve technology
for road safety as well as enforcement and other measures, we can
reduce road safety deaths. But your thoughts about the injury trend?>>Yeah. The trends in a
lot of European countries over the last few decades have
been remarkably favorable. And traffic death rates have
really gone down a lot in countries who have actually got the money. And I think if you look at
places like India and China where you’ve got still very high death
rates from traffic, I mean it probably — there are things that could be done that
would be acceptable that would reduce — I mean traffic — road traffic accidents on
their own are more than a million deaths a year, and total injuries is about 6 million a year if you include suicide and
deaths a year at all ages. But I think — I think it’s not a sort
of priority as countries are struggling. You know, countries sort of come in, and they
want to do something about giving medical care to the population, but they don’t — they
don’t think of safety as being a sort of political priority, work
safety or traffic safety. So I think — I think this change
has happened in rich countries. I think it can spread to other countries, and it
could well be that some big political decision in a country like China or India
could really produce striking results. But I don’t — I don’t know
exactly what I would recommend, and there’s so many different things involved in
this to achieve — well, to reduce death rate.>>Of course, the Bloomberg Initiative has
attacked road safety as well as tobacco. So they’re getting some results there.>>You mean the supported road safety?>>Yes.>>Okay.>>Judy. Judy Monroe is the President and CEO of the CDC Foundation,
and if you’ll come to the stage. [ Applause ]>>So thank you, Tom. Good morning, everyone. So as you’ve heard this morning
from our speakers, cardiovascular disease is certainly a growing
global burden, and especially in lower and middle income countries
where access to primary care and essential medicines are lacking. So that’s why it’s really important
that we’re having this dialogue today. It’s really an incredible
opportunity to share knowledge from so many large population-based
studies all over the world in hopes of identifying even more
information that will be essential to improving national health policies
and care in areas that need it most. The CDC foundation is pleased to work
with CDC to draw attention to this issue and to hopefully draw support toward it as well. We need to tackle this challenge head on. It’s a health threat that continues,
and we need the continuing — the vital research that is
needed to address this. Some of you may know that the CDC Foundation
was established by Congress over two decades ago as a non — an independent
nonprofit to support CDC. We like to say that we help CDC do more
faster in its vital life-saving work. We have over 300 programs around the world,
and they range from preventing infections in cancer patients in the
United States to measuring adult and youth tobacco on a global scale. So it’s certainly our privilege
and honor to support CDC. This morning, in addition to diving into
extensive research on the global toll of cardiovascular disease, we also have
the privilege of honoring an individual who has been integral to this work for decades. His work and tenacity in presenting
his findings without hesitation, and no matter how seemingly
controversial at the time has saved and improved the lives of millions of people. The CDC Foundation for a number
of years has honored individuals that we believe are heroes in
the field of public health. We’ve done so through the presentation
of the CDC Foundation Hero Award, which was first presented in 2005. With it, we recognize an
individual’s significant contribution to improving the public’s health through
exemplary work in advancing CDC’s mission of protecting the health, safety, and
security of America and the world. Among the eight previous recipients of the
award are former US President Jimmy Carter, the founder of the world-renowned
development organization, Partners in Heath — that’s Paul Farmer — and the man who helped
lead the global eradication of smallpox, former CDC Director Dr. Bill Foege. So today’s recipient is among good company. Today, we are honored to
recognize Sir Richard Peto, one of the world’s foremost epidemiologists. We present this award to Sir Richard
for his career of visionary leadership and tireless advocacy in uncovering the root
causes of cardiovascular disease and cancer and in bringing data to bear
on public health policies that have saved countless lives worldwide. Sir Richard’s legacy in this
space is long and varied. Among his many accomplishments and
discoveries, he was the first known expert to describe clearly the future worldwide
health effects of current smoking patterns, and his work helped alter policies and popular
attitudes in many countries towards the hazards of smoking and the large benefits of stopping. He has — he was co-author with the late
Professor Sir Richard Doll of a series of reports on tobacco use in cancer, a 50-year
prospective study of mortality in relation to smoking among British doctors, and an influential 1981 report quantifying the
avoidable causes of cancer in the United States. He was elected a Fellow of the Royal Society
of London in 1989 for the introduction of meta-analysis of results from related trials
that achieve uniquely reliable assessment of treatment effects and was knighted by
Queen Elizabeth II in 1999 for services to epidemiology and to cancer prevention. Sir Richard’s tireless work continues, and
he does not rest on the laurels and findings over the course of his storied career. He recently collaborated in major studies of
alcohol in Russia and of malaria in Africa and India and is working diligently
with his Oxford colleagues to ensure that large-scale randomized trials,
meta-analysis, and prospective studies on cardiovascular diseases and cancer
can continue well into the future. So on behalf of the CDC Foundation and our
Board of Directors, it is my distinct pleasure to present to Sir Richard Peto
the 2016 CDC Foundation Award — Hero Award in recognition and
appreciation for his forward thinking work in protecting the health of so many people. Sir Richard embodies the spirit that CDC
Foundation Hero Award seeks to recognize, and his work and research is an
incredible example of the role of public health in our communities. So with that, I’ll uncover the magic sheet here. [ Laughter ] And Sir Richard? [ Applause ]>>I accept this award on behalf of
myself and also Richard Doll’s tie. [ Laughter ] Because when he was lying dying in
the hospital, I was saying last night that when he was lying dying in the hospital, I
said, “Look, I don’t want anything else off you after you’re dead, but can I
have some of your flashy ties.” He said, “All right. They’ll be no use to me.” [ Laughter ] So, you know, after he died,
I got this and a few others. So every time, particularly
occasions like this, I just wear them. Just –>>Well –>>And this actually not
only is Richard Doll’s tie, but it actually has some of
Richard Doll’s DNA on it. It hasn’t been washed in ten years. [ Laughter ]>>Congratulations.>>Thank you. [Inaudible] indeed. Thank you.>>It was a wonderful session this
morning, and now to continue on, we have — I have the pleasure to invite
Dr. Gajalakshmi Vendhan, Director of the Epidemiological
Research Center in Chennai, India.>>Good morning. The title of my talk is Risk Factors
for CVD [Inaudible] in India. I am going to show you the new
evidence from Chennai Prospective Study, which is the largest study in India. So you see south India. I’m coming from Tamil Nadu State,
and Chennai is the capital. And we are doing the study in Chennai City. The survey matter is a house-to-house survey. It was started in 1998, and we
completed recruitment in 2001. We have in — we have interviewed and measured
over 500,000 individuals, age 35 and over. That is we have half a million
individuals in our cohort. And [inaudible] rate in 2013 10,000
individuals, and as a quality control measure, we randomly select five persons
of the [inaudible] center, and we again repeat and remeasure all the data. This was done both for the baseline
data and for the resurvey data. Follow methods, so we follow them
regularly, all 500,000 individuals. We do it by two methods. One is active, and the other one is passive. By active method, we do the house visit
again and find out what is the status of the individual, and then we
collect data on that information. By passive method, we link our database with
data systems at the Vital Statistics Department. So we are using the data collected
through 2015 for this presentation. And while an autopsy was done for
all deaths to get cause of death because in the [inaudible]
environment, the cause of death written on the death certificate is less
reliable and also you see about 30% of deaths you don’t get cause of death. So statistical methods, all analysis
adjusted for the age of risk, sex, education, and socioeconomic status. To limit reverse causality, we excluded people
with chronic disease at baseline, like asthma, COPD, TB, heart disease, stroke, cancer,
and also the deaths that occurred in the first two years of follow up. For the risk factor analysis of systolic
blood pressure and BMI, we excluded smokers. So this slide shows systolic blood pressure versus ischemic heart disease
and stroke mortality. On the left side, you have ischemic heart
disease graph, on the right side, stroke. So death rate ratio [inaudible]
against systolic blood pressure. Here we have usual systolic blood pressure. That was explained by Sarah, and this
one is actually baseline blood pressure that is collected for regression dilution
bias that is coming from our resurvey data. So this slide shows the systolic. With increase in systolic blood pressure, the risk goes up for both
ischemic heart disease and stroke. So systolic blood pressure has a strong
past relationship with vascular mortality. So we have total about 10,000
vascular deaths in the study, and this slide shows strong positive
relationship with systolic blood pressure with the 20 — every 20 millimeters of mercury
lower systolic blood pressure halves the risk. This is a result seen in high income countries. So our baseline data shows ten units higher BMI
increases approximately ten millimeters mercury of systolic blood pressure. This is also seen in high
income country studies. So this is BMI versus vascular mortality
among 10,254 deaths, most nonsmokers. So we have seen BMI increases
blood pressure, and blood pressure, systolic blood pressure,
increases vascular mortality. So we might expect that BMI will
increase vascular mortality. But our study shows little effect
of BMI on vascular mortality. So we adjusted the analysis with
usual systolic blood pressure. Then BMI is inversely related
to vascular mortality. This is seen on the next slide. Here we adjusted the data for
usual systolic blood pressure. Then BMI is inversely related to vascular
mortality, especially in the lower BMI range. Same is seen for ischemic heart disease. So the left graph, it is not adjusted
for usual systolic blood pressure. The right side, it is adjusted
for usual systolic blood pressure. So BMI is inversely related to the
[inaudible] mortality, especially middle 25. The same is true for stroke mortality. The left side is not adjusted for
usual systolic blood pressure. The right side’s adjusted, shows
BMI inversely related to stroke. So next we move on to smoking
and drinking as risk factors. Here analyze what we meant, because in our study
data only less than 1% of women smoked or drank. And then we resurveyed 10,000
individuals in 2013. Almost all who were resurveyed
in 2013 reported the same smoking and drinking that they had at baseline. So smoking and drinking,
what’s the risk of mortality? Death rate is shown here 35 to 69 in Chennai. The first line is neither, is a nonsmoker
and nondrinkers, next smokers only, and then third drinkers only,
and fourth smokers and drinkers. So if the men who smoke only
and they do not drink, then they have 40% higher risk
compared to nonsmokers and nondrinkers. If they drink only and they do not smoker, then they have 30% higher risk
than nonsmoker, nondrinkers. So the risk is lesser than smokers only. But if they do both smoking and drinking, the risk is about 75% higher
than nonsmokers and nondrinkers. So if they have both the habits,
the risk is significantly higher than having only one habit, smoking or drinking. The next slide, smoking and drinking
versus all causes of mortality. The first line is nonsmokers/nondrinkers,
second smokers only, third drinkers only, and fourth both smokers and drinkers. So if they smoke and drink,
the risk is about twofold. That means 50% of those who drink
and smoke will die due to that habit. The smokers have 40% higher risk
than nonsmokers and nondrinkers. Drinkers have 50% higher risk
than nonsmokers/nondrinkers. That translates to one-third of those who are
drinking or smoking will die due to the habit. So when we did the study, the age at
starting smoking was 22 among men. So the age — the younger the age they
started smoking, the greater the risk. So in 2010, CDC did the tobacco survey in India. That study does show the age they
started smoking now reduced to 18. So in future, the smoke-related risk will go up. This may go up from 1.4 to maybe two. So then what happens? Those who smoke, 50% of them will
die due to the smoking habit. When that goes up, those who
smoke and drink will also go up. So Chennai Prospective Study
highlights, the 20 millimeters of mercury lower systolic blood
pressure halves vascular mortality. So ten millimeters of mercury lower systolic
blood pressure reduced approximately 30% of vascular mortality. BMI had little effect on vascular
mortality, and low BMI when adjusted for blood pressure was [inaudible]. In men, both smoking and drinking are strongly
and independently related to risk for mortality and were associated with substantial
excess mortality during middle age. This slide shows the prospective study
collaborators in Chennai and in Oxford. We [inaudible] — we thank all the participants,
half a million participants and funders for their support and cooperation. Thank you for your kind attention. [ Applause ]>>We have time for one or
two questions from the floor. The mics are on the table, and
there is a standing mic as well. Yes, Jeremy?>>Hello. Thanks so much for this presentation. I’m Jeremy Hussein from Global
Noncommunicable Disease Branch. So what are the risk factors for the women
in India, for the CV disease risk factors, and how different are the CVD mortality
rates are between men and women? That’s my first question. And the second question is given India’s
heterogeneous regional attributes, do you think the CVD risk factors are
different for different regions, and if so, the findings from these prospective
studies, how much generalizing are — they are in informing a national
level CVD intervention package?>>The women, they don’t smoke or drink. Just one in 1,000 smoke. So smoking [inaudible] blood pressure,
when blood pressure increases, cardiovascular mortality goes up, and also BMI. And what did you ask on the next one?>>The regional heterogeneity.>>Yes. The smoking prevalence varies, and also
intensity varies, and the type of smoke I find in [inaudible] India, 35% of not
[inaudible] Chennai, 30% of men smoke, but out of them, 75% smoke only cigarettes. And 25% smoke beedis. But if you take the whole
India, 70% of men smoke beedis. Only 25% smoke cigarettes. So our study shows if you smoke beedis, the
risk is higher than you smoke cigarettes. And so, you expect more higher risk in a
rural area for smoking than in an urban area. And also chewing tobacco, that is also
different in rural areas, and like [inaudible]. That is also different. And it’s also different in
different regions of India.>>Thank you.>>Thank you very much. Thank you, Gaja. [ Applause ]>>And now — [ Applause ] Here is Ben Lacey, and he’s the
epidemiologist at University of Oxford.>>Thank you, Samira. And good morning and many thanks for
the opportunity to present today. So my name’s Ben Lacey. I’m a public health doctor and
clinical research fellow at Oxford. And I’m going to present some findings
from the China Kadoorie Biobank on blood pressure, adiposity, and stroke. So the study design of the China
Kadoorie Biobank is really very similar to the Chennai Prospective Study.>>Get nearer to the microphone.>>Is that better? Can everyone hear? Is that much better? The study recruited half a million
adults from ten diverse areas in China, ranging from the freezing north
to the more tropical south, the wealthier coastal eastern areas,
and the poorer western rural areas. So participants completed a
computerized questionnaire, which included lifestyle questions
and information on medical history. Measurements were taken, including blood
pressure and height and weight to allow BMI to be calculated, and a small
sample of blood was taken for long-term storage in Oxford and China. So as with the Chennai Prospective
Study, there was a resurvey of about 5% of participants several years
after the baseline survey, and that allowed for correction
for regression dilution bias. The — I’m not going to show the statistical
methods because they are very similar to that that you heard from Gaja. But follow up is ongoing. But I’m going to present the results to 2015. So events were collected by the study through
electronic linkage to death registries, but also nonfatal events through linkage to
hospital records and CT and MRI scans used to diagnose [inaudible] quite widespread
in China now, and that’s really important because it allows us not just to look at
total stroke, but also to differentiate between the major types, such as ischemic
strokes where you have a blockage of the artery, or hemorrhagic strokes where you
have a bleed into the brain tissue. So if we look at the relationship between
systolic blood pressure and ischemic stroke, we see a strong positive relationship
throughout the blood pressure range examined with no evidence of a threshold below which blood pressure’s not
associated with increased risk. And the strength of that association,
so for every 20 millimeters of mercury, you see a doubling in the risk. And that’s what we see for
total stroke in countries like the US and other high income countries. But by contrast, when we look at hemorrhagic
stroke, we see a much stronger relationship. So for every 20 millimeters of mercury,
we see a tripling in the risk of stroke. So blood pressure drives stroke rates. When we looked at the baseline
survey in the China Kadoorie Biobank, we found that for every ten units of BMI, that increased blood pressure
by 20 millimeters of mercury. So that’s much stronger than you see in the US. You may remember Andy Gadget’s
[assumed spelling] presentation where ten units associated about ten millimeters
of mercury, so quite substantially stronger. So now if we turn to BMI, we find as with
blood pressure a strong, positive relationship, no evidence of a threshold throughout the range. And the strength of that association
is for every ten units higher BMI, you see a doubling of the
risk of ischemic stroke. So that’s what you’d expect from
the effect of BMI on blood pressure. Now we know that blood pressure
increases the risk of hemorrhagic stroke more
than it does ischemic. So we might expect an even
stronger relationship with BMI where it’s an actual fact
we see very little evidence of a relationship except
perhaps such a very high BMI. So as you saw in Gaja’s presentation,
if we adjust for usual blood pressure, we find little residual effect from BMI when we
look at relationship to BMI and ischemic stroke. And when we do the same for hemorrhagic
stroke, we see a strong inverse association. And this isn’t really understood, and we’re
currently investigating as to why this might be. So in summary, BMI drives blood pressure,
and blood pressure drives stroke rates. So this is my last slide, and
I wanted to mention treatment. Those previous analyses look at
participants without prior stroke. But if you look at participants with prior
stroke and within the normal tendency range, that is where most of the
participants with prior stroke were. We find that only one-sixth
have any long-term medication. So that’s blood pressure statins or aspirin use. It’s much, much lower than we find in the West. So lastly, I’d like to acknowledge
the principle investigators, first Lee Win Chang [assumed spelling] and
Lee Win Lee [assumed spelling] in China, the half a million participants,
and the funders. Thank you. [ Applause ]>>We have time for questions.>>Yeah. Thank you. I’m Henry Khan from our Division
of Diabetes Translation. Some of these results shown from India and the Kadoorie Data Biobank are
stimulating and provocative about the BMI. As I recall, the Kadoorie Biobank also
has measurements of abdominal obesity, and I wonder if those would be parallel to
the BMI findings or whether they represent and different phenotype expression of so-called
adiposity that might be very informative.>>So, yes, the kind of range of
adiposity measures always come from bio impedance [inaudible]
facts and things like that. They’re all very strongly correlated
in this, in this population. So the relationships are basically the same,
but you do see subtle differences in terms of the relationship with blood
pressure and with diabetes and some of which have been published
and some of which will be. So there’ll be subtle differences, but
essentially the messages will be the same.>>So the findings of BMI are very
interesting, and also interesting in light of the Chennai findings. Finding the inverse relationships
that we saw in Chennai, [inaudible] adjusted for
systolic blood pressure. So that’s what made your
relationships be inverse I think I saw.>>That’s right. So it was the same in Chennai, whereas in
Chennai we saw no relationship with kind of vascular disease irrespective of whether
it’s — whether it’s ischemic heart disease or stroke, and, yeah, very
strong relationships with stroke. But all the excess risk from higher BMI
is really attributable to blood pressure. That’s really what the China results are showing
so that when you take out the effect of BMI on blood pressure, you see the residual effects of the other pathways that
the BMI will be acting. And those give you that inverse association.>>And BMI [inaudible] about
what these other pathways are?>>That’s an idea, that it really
is a kind of unsolved question. I mean they could — they could be cholesterol,
and that’s kind of one of the, well, certain fractions of cholesterol, and
that’s been kind of being investigated. But there are kind of other ideas. But hopefully over the next
few years, we might solve.>>[Inaudible] there really are some other
mysterious factors determining stroke risk. But we really don’t know why
[inaudible] stroke death rates were down tenfold during the 1950s
and the Depression. But, you know, it was just [inaudible]
’60s, ’70s, when there were [inaudible]. And people had really stopped smoking by then. And we don’t know why there’s this
vast increase in stroke death rates in Japan between the 1950s and the 1980s. It went from being absolutely
predominant, you know, vastly more important than all the other causes of death put together, into being just one disease among
many just over a 30-year period. And within Britain in the 1960s, in those
days, the rich and the poor smoked the same. In those days, the rich — the
poor weren’t fatter than the rich. There wasn’t very much difference in adiposity. And there weren’t any treatments that worked. And yet, the poor had five times the
stroke death rate that the rich have. And now there’s reports coming out of China
from North China where stroke predominates of striking decreases in stroke there, which
can’t plausibly be accounted for by any changes in treatment or hypertension control. So there’s some — we’re missing
something really major in terms of what are the causes of stroke. Yeah, smoking matters. Yes, blood pressure matters. But there’s something else that really
matters that we don’t understand. And it — from the data you’ve
shown, it seems to be particularly for hemorrhagic stroke rather
than ischemic stroke.>>Yes. I mean I agree, and I suppose one of
the real strengths of this study is that rather than looking at kind of total stroke,
you’re able to look at finer detail of those particular subtypes and maybe even
possible finer than just purely ischemic or hemorrhagic, you know,
subtypes of the subtypes. And that might kind of reveal
some more information about the kind of causes of these diseases. But, yes, I agree. We — you know, we know more about ischemic
than we do hemorrhagic, but, you know, there are lots of things, important
things that we need to find out. I think especially in the kind of parts of China
where there’s very high hemorrhagic stroke rates where you do find higher rates than
you do in the West than, you know, the effective blood pressure is going
to be even greater than it is in areas where there’s you know, high rates of ischemic.>>Another question?>>Thanks so much. I have a quick comment and observation. I mean in the Chennai case, as it
was in China, is the fatality rate from the CVD events are much
higher, I assume, in India or China. So is that a factor? And also the BMI range starts from
quite low, from 15 onwards for India. So there is malnutrition stunting
all these issues involved. So the fatality rates from
the CVD events are higher. Is that something that explains the high
death rates and onset of the CVD events?>>So there are differences in the
cardiovascular types in two countries. So within India, you saw there’s much
higher rates of ischemic heart disease. And in stroke it’s really — sorry
— in China, it’s really stroke. So of the 20,000 vascular
events that we’ve recorded in the China Kadoorie Study,
17,000 have been stroke. So it’s quite a kind of marked
— a marked difference. We can — with a lower BMI, you find in
the Chennai study than you do in China, and there’s lower BMI in China than
you find in the Western countries so that the different strengths about examining
those relationships at different levels of BMI where there’s just not enough
people with a high BMI in Chennai to give us reliable information
about that relationship.>>Thank you again, Ben. Thank you. [ Applause ] We will take a short break. While we get settled, I just wanted to
let everyone know that I just learned — we just learned that many
people have joined by IPTV. So thanks to everyone who are listening
to this very interesting session remotely. And I with like to now invite
back Sarah Lewington to take us through the remaining of the sessions. Sarah?>>Thanks, Samira. So we’re going to hear now from
Mexico, Russia, US, China, and Cuba. And so, I’d like to start the next session
by inviting Dr. Jesus Alegre up to tell us about diabetes and death in Mexico. Thanks, Jesus. [ Applause ]>>Thank you very much. I’m more than honored to be in this — in this
building speaking about our results in Mexico. Let me show you that originally in Mexico
the problem is the glycemic control. The diabetes in Mexico has been a
problem that we have noticed long ago. But the study that we conducted originally was
appointing the follow up of tobacco consumption, probably alcohol consumption,
and in time we discovered that the problem from Mexico was diabetes. So we did this study of 150,000
people which have been followed. We ask them a lot of questions in a — in a
[inaudible] questionnaire, a lot of questions about their behavior, tobacco consumption, of
course, if they have or not chronic disease. Hence, we obtained the information
about being or not diabetic from this and then also a blood sample. We measured the blood pressure. And we make [inaudible] for
the BMI and waist and hip. When we have discovered is that the
prevalence of diabetes is 10% — more than 10% over the age of 50, and
by the age of 60, it’s more than 20%. And it’s an — it’s a 400% excess
of risk before the age of diabetes. And I will show the evidence in the next slides. So if you see here in this slide it’s
very clear that by the age of 50 is more or less 50 — 10% the prevalence of diabetes. And by the age of 60 is more or less 20%. Let me tell you that we have a national survey, and it is a national survey
called ENSA [assumed spelling]. We have the same — more
or less the same results. So this study is very similar, exactly
the same results to national — to our national survey, although we
have conducted only in Mexico City. So with the blood samples, what
the people of city used these for is the measuring of the HbA1c. So the people who told that
they were not diabetic, precisely they have good limits
of glucoside and [inaudible]. Well, the people who said they were
diabetic not only were diabetic — were diabetic, but with a poor glycemic control. In average, they have a nine measurement. What — and the people 50 years and under
were in the worst state, were above nine. So what we see here is that
in Mexico, we do not have — we haven’t achieved the proper
control of diabetes. If you’re a diabetic in Mexico,
you have low possibilities to be well treated is what this says. So we start to follow up 12 years ago
the mortality, and if you see any cause of mortality — any cause of
mortality in Mexico is impacted — has an impact in the — because of the diabetes. Diabetes causes five plus
mortality in people from 35 to 69 age old in any cause, in any cause. But I will also show you that this
impacts more in kidney insufficiency and in the cardiac insufficiency,
which is more important to show. So by this moment — in this moment,
we have in the follow up we have more than more or less 18,000 diseases. From these 18,000 diseases that originally
were recruited for the cohort, we have found — we’ll rehab the information of the
— of these death certificates. Let me tell you that in Mexico death
certificates are 99% all the country done by doctors. I mean it’s somebody’s doctor
who fills all the fields. And since we follow the rules very, very well, we don’t have a very clear
picture about diabetes. So many times, we see diabetes
written in the fourth — in the main [inaudible] of the fourth row
of the causes and to do a corporate analysis of the causes, what we come to do is
that only the people who have a program with their glycemic levels would be
considered dead because of the diabetes. I mean I [inaudible] all
people with hypoglycemia. And the people who died of whatever they die,
I mean cardiac arrest, kidney insufficiency, will be considered dead as — dead of cause. And we take all of the diseases among
these 18,000 and obtain the next numbers. If you see from all the causes of mortality,
including the [inaudible] the cardiac, the infected, the solid part is
the part that involves diabetes. And if you see this diabetes
block is very much important. But what is sadder, the most sad part
of this is that we still have a lot of people dying of diabetic crises. If you see the first four causes
are the main causes of mortality. So although if self — if diagnoses
diabetes by clinician is about 30%, or diagnosed by a [inaudible] is 34%. What we may say is that about 40%
of all — of all the problems of — of our [inaudible] infection is due to
diabetes in one-third of the mortality. But what we should do is to prevent by — and you will hear about methods
of prevention and treatment. The term [inaudible] will expose in a second. And to say in summary that
the most common cause of death in Mexico will be attributable
— attributed to diabetes. In Mexico, diabetes is poorly controlled,
causes unnecessary high mortality, and 40% of all deaths before —
between 40 and 80 — and 80 years old. This is where we would like to thank
too, especially Dr. Richard Peto, who has been a great support since the
beginning of our study, and we haven’t — we couldn’t really have done
anything without his support. And all the people from Oxford have
been very supportive and kind with us. Thank you very much. [ Applause ] I have an extra one. I have an extra one, extra time. Sorry. This is the real risk of
dying of cardiovascular diseases and from renal diseases in
Mexico if you are diabetic. You can see here you have diabetes and
you are between 35 and 59 years old, you have 30 times more risk in all percentage
times, more risk to die of renal insufficiency, and five times at least of cardiac disease. That’s it. Thank you very much. [ Applause ]>>Thank you very much, Jesus. I think we’ll take questions after both
talks because we’re now going to hear from Pablo Kuri-Morales, who’s
the Undersecretary of Prevention and Health Promotion, also Co-PR of the Mexico
Prospective Study on what you’ve been doing for diabetes control recently
and your plans for the future. Pablo?>>Thank you. First of all, thank you for the
invitation, and I will start with a thought around what we’re going to see, and the
first thing that I want to mention is that we’ve have heard a very nice set of data. But it becomes nicer when it
becomes translating to policy because actually having this data
published, very important, very nice, but when you start doing something with
it to deal with the population and to deal with the health of the population,
it becomes much more important. And this is a story that started 22
years ago when I met Richard Peto, that time Professor Peto
here in the US at Miami. And we started talking about the study that
Jesus just described, and it was the leverage that has translated into public policy
after almost two decades in Mexico. I’m going to talk about what this story
has become in terms of public policy. As you see, there was a launching the
way Richard was in Mexico at that time. He was — he was witness of this launching by
the President, and the first thing that I have to say is that to have a sound policy in our
country, you have to have political wield. So this was launched by the President of
the country, by the Mexican President, and the objective is quite clear. It’s to improve the wellbeing of the Mexican
population by slowing down the increase of — increasing prevalence of overweight
and obesity, reverting the epidemic of noncommunicable diseases, particularly Type
II diabetes through public health interventions, a comprehensive healthcare model,
and [inaudible] healthcare policy. And the results on the study that the Mexican
prospective studies, one of the key pillars of making this possible, not the only one. As Dr. Friedan said, we have to have
different sources of information. But this is one the main sources. And this is the conceptual
framework of this strategy. You have three pillars, public health, medical
care, health regulation and fiscal policy, guiding principles, and the first one you see
there is research and scientific evidence, but also shared responsibility, transversality, and intersectionality, evaluation
and accountability. And what for? To increase public and individual
awareness of obesity and NCDs. And let me tell you something. The name is something that had
to do with social marketing. It was very long to put hypertension
and dyslipidemias and all that. So we decided to put overweight,
obesity, and diabetes. But we didn’t forget the other things. It’s just a matter of social marketing. And what we wanted this to develop our
guidance, the guidance of the national system of health towards early detection, which is
very important, to detect and control this type of diseases at the first contact at
the primary healthcare facilities, and to slow down the increase of prevalence of
obesity and noncommunicable diseases in Mexico, which has been growing out in the last I
would say precisely two or three decades. And Richard posed a question at the end
of the last talk of the first session, which was there’s something that
we don’t know that has influence. And I think the answer might be
in social determinants of health. There are many things that are outside of
health that contribute to health problems and that have impact in health,
sanitation, water, education, conditions that change inequities that have
endured and that we don’t measure properly, but at the end of the day have
impact on this type of diseases. What advances do we have? We had what we call the check
yourself, limit yourself, and move yourself mass communication
and social marketing campaign. And the results were that 85% of — there was
an 85% of remembrance rate of this campaign. Ninety percent of the respondents
said the campaign encourages people to improve eating habits and physical
activity, and 61% of the respondents agreed that the campaign has changed their behavior and
perception related to overweight and obesity, which was one of the first
issues that we mentioned. We had to make the public aware of the
problem because everybody talked about it, but there was not a social awareness,
and there was not a response. Then accountability, we developed the Mexican
Observatory of Noncommunicable Diseases. By the way, you can go into
the link that is posted there. You can see what — it’s in Spanish, sorry, but you can see what actually
that observatory is all about. It’s not a governmental observatory. It has to be outside the government. It’s administered by the Autonomous University
of Nuevo Leon and measures the impact and evaluation of this public
policy from national strategy. And there’s some figures there. From January to June 2016, we have more than
152,000 interactions, more than 23,000 users, and you see the countries that — which have visitors from to see what
we are doing and what’s posted there. Some advances in medical care, as mentioned
in the first session, you need information. Of course, the Mexico City prospective
project, as I said, was one of the main drivers and the leverage — on the
leverage for this strategy, but now we needed to see what
actually was going on in field. So we developed a chronic disease information
system, which allows health personnel to register the health treatment provided
to patients that live with chronic diseases. And it accounts for what happens in that number of primary healthcare facilities
in our 32 states. Mexico is a country — a
federated country with 32 states. And those are results, and as you can
see, the update is of two days ago because it’s updated on a regular basis. And we know of the people that get
to those health facilities of more than 1.7 million people have at least
one consultation in the last year. And the interesting thing is the figures
that you see on the right side of the slide. Of those, about almost a million have diabetes. The important thing is that
we’re not doing the proper job. We’re just measuring 22% for HbA1c, which
should be 100% because we have all the financial for doing that on all the population. But of those, 40% are under control. In hypertension or in blood
pressure measurement, that’s better. We have 75% of people measured
with 66% under control. With obesity, that’s a figure there. I’m not going to read it. You can read that on the slide. And with dyslipidemia, specifically
with measurement of LDL, here you have 28% and 41% under control. This is not enough. So we needed to improve the
quality in medical care. So we developed what we the Excellence in
Diverse Care Networks, with four components, early detection, quality of care,
training health professionals, which is something very important because
you ask health professionals to do things, but you don’t give them the information
and the knowledge to do the proper things at the first — at the primary healthcare. And one thing that’s very
important, to have the supplies. You have to need — you need to have the
reagents to do the diagnosis and then, of course, the Metformin and the
— and all the — what is needed. And now we have that in 26 states. We reach almost 6 million beneficiaries and
strengthen the reference and counter-reference of patients in the — in the system. Other results, well, those are
the achievement indicators. We are not yet there, and you
will see we’re not yet there. But we have those achievement
indicators perfectly well defined, and we all know what we are striving to achieve. And this is where we are. We have not yet achieved those. In fact, when you compare these to the
regular program, the differences are basically in the people that are being measured. In the control, we’re still way behind of
what we want to be — where we want to be. But now we have a system that allow us
— allows us to go to each of the — of the healthcare units and see what’s going on within each healthcare
unit, which we didn’t have. Something that’s very important, we have
many changes from the guidance for the sale and distribution of food
and beverages in schools. We had an amendment in the constitution
whereby foods and drinks that do not contribute to health of students are banned from school. Now the problem is to implement that. That’s a challenge because we have
a lot of resistance from the — even from people, of course, from
the industry, but from many others. An amendment of these articles of
the general — of the education — of educational physical infrastructure related
to the installation of water drinking fountains in elementary schools because you
tell them you have to drink water, but you don’t give them the means of doing so. So we changed that, and we
are in the process of that. And we updated the nutritional guidelines for
the sale and distribution of food and drink — and drinks in schools within the national
education system, and it applies to everybody. And of course, it promotes
drinking plain water and consumption of natural foods such as vegetables and fruits. Front of package labeling and targeted
advertising to young audiences, we had an amendment on the regulation
of sanitary control of products and services published in the official
gazette in the — about a couple of years ago, which made mandatory front of package labeling
the nutritional quality seal and restrictions on advertising of high calorie
density food and beverages in open and paid TV as well as movie theaters. And what are some of the results? A decrease in almost 27% in advertisement
and of the people, there has been a request for 798 sealed requests and only 126 have
been granted because deals are not — have not fulfilled and are under review. We have 297 requests. Something that has made a lot of — has come
to a lot of interest everywhere in the world because we’re always asked about it,
it’s important, yes, to have the taxes, but it’s not the only thing we
have to do, as I have shown. Of course, we passed — well, not we —
the congress passed a change in taxes. Basically, energy drinks and flavored and sugared sweetened beverages is
one peso per liter, which is a lot, and a 8% increase in all the products
that you see that are in the list. And the results are there. I mean I don’t know if you knew,
but Mexico was the largest — world’s largest per capita
consumption consumer of soft drinks. And the tax was implemented in 2014 and after
one year, what we observed and is published in the British Medical Journal is that there’s
a decrease of about 6% in the taxed beverages, and it goes up to 12% in
the lowest income deciles. So price does matter. And we have an increase of
plain water for around 4%. Now with need to assess and that’s a process that we’re working on, the
sustainability of this. Next steps, in according to what we presented, we need to assess the health
impact of these measures. We are currently carrying out what we call the
National Household Survey of Health midterm, which is currently carried out, and
we’ll have the first results on October to assess precisely the national
strategy for prevention and control of overweight, obesity, and diabetes. But it’s important to let you know that the
results that Jesus just presented have led us at the policy making level to conclude that we
need to rethink and review the current policy and if necessary make adjustments
and strengthen the three pillars that I just mentioned of this strategy. So again, we are in a properly — I would say
based decision making process using scientific evidence from studies such as the
prospective Mexican study, Mexico City study, as the household surveys, as the chronic
information system, to adjust our policy. And we think that it took — it took us about
30 years to get to the plan that we have. So it’s not going to be solved
in two years or three years, but I think with this strategy we have
set the foundations in my country to deal with the problem in a I would say comprehensive
way, thinking that it’s not only a problem of health, that you have to deal
with other things outside of health to really have an impact on this. And I just want to acknowledge that to come up with this strategy 14 government agencies
were involved, seven private institutions, four international agencies,
three academic organizations, and one nongovernmental organization, and
of course, for this specific presentation, I want to face — say thanks
to my staff at my office. Thank you. [ Applause ]>>Thanks very much, Pablo. So does anybody have any questions
for either Jesus or Pablo?>>I have a question.>>Mike?>>Were the people not being treated
because they didn’t have access to care or because it wasn’t normative
to treat people with diabetes?>>No, they were not being treated because of
the lack of compliance from the health services. They all — they all have access to — well, not all, but access to healthcare
in Mexico is quite widespread. We talk about universal coverage,
which does not mean efficient access. But everybody has the right to be covered. But the differences are so big
that actually what happened is that they were not doing properly
what they had to do at the — at the primary healthcare facility. So that’s where we’re working, which is
training and making all the — all the — what is needed medicines and all what is
needed to be there at the — at the contact.>>Does the health information
system have the capacity to record uptake and adherence to treatment?>>Record what we are doing?>>To record and follow the uptake of treatment?>>Yes, and –>>Adherence.>>Yes, because what we did is basically we have
internet access in about 50% of the facilities. And those that don’t have it,
they can download it and then go into a place where we have interact access. So that system is online. It works for all the system
of the Ministry of Health. Let me just remind you that in Mexico
we have a social security which accounts for about half of the — of the population. But it’s interesting. They don’t have the same plan that we do have
basically because of social security accounts for people with higher income
and that have better access. And — but we are dealing with the poorest
of the poorest in Mexico within the system. I have a question about — in the
mortality of the renal disease. It is a striking finding
that the number one cause of death among people with
diabetes is renal disease. It’s much higher than — the risk of death for renal disease is a much
higher among Mexican people with diabetes compared to other populations. So could you explain why this is the case? Any of the chemical mechanisms
or environmental [inaudible]?>>There are several — there are several –>>Underlying this strong association?>>There are several that in fact are published. There are several studies published
about genetics in Mexico on diabetes. There are environmental issues,
and there’s still a lot to learn why is it that we’re
having that problem. But there’s some things that have been
published, specifically by Aguilar Salinas on the genetics of diabetes
in Mexican population. And there’s studies way back that were done
in Mexican American population, and we have — living in the US, and we have higher rates
than the population — the native population. So there’s a genetic factor among others.>>Okay. Thank you. So in the interest of time,
I think we’ll move on. Now unfortunately, David
Zaridze can’t be with us. He wasn’t able to get an interview
in time to obtain his visa. So Richard is going to present
David’s results on his behalf. Thank you, Richard. [ Applause ]>>All right. Okay. You’ve got to pretend that that’s me. [ Laughter ] Actually, David Zaridze is leading the
largest prospective study in Russia, and I’m presenting some results
from it on his behalf. He was the first cancer epidemiologist
who actually managed to work outside the Soviet Union back in the
1970s, and he came, worked with Richard Doll, and became determined that he was going to
try to get tobacco taken seriously in Russia. And in the 1980s, he was discussing
with Richard Doll, with various people, about the possibilities of getting
studies going and then things changed in Russia, and the system really collapsed. But they did — they did
manage to get some changes to cigarette manufacturing,
which were probably favorable. But he’d always wanted to do a prospective
study, and then as things settled down after the transition from communism,
he finally did start to do this, but also did retrospective studies. So — and he’s been a collaborator
with us for about 40 years. And he works — I mean all of these
studies, they obviously for obvious reasons, they run through the generations. So he works with a younger
colleague, Alex Boroda, in Moscow, and Sarah Lewington who’s chairing this session. And I’ve worked with him and
with other people in CTSU. The [cough] study areas are three cities
in western Siberia, Tomsk, Barnaul, Biyk. It’s the outside region there
in western Siberia. And they’ve got fairly typical
mortality patterns. And when I say typical, I mean catastrophic. Russia has dreadful adult
mortality statistics, and the — and I’ll come to the conclusion and
then give you the evidence for it. The Russian death rates shoot
all over the place. They just go up and down, but they’re
very high, and also they fluctuate widely. I mean you get a twofold difference in mortality
within a few years, and the reason for this is that they drink a lot of vodka or
strong spirits, but mainly vodka. There are some newsworthy things about, you
know, drinking alcohol that was intended for industrial, and, you know, they all
drink various other odds and ends as well. But it’s predominantly vodka. And so, I’ll call it vodka. Then what I’ll show is expressed
in vodka equivalents. Okay. What are these high
rates and rapid fluctuations? I’d like to illustrate it by
saying what’s the probability that 15-year-old is going
to die before the age of 55? So I’m going to look at the probability that
a 15-year-old will die before the age of 55. And here is British data for the period
1980 to 2014, so over a 35-year period. And you can see the top one is males. The bottom one is females. British death rates have been drifting
down, and now at current death rates, you’re talking about 6% of the
males would die before they’re 55. So there’s 6% of the men in Britain
and about 3 or 4% of the women in Britain would die before they’re 55. And if the men didn’t drink or smoke,
then we’d probably be down to about 4%. Now what about the women? I’ll look at the women in Britain
and the women in Russia now. So I want you to look at the females in Britain,
where you got 3 or 4% dying before they’re 55. And I want to compare them
with the females in Russia. Now you can see in the Russia female data these
large waves, this up and down in the death rate. Now this is real. This is not in any sense an
artifact of registration. They — the deaths in Russia
are accurately counted. The denominators are accurately counted. There was no disruption to the system
of accurate counting when it changed from being part of the Soviet Union
into being an independent state. And the reasons for these large
changes lie in alcohol use. So if we — I can make this work. No. There we are. Ah, here. Here we are. Good. Okay. Over her at the left, we’ve got that there
is Gorbachev’s first year as Russian leader. He came to power in March, and in May that
year, two months after coming to power, he’d issued his first anti-alcohol law. And it was quite strict. It produced about a 25% reduction
in total alcohol use. It wasn’t a bigger reduction because a lot of
the alcohol use in the Soviet Union was illegal. It was drink that was made in the
villages, distilled in the villages, and then sold in the towns illegally. So it’s difficult to control
illegal supplies like that. But anyway, he came up with this. And so, even in the year where that law came
in, you’ve got some reduction in mortality, and then in the first full
year of its operation, you’ve got quite a big reduction
in female mortality. I mean it took the women halfway from
being Russian women to being UK women as far as death rates are concerned. Things stayed, you know, for the next few years
— this is the rest of the Communist period. So there’s the end of Communism here. The restrictions on alcohol were breaking down
because, you know, when you get restrictions on official supply, then the unofficial
supply comes up to meet the demand. And so, in the late Gorbachev years,
as Communist authority was going, the alcohol policy, the anti-alcohol
policy was weakening. And then just overnight at the end of
that year on December 31st of that year, the Soviet Union ceased to exist, and
all of the anti-alcohol laws went. So overnight, there was no
restrictions on vodka sales. And I can remember being in Moscow at that time. You could see the old grandmothers from the
villages with their sort of liter bottles of vodka distilled in the villages selling them
by the roadside outside the Cancer Institute. And Russia collapsed. Industry, half of all industry collapsed. So half of all jobs went. If you had a job, you probably
weren’t getting paid anyway. And if you got any money, then
within about three or four years, that would all go because
96% of your money would go because there was 2,500%
inflation between 1991 and 1994. So there — and there was
just complete collapse. And it was — the whole country
was so demoralized that even the birth rate went down by a half. You know, there was a total — it
was a total collapse of morale, collapse of infrastructure,
collapse of food supply. You know, food prices went up. The only thing that went down was vodka. And so, there was a huge increase in vodka. And the overall death rate you can see is
almost doubled, up to its peak in 1994. And at that point, they got the system a bit
under control, and, you know, things came — well, for the middle 1990s, the
inflation was about 7% per year instead of 2,500% over a three-year period. And then things got a bit better. Vodka consumption went down. And then here, there was again a banking crisis. The Asian banking crisis spread to Russia. The ruble collapsed in value again, and
there wasn’t any sort of trading system. If you’d got an industry running, you
couldn’t even arrange to sell your goods and buy components for your factory because
there was no trading system working. You know, US dollars weren’t available. The ruble wasn’t trusted at all. And so, you were — you were finishing
out getting barter between factories. You know, we’ll give you 100 tractors in
exchange for 1,000 sheep or something. And it carried on absolutely
chaotically until about 2005. Things started to get a bit more under control. They got sort of gas production. They got agreements to sell gas abroad. And they all — they introduced
an anti-alcohol law. There were alcohol regulations, and over the
next five years, the consumption of spirits, vodka and illegal — including illegal
stuff, went down by about a third. And the death rates went down by about a third. Now you may think, look, okay,
he’s saying all these things. But how do I know it’s true? And the — I’ve shown you the women just
deliberately so that I could tell the story. But here, on exactly the same
scale, is the results for the men. Look at it. You know, you’ve got — I mean at this point,
the death rates here of 1994, you have something like 37% dead before the age of
55, whereas in Britain it’s 6%. You know, it’s gone drifting down
from about 10% down to about 6%. But here it was 37. And look at that huge increase, because of
course vodka is predominantly a male problem. Some women drink. The main female problem for vodka is
probably being hit by drunken men. And then here you see things getting better. There you see the ruble collapse, and there
you see the effects of the anti-alcohol laws. And as you see here, it comes down
from about 36, 37% down to 24%. You know, the vodka went down by a third. They went down from drinking nine times as much
as the British here down to drinking six times as much as the British down there. So it’s gone down from nine times what
we drink down to six times what we drink. And overall mortality went down by a third. Okay. And the main reason for this was
alcohol, and I’ll show you the two main studies that David did, which illustrate this. Okay. First was a retrospective
study of 50,000 dead people where you may think dead people aren’t much use
to epidemiologists, but they’ve got relatives. And if you want to actually know what
a Russian man drinks, then you go and ask his wife after he’s dead. She’ll tell you. So heavy drinking is strong here. So ask the families of 50,000 dead people
what the dead person used to drink. And you find there’s just some particular
causes of death strongly associated with drink, particularly diseases and deaths
from accidents, violence, suicide. Also, you could look at these national
mortality trends, and when you — trends down by what diseases people are
dying of, you find where’s the variation. Well, you put these together, eight
diseases plus injuries, put that together, and that’s what’s accounting
for this vast variation. Cancer rates didn’t vary. Child mortality didn’t vary. Death rates in old age didn’t vary. Even death rates of myocardial
infarction didn’t vary. Lots of things didn’t vary. It was the things that were closely associated
with alcohol, the things that were found by this retrospective study to be
closely associated with alcohol. They were the things that were driving these
wild fluctuations in national mortality rates. Okay. So I remember when I presented all
this evidence with [inaudible] Boston at the Edinburgh International Epidemiological
Association, people said, oh, my God, I thought Richard Peto was
a serious epidemiologist. All this retrospective stuff, all
these national trends, you know, you got to do proper prospective studies. Well, fortunately, David had already got on with a proper prospective study
and was about ten years into it. And so, the results came out a
year or two later in The Lancet. And they found that in male smokers — I mean there’s not as many heavy
drinkers in a prospective study because they won’t come into your study. The real skid row drinkers aren’t —
don’t ever get into prospective studies. But in the ones who were heavy drinkers, heavy
drinking was strongly associated with death from the aggregate of these
causes, not the aggregate of deaths from all other causes, so again the same thing. And the association was very strong. Okay. So I’ll give a bit more
detail, these studies now. So now we’ve got our prospective data,
and we can be respectable epidemiology — actually, in this case, because the
effect is so extraordinary, the — actually the retrospective studies,
which start with 50,000 deaths, are in a sense they’re cruder, but
they’re almost more informative about how bad the problem is. The national mortality trends are cruder still, but they actually tell you
how bad the problem is. And the prospective study, although it’s
useful scientifically, it actually — you know, you can take — get sort
of all possible biases and objections because you underestimate the problem. You can’t calculate the attributable fraction
because you don’t get the ones who are most at risk of death into your study. Okay. The retrospective study, 50,000 Russian
deaths in 12 years, 1990 to 2001, so in 2001, we started interviewing, interviewed 50,000
families, and they were collaborative, very collaborative, about the smoking
and drinking habits of the deceased. And the controls were 5,000 of these deaths,
and they were from diseases that were unlikely to be related to smoking or drinking. The cases were the deaths from each of the
other causes, and here are the results. So here are the diseases that were much
more common in people who were drinking, and we took our exposed group as the ones who
were averaging about a bottle of vodka a day. But actually, in Russia, a bottle of vodka
a day doesn’t mean a bottle of vodka a day. It means two or three bottles of
vodka every two or three days. You know, they blitz drink. And that’s why it’s so dangerous,
versus reference men. Well, I’d have like to have had
nondrinkers as a reference group. But we couldn’t because there weren’t any. So we had to actually take the ones who were
drinking less than a bottle of vodka a week. So that’s our control group. So we’ve got twice the risk of
liver cancer, three times the risk of what they call non-MI acute heart disease,
which basically just means heart stopped for no obvious reason, but not a heart — not an
MI, three times the risk of death in pneumonia, four times the risk of death from cancer,
and that’s in esophagus, four times the risk of death from TB, six times the risk of death
from liver disease other than cancer, you know, particularly cirrhosis obviously, seven
times the risk of pancreatic disease, mostly pancreatitis, and then eight times
the risk of dying from an unspecified cause. So you put all these together, and you finish up with four times the risk
of dying from any of this lot. And then for the other diseases that weren’t
control diseases, we had a relative risk of 1.4. And you can work out from these
that the relative risk for death from any disease was about twice. So they’ve got twice the death rate
of disease, these heavy drinkers. But then what about other things? What about accidents and violence? And that was extraordinary. So here we’ve got that the drinkers had
twice the risk of dying from a medical cause, but they had four times the risk of dying in a
traffic accident, six times the risk of dying of some other accident, eight times
the risk of committing suicide, and ten times the risk of being murdered. And for the women, it was even worse. They’ve got 15 times the risk of committing
suicide and 20 times the risk of being murdered, you know, because drunken women
hang out with horrible men. And we could take our study areas. You remember those three cities? That was Tomsk and then the [inaudible]
republic area just next to Tomsk. And we could look at the mortality
trends there, and here we are. So here at age 35 to 54, that’s 1990 here. That other dotted line vertical
is the year 2000. So this goes from 1990 to 2001. So that’s the time we were doing the study. And you see absolutely that spike in
mortality you see in Russia as a whole. And then we can say, well, in
this area, what was the variation? What were the — what were the
mortality trends in the diseases we found to be closely related to drinking? Well, there you are. And what about mortality in
these from other diseases? Not very much. So it was the diseases that are strongly
related to drink in the retrospective study that were driving the mort —
the total mortality pattern. Then there’s the same thing for 15 to 34,
and there’s the same thing for 55 to 74. And again, you can see what’s driving things. Then here’s the prospective data. Now the solid black is deaths from
the disease that was pre-specified as alcohol related, so the eight diseases. And this is probability of death at 35 to 54. Look at it. You know, the 30%, remember
that, you know, in Britain, 6% of men would die at that
age, in that age range. But here in the top group, the ones who
were drinking say about — this is now — sorry, not in bottles today —
this is now in liters a week. So two and a half liters a
week, that’s five bottles a week because their bottles are half a liter. So you get this very strong — the black
is death from the diseases that we were — we’d found in the retrospective study associated
with drinking, and there’s the same 55 to 74. The dotted line, that’s almost too faint
to see, is deaths from other causes, and that wasn’t related to drinking. And so red is the total. So again, the prospective
data really sorted out. Now the problem is that in Russia, drinking
and smoking goes so closely together. If you’re a heavy drinker, then
nearly all of them are smokers. And those who aren’t smokers
used to smoke in the past. So nearly all heavy drinkers have smoked. So you got this total confounding. It’s difficult to do. So the main analyses of drinking have
to be restricted to people who smoked. So it’s just that what I showed you is among
smokers, what’s the relevance of drinking? But if you want to know the effects of
smoking, you’ve basically got to look at that in the people who drink less
than a bottle of vodka a week, you know, just exclude the heavy drinkers. And I’ll just show you one result on smoking. So here, this is among Russian men with no
prior disease who didn’t drink much, then, a, stopped smoking versus all cause of mortalities. So here’s the ones who keep on smoking
relative risk 1.66 for all causes of mortality. Now here’s the ones who’ve stopped at about
30, stopped at about 40, stopped at about 50. And you can see that if you stop at around
age 30 or 40, then you avoid most of the risk that you’d have if you’d kept on smoking. So again, it fits again with
the Western evidence. It — and 1.66, just to remind you, if you
had a relative risk of 1.66, that means that, you know, 66 out of every 166 deaths
wouldn’t have happened if the drink — if the smokers had had nonsmoker death rates. In other words 40% of the deaths among smokers
wouldn’t happen at nonsmoker death rates, 40%. So it’s — you know, it’s something,
sure, Mike Toon’s done studies where it’s 60%, but still 40% is pretty bad. So Russian male deaths in middle age,
what are the attributable fractions? And we have to say for drink
we still don’t know. Most of them, and you can say
confidently that most of them are the way that drinking has dominated the national death
rates, but that relative risk of 1.66 suggests that about 40% of the smokers’
deaths are from tobacco. The global adult tobacco services that get
done here — well, in collaboration with here, say that 62% of the Russian men smoke in
middle age, put these two numbers together. That suggests that smoking’s causing about
a quarter of all Russian male deaths. And, you know, other methods suggest,
you know, indirect arguments looking at national lung cancer rates
suggest it might even be a third. But for alcohol, what can we say? Smoking and/or alcohol all [inaudible] may well
cause about three-quarters of all male deaths. How exactly are we going to measure this? You can’t get a representative survey of people
and get accurate information about drinking. It’s a very difficult exposure study,
much more difficult than smoking. Anyway, thank you very much indeed, David. And thank you to Alex Boroda. Thank you to Sarah Lewington. I’m sorry David couldn’t be here. There were just visa problems. And it wasn’t the Americans being
obstructive or the Russians being obstructive. Somehow they’ve got their rules,
and they were trying to be helpful. And that was even more true of Muris
[assumed spelling] when she was — you know, the American Embassy staff
yesterday were desperate to try and help Muris get out of
Cuba, get this meeting. And somehow, their own systems, they’ve got
so many different ways and checks and things that they’re completely tangled up. It’s not malice. It’s just somehow somebody needs to take an
axe to all the systems and just simplify them. But the American Embassy staff were
trying to be as helpful as possible. They’ve got the plane waiting on the runway. It’s going to take off at 4:30. That’s yesterday afternoon. If she could get the visa to
the American Embassy by 2:30, then she’d be on the plane,
and they’d hold it for her. But even then, that didn’t happen. So thank you to the Embassy staff
for trying to help Muris get here, and Sarah will tell you what she
would have said if she had been here. So thank you. [ Applause ]>>Before –>>[Inaudible] if you want to ask any questions.>>You don’t want to answer questions?>>I don’t know anything else.>>He doesn’t know anything else. [ Laughter ] So worst case, we’ll move on, and we’re very
grateful to Mike Toon, who’s going to talk to us about the tobacco evolution of the epidemic in
the US and China on behalf of obviously himself and Professor Zhang Min Chen [assumed spelling].>>Thanks. It’s a pleasure to be here. And –>>Just click it.>>How do I advance it? No.>>Enter.>>Oh. The — all the slides from China
are from Zhang Min Chen and Richard. They’re the ones in black and white. So the question I thought about is how did the
evolution of the tobacco epidemic in the US, how does that inform disease control in China? So this is a map of smoking prevalence in adult
men in which China and the former Soviet Union and Eastern Europe have prevalences
above 52%, 53%. And if we were to have the map in women,
it would essentially be the inverse. The West would light up more than —
more than — certainly more than China. So I’m going to talk about the
timing of the epidemic in the — in the two countries, the difference in the
stage of the epidemic, and also with respect to gender between the US and China
and sort of where each country is in terms of the stage of the epidemic. And I’m going to emphasize the importance of
local data, especially from large cohort studies for monitoring the effects
of major risk factors. So this is a very familiar slide to you. It’s featured in every Surgeon General report. It’s the rise and fall per capita cigarette
consumption in the US from 1900 to 2012. And I just want to point out two things. One is that the rise began at the beginning
of the 20th, and the peak occurred in 1963. And then subsequently, we’ve had a fall that
takes us back where we were before World War II. These figures on your right, the ten a day and
three a day, that’s averaged over all adults. If you restricted it to smokers, so in 1963,
about 40% of the population were smokers. So that would actually be 30 a day. That’d be a pack and a half, and
I’m telling you this to contrast it with what I’m going to show you for China. So here’s annual Chinese cigarette
production from ’52 to 2011. All of the consumption essentially is
domestic, and all of the production is domestic. So if you look at 1963 on the left
side of that, they hadn’t even begun. This is when the US peaked. And it’s been pretty much uphill ever since. And the y axis is in billion. So if you look just at global cigarette
consumption in the world in terms of billions of sticks, and I’m actually going to
look at in terms of trillions of sticks, you’d see that by 2000, we were up to
5 1/2 trillion cigarettes in the world. That’s not what I want. I don’t know how I get it to go back. Okay. And so, China is consuming between a
quarter and a third of all of the cigarettes in the world back at the turn of the century. Okay. We’re back to per capita
cigarette consumption, and what I’m going to do is superimpose this line on the age
standardized lung cancer death rate in men and in women, men in yellow,
women in the purple. And what is striking about this is that the
peak cigarette consumption occurs in 1963, but the peak lung cancer death rate in men,
the age standardized rate is 25 years later, and in women it’s 35 years later. And this was occurring despite
a big decrease in the percentage of men and women who were smoking. I just showed the men here. So as we go to China, the fact that the peak
impact on mortality occurs 20 or more years after the peak smoking is
going to be highly relevant. So why an over — why this overshoot? It could reflect that cessation takes
a long time to produce benefits, which I’m going to show you in a minute is
not true, or it could show you that the risk in people who continued to smoke,
who haven’t quit keeps going up. And that I’m going to show is what’s going on. So this is the slide from American Cancer
Society Cancer Prevention Study II, and it’s very similar to
British Doctors’ slides. The red is the cumulative probability of death
from lung cancer in someone who continues to smoke, and then the dashed lines below it are
people who quit at progressively earlier ages. And so, those lines diverge from the
red line at the age of which they quit. And about five years after you quit, you start having substantially lower
risk than if you had continued. So the idea that it took 20 years to have a
benefit from smoking for lung cancer is wrong. The interesting thing is that quitting
smoking produces a benefit much sooner than the harm when you start. [ Laughter ] This just shows you how the trends in smoking
prevalence in the US went from 1965 to 2010, and you can see that a big part of the reduction in prevalence occurred between
1965 and about 1990. And at the same time, the machine
measured tar level and nicotine level in US cigarettes had decreased normally. So — enormously — so everyone is expecting
that the risk would be going down in smokers. So what we did is we studied mortality
risks from smoking in three — in large cohort studies across
three time periods. 1959 to ’65 was the American Cancer
Society Cancer Prevention Study I. ’82 to ’88 was II. And 2000 to 2010 were five large
pooled contemporary studies from the NCI Cohort Consortium. And they’re the cohorts that everyone’s familiar
with, and there’s this health study, et cetera. So basically, we’re going to be looking at the
early ’60s, the mid-80s, and the early 2000s. And so, these contemporary cohorts,
they had to be in North America. They had to have at least 50,000 total men
and women and be followed beyond the year 2000 and have updated information on smoking,
which we didn’t have CPS I and II. And this just shows you that if you combine
men and women, we had more than 500,000 in all of the cohorts, and we had
at a minimum 22,000 deaths. Okay. So what I’ve done is shown
you in relation to the curve for the age standardized lung cancer death
rate in men and women where CPS I, II, and the contemporary cohorts, what they’re
capturing, their slice of the epidemic. You don’t think of cohorts as cross-sectional,
but really they do represent a slice in time. And you can see that they’re
nicely spaced across this epidemic. So what we found, and I’m going to start with
the men down below, is that the relative risk for death from lung cancer back
in the early ’60s was 11.4. And by the ’80s, this had increased to 22.4. And it stayed essentially stable. It went up to just under
27 in the pooled analysis. Now moving up to the women lung cancer,
their relative risk in CPS I in the — in the early ’60s was 2.7 because
they hadn’t been smoking long enough. CPS II, it was up to 11.9, almost where
the men had been in the first study. And by the pool, they’d caught up with the men. And very similar pattern —
slightly different pattern for COPD, the big increase in the relative risk
for men occurred in between the 1980s and the early 2000s, just a huge jump
in the relative risk from 9.7 to 24.5. And in the woman, you have
pretty much progressive increase from the early ’60s, 5.9, to 23.5. Basically, the men — the women have
caught up with the men in both of these. So the key features that one can take from the
epidemic in the US just in this brief time is that the early studies massively
underestimate the ultimate risks, that — and I haven’t shown you this — the risk
is highest in those who start young, smoke more cigarettes, and don’t quit. And over time, as smoking became the
norm, this pattern became more dominant. And then in Richard Peto’s words,
women who smoke like men die like men. So I want to switch to China
where you have a massive — so you might say, well, if we know all this
about smoking, why would you need to study it in cohort studies in China where you have
obviously massive numbers of male smokers? We know that certainly the background risks, and we’ve seen this morning even the
relative risks, vary across countries. And the — really the most important thing is that policy makers mostly
pay attention to local data. They don’t believe data from other countries. They believe data that happens to the people
who they’re supposed to be responsible for. Just to show you how many
smokers we have in China, the 300.7 million smokers just
way overtops any other country. The US didn’t even make this slide. So I’m going to be looking at two large
cohort studies that the Oxford researchers and their collaborators have
carried out in China. The first one is the Chinese prospective
smoking study, and it’s basically the 1990s, and the second is the one
you heard about already, the China Kadoorie Biobank Prospective
Study, which is from 2006 to 2014. And this slide here shows where these two
studies started in relation to this increase in annual Chinese cigarette production. So the people in the first study, you
could see that the number of people who were long-term heavy
smokers would have been small. These next two slides are going to show
that pattern — the next three slides — the patterns of smoking in
China first among women. So China I think is anomalous
in that back in the 19 — mid-1930s, there was a lot of
smoking in two areas of China. In the northeast and the southwest,
just under 30% of women smoked. And it has decreased over
time so that it’s down to 2%. So have an anomalous thing in which — in
which women in China are not currently smoking. And the same is true in these
other areas of China. Now the opposite has happened in men. On the left, you see the — what’s
happened to the prevalence of smoking and the white squares show rural areas. The black squares show urban. So if you just look at the — at
the prevalence in the white squares, you see that it’s higher in
the country than in the city. You see that increased over time
until between 1960 and 1970. If you go to the right and you
look at cigarettes per day, you see that it’s increasing particularly in
the rural areas in men, and it’s increasing some in the urban areas, although it
seems to have leveled off in both. This is really important. If you look at the age when people started
smoking regularly, and you had the year of birth along the x axis, you
see that with each birth cohort, the average age when people
started smoking went down. So as in the US, as the smoking
epidemic progresses, people start smoking earlier in life. And they’re more likely to smoke cigarettes
when first started, as you see on the right. So these are the — here’s the
relative risk for all-cause mortality by age of starting smoking in men. And here, the x axis has
the younger age on the left. So as you go from — towards a younger age
that’s starting, the relative risk goes up. And it goes up even more in the
second study than in the first. Yeah, so there’s several patterns here. The relative risk is higher in the
urban areas than in the rural areas. That’s the opposite of prevalence of smoking. The all-cause mortality relative risk
goes up with earlier age at starting. And it’s higher in the second
study than in the first study. So these — the relative risk in
let’s say the top bar in those who — this shows that the relative
risks for all-cause mortality goes down when you quit compared to if you continued. So quitting is helpful in
China as it is in the West. So this shows is the relative risk for
all-cause death rates in the urban men in the top five lines and rural men in
the bottom five lines in the two studies. Patterns I want to point out is again
you see that the earlier people start, the higher their relative risk and that the
relative risks are higher in the second study than in the first study, and they are higher
in the urban men than in the rural men. But what I want to point out to you is that
even in the second study and the people who — the men who started under age 20, the
all-cause relative risk is just under two. So in the modern cohorts in the US,
the all-cause relative risk is three. And this — you see exactly the
same pattern for lung cancer. It’s higher in the urban than the rural. It’s higher when you start earlier. It’s higher in the later
study than in the first study. But again, when you look at the hazard
ratio, the relative risk in the second study for the men who start under age
20, the relative risk is 3.8. In the US, it’s 25. For stroke, you see exactly the
same pattern that you saw for — pretty much the same pattern
you saw for all-cause mortality. So the key features of the epidemic in
China is that it’s currently limited to men, but to massive numbers of men, that the risks,
both the absolute risks and the relative risks, are lower than in the US, but they’re
rising in birth cohorts that began smoking at younger ages and smoke more heavily. Quitting reduces risk, and remember cigarette
production and consumption are on the rise, and the population’s getting older. So this is from Richard that whereas in Western
Europe and North America cigarette sales and male lung cancer and male deaths from tobacco are decreasing,
the opposite is true in China. You have rising cigarettes sales, rising
percent of male deaths due to smoking, such that in 2010s, it’s about 20%,
2030s, Richard projects about 30%. So the conclusion for China, the main
conclusion is that in the absence of widespread quitting, the
worst is yet to come. And then this is rather bizarrely formatted. It’s really Richard’s message
is prolonged smoking kills. Stopping smoking even in middle
age avoids most of the risk. Smoking cessation is the principle
intervention for the people who already smoke. Reductions in initiation are the ultimate goal. So he always says prolonged smoking kills. Prolonged stopping works. And I will stop there. [ Applause ]>>Thanks, Mike, for that kind of fascinating
comparison between the US and China. In the interest of time, if there’s just
one question, anybody has a question?>>Oh, you know what I didn’t mentioned? And probably most everybody knows it, but the Chinese government is the
producer of the cigarettes in China. And the proceeds from the cigarettes
support the educational system. So in terms of changing social norms, it’s going
to take even more in China than it takes here.>>Thanks very much, Mike. So final presentation before we
wrap up, and now you have to — you’ve pretended that Richard was David,
and now you’re going to have to pretend that I’m Muris Armis [assumed spelling], who as you’ve heard despite heroic
efforts I think may be Samira, Lauren, and Marcella deserved a hero award for
their efforts to get Muris here today, but unfortunately as you’ve heard, the last
hurdle we just weren’t able to manage it. So I’ll be presenting the results on her behalf
and of Alfredo Dwinnus [assumed spelling], who is co-PI of the Cuban Prospective Study. And I should just say that as a photograph,
when Muris was visiting Oxford last year for the Richard Doll Consortium
Meeting, meeting of prospective studies, all looking at these major modifiable
risk factors for cardiovascular disease. And I think this meeting
really comes out of discussions that we started with Samira at that time. So the Cuba study, this is also a study of
150,000 adults who were recruited between 1996 and 2002 in the five urban and rural areas
that you can see from the arrows in Cuba. They recorded — it was a very simple
questionnaire, just a single sided A4 page, but managed to cram really quite a lot of
information, particularly valuable information, on the smoking, drinking, SBP, BMI, et cetera. No blood was collected for this study. Again, in the usual Oxford collaboration
tradition, we had a resurvey ten years later, so 20,000 participants were resurveyed
some years later to check for consistency, correct for regression dilution bias, and
the follow up is certified causes of death. And I’ll be presenting again. That’s ongoing, but I’ll be
presenting data up to 2015. And again, we’re using the same statistical
analysis approach, the same attention to detail, that we described — that
Gaja described in more detail. So the main analyses that I’m going to show you
from the Cuban study are of smoking, drinking, which in Cuba is mainly rum, and the
combination of smoking and drinking. And then — and then I’ll go on
to look at blood pressure and BMI. So first of all looking at smoking,
and here we’re looking as my — sorting them out by the birth year. So we see quite strikingly — so men
and women separately by birth year, and I think you see quite striking
little effect in either men or women among those born before 1930. And so, the main results that
I’ll be presenting will be of men and women combined born beyond 1930 or later. So similar size of relative effective smoking
among the men and the women born after 1930. And one of the striking things in
Cuba is the number of people who — or the number of children who
start at really young ages. So we have substantial numbers
starting before 15, but even substantial numbers
starting before they’re ten, so between the ages of five to nine. And I must say here that in the data
cleaning, we actually removed people. We assumed it was an error for the people
started smoking before they were five. I hope that’s true, but in a — maybe not. But even at ages five to nine, substantial
numbers starting to smoke really very young. And we see that as you see from China and
from the USA and from most other studies, the younger people start to
smoke, the greater the risk. And so, substantially greater risks for these
people who started before they were ten. And this is adjusted for cigarettes per day. So this is assuming the same
number of cigarettes per day. What’s the effect even in middle age? And these are — deaths are occurring in
their 50s, 60s, and 70s, having started even for people who are dying in their 70s, having
started young gives higher death rates. And then if we look by cigarettes per day,
so given age began for people who began at the same age as you might expect, the more cigarettes you smoke,
the higher the risk of death. And so, if we put these together for
the people who began smoking young and smoked considerable numbers
of cigarettes a day, then the risks are approaching those in the US. So at present though, in the whole study, the
relative risks are lower in Cuba than in the US. So we see lower risk for lung cancer
and for vascular disease, but I — if you remember from Mike’s talk, these
relative risks are maybe comparable to what we were seeing among
women in the CPS I in the 1960s. So I think if Cuba carries on
to smoke like the Americans, if women smoke like men, they’ll die like men. If the Cubans smoke like the Americans,
they’ll die like the Americans. [ Laughter ] But again, quitting works. It really does. So we have limited data on quitting, but
among the people who were still alive at 55, if they’d stopped before the age of
45, then they’d have avoided most of the risk of carrying on smoking. And we have limited data unfortunately
on the people who’d stopped after 45. But certainly, these data are consistent with all the other data you’ve seen today
showing that stopping really does work. And if you can stop early
enough, then you can avoid most of the excess risk associated
with carrying on smoking. Moving onto alcohol, and here I’m going to show
the total alcohol given in equivalents of liters of rum per week, because as I said,
the main alcohol that is drunk was rum. There were a few nonsmokers. Sorry, the nonsmokers in the women drank much. And so, these analyses will be
among — largely among males. So looking at the amount drunk
versus all-cause mortality, and I think this is strikingly
different to Russia. So here in Cuba, so the — these are compared
with people who drank less than weekly. So this is comparing people who drank
more than weekly or at least weekly with people who drank less than weekly. And we see that actually among
people who drink less than a bottle of rum a week, there’s really no excess risk. We don’t see any excess risk among those people. On average, they’re drinking about half,
just over half a liter of rum per week. But that seems to give little
excess risk of death. And even the people who are drinking quite
heavily, so drinking about two liters of rum per week, that’s only about a
20% higher risk of death than the people who are drinking less than weekly. So if we put those together and again think
back to the Chennai study where the people who were drinking had similar risks,
similar relative risks to smoking, here compared with people
who neither drank nor smoked, there was about a 60% increase risk among
the people who smoked but didn’t drink or were only drinking light drinking. Among the people who drink — who were just
drinking, not smoking, of course, as we’ve said, there aren’t as many of those, but
there’s really no strong evidence of a strong excess risk. But the people who are both drinking and
smoking have a relative risk of about two, so a twofold risk compared
to the people who do neither. So we’re now seeing that 17% of
all male deaths are due to smoking, compared with 1% of all drinkers
will be killed by their habit. But half of the people who both smoke
and drink will be killed by their habits. Moving onto blood pressure and BMI, and I
think you might have gathered this by now, that in all of these studies,
we see about a 20 millimeters of mercury lower systolic blood pressure
approximately halves vascular mortality again throughout the range, again without any
evidence within the range that we can study of lower blood pressure not
being associated with lower risk. Now in Cuba, the BMI wasn’t quite as strongly
associated with systolic blood pressure as it is in the high income studies
and as you saw in China. So here in Cuba, about ten units of BMI
is associated with about eight millimeters of mercury higher systolic blood pressure. So looking at the associations
of BMI with vascular mortality, we see the strong positive association that can
largely be explained by the association of BMI with blood pressure down to about
22 kilograms per meter squared, so about 22 units BMI continuous association,
lower BMI associated with lower risk. And again, this given blood
pressure, so this tells us that most of that excess risk can be explained by
the effects of blood pressure on BMI. But beyond the effects of BMI,
there’s little further — sorry — beyond the effects of the BMI on blood pressure, there’s little further effect
of BMI on mortality. And then finally, looking at previous diagnose
— previous diabetes versus all-cause mortality, so compare with the Mexico study about a
sixfold increased risk, in the Cubans under 60, a threefold risk of death —
all-cause mortality for the people who reported have diabetes compared with
not, so 160% excess risk due to the diabetes. I say compare that with Mexico
where the relative risk was 5.4, in other words a 440% excess risk
associated with the diabetes. So to conclude, we found that smoking
was responsible for 17% of all deaths, relative risk of 1.6 for all-cause
mortality, but importantly, those who started before they were ten and
smoked heavily had much higher risks than those who started later and smoked lightly. And importantly, stopping works. It works in the US. It looks to be working elsewhere. And it’s working in Cuba. Drinking, on the other hand, was only
associated with about 1% of all deaths, and there was little hazard, maybe surprisingly,
until more than a liter of rum per week. And even then, it was quite a small
excess risk due to — due to the drinking. As in all the other studies
we’ve seen 20 millimeters of mercury approximately
halved the risk, but the — and the BMI factor on mortality appears to be
largely due to its effect on blood pressure. And finally, diabetes, if it occurs, it seems that the Cubans do much
better I’m afraid than the Mexicans. So again, just like to say that it’s — we’re sorry Muris couldn’t be
here to present these findings, but this is very much a collaborative effort
between Muris and Alfredo in Cuba and Muris, who really does make this study happen. It’s Muris who keeps this studying
happening, keeps the data coming. And then the team in Oxford,
who we’re very grateful to for preparing these analyses for us today. Thank you. [ Applause ] So I think now I’ll go back to chairing. I think Richard has just got a few words to say
to sum up the morning, talking about studies in lower and middle income countries. Thank you very much. Well, I’ll be very brief
because we’re running over time. I think that most of the points I want to make
could be made by the things that have shown, that all of these studies have
produced that were not expected. [Inaudible] — [ Inaudible Comment ] Okay. Take this back. All of the studies have produced
results that weren’t expected. In China, the stuff that Ben showed about
this very strange relationships of BMI, blood pressure, and hemorrhagic
stroke weren’t expected. They’re similar to things that Gaja’s found. And in Russia, the effects of
alcohol are quite extraordinary, and they had to be documented by studies. And we need further studies to find
out it’s such a major cause of death. It’s probably half of all the deaths of
Russian — well, of the Russian men anyway. Mexico, diabetes, the extraordinary
effect of diabetes, not only is it common, which comes from obesity, but when you get
it, then you have this much greater risk. And this will have effects
on the treatment strategies that the Mexican government approaches. Cuba, you have this very odd result
about starting to smoke before ten. Nobody else has got that, and I believe that
this will lead to real efforts to make sure that kids do not start before
ten or actually before 15. And in the United States, well, when Mike Toon
wrote up that stuff for the successive studies, the 1960s, 1980s, 2000s,
he published it in 2013. And I just thought it was a beautiful use of
prospective study data, really imaginative. But it isn’t what was being looked for. All these results we’ve got,
they’re relevant to practice. The Chinese results will affect
what the Chinese government does. The Indian results will affect
what the Indian government does. The Cuban results will result in
legislation, the Mexican results. But we didn’t know at all. We didn’t know in the 1990s. If you were to ask us in the 1990s
what the point in doing these studies, our answers wouldn’t have
had any relation to reality. The only answer we could have given then that
I’d give now is that if you want, you know, if you study big causes, you’ll find
something that is unexpected and important. And you don’t know what that is. And of course, that sounds very attractive
to people who have been given grants. I’d give the same answer now. If you keep on studying big important
causes, you’ll find things that are odd. The things will come out in
the 2020s, 2030s that are odd and that you weren’t predicting
back in the 2010s when you wanted to either start these studies
or keep them going. So just — you can avoid more deaths
by moderate reduction, a big cause and a big reduction of a small cause. So study big causes as well
as doing other things. Oh, one other thing I was going to say, you
know, when those Russian results first came out, of course, Gorbachev was the only hero because he was the one who’d introduced
the reductions in alcohol consumption. And he — so Gorbachev, when this study was
published in The Lancet, he was a bit annoyed about the Russian people rejecting
him when he was trying to be nice. And he went on nationwide Russian TV
saying it was a really important study. And, well, that’s good. We’ve got Gorbachev saying
what we had done was important. And then — but the next thing, the day
after that, Medvedev went on Russian — nationwide Russian TV saying
these results were important, and I’ve told the Health Minister
to make an anti-alcohol plan. So the whole world, you know, words. But then they came up with an
anti-alcohol plan about a year later. But Putin — see all this
anti-alcohol stuff since 2005, that was Medvedev who started it, not Putin. And then — and for a time, Putin actually
stood well back from it because he knew that Gorbachev had got into
political trouble for, you know, trying to actually restrict
alcohol, as in fact did the czar. You know, the czar restricted
vodka supplies to Russian troops in the First World War, and
that caused the revolution. And then Gorbachev, you know,
withdrew vodka from the revolutionaries in the 1980, and that ended the revolution. So Putin was keeping really quiet on vodka,
but he found that it was actually popular. An elected government can do
things that dictatorships can’t do. And actually, the Russian government liked
— the Russian people liked the restrictions. So after a couple of years of watching
carefully, Putin came in, and he said, well, you know, I support these efforts to try
and restrict vodka use, and, you know, we don’t want the Russians to be a population
where so many people are dying before 70. So we need to actually do something
about vodka and about smoking. And, you know, that’s what he said. So, again, we’ll see what happens. But if they really act on it, then by the 2020s, Russians could have West European death
rates instead of East European death rates. They’re not genetically different. They’ve just got too much
vodka, too many cigarettes. So these things do have effects on
what happens, and they’re interesting. Okay. Thanks very much. [ Applause ]>>Well, what a wonderful morning
and an outstanding program of experts and practitioners from around the globe. So I want to thank all of you
for being with us this morning and especially those who’ve traveled
from afar to be here at CDC with us. Your willingness to share in all your
experiences to continue to inform the field, increase the commitment we all
share in the important goal of halving global cardiovascular
disease mortality rates, a goal achievable through
sustained focus and the coalescence of committed partners to fortify the effort. It has really been a privilege
for the CDC Foundation to join CDC in hosting this important symposium, along
with all our global colleagues and partners. And as we’ve heard this morning, there’s so much
excellent work being done all over the world, but again, there is a tremendous
need to address the serious issue of increasing cardiovascular
disease rates and risk factors. No work is more important
than that of saving lives. At CDC and the CDC Foundation, we’re
fortunate to partner with renowned and prestigious institutes, partners, and
organizations that provide the data needed to implement life-saving solutions. Equally important is ensuring sustainability
so that the work can continue to be built upon and crucial to continuing this work is a strong
commitment to both resources and partnerships that are so vital to the landmark studies that inform the cardiovascular
field both now and into the future. This symposium and the opportunity to
partner with Oxford University comes at a very fortunate time when there is really
unprecedented alignment among our partners who are passionate about tackling this issue. So we look forward to the future when we
can all look back and see how far we’ve come and how much progress has been made. And so, with that, I do want to just thank all
of you for being with us again this morning and certainly hope that you will join
us on this journey that we’re beginning as we embark together to make a big difference. So enjoy your afternoon. Thank you. [ Applause ]