Dietary Supplement Practicum (9 of 21): Efficacy & Effectiveness in Dietary Supplement Research

>>Cindy Davis:
Please join me in welcoming Dr. Sesso. [applause]>>Howard Sesso: Thank
you for that kind introduction, Cindy. It’s a pleasure to be
here this morning. I appreciate everyone’s
willingness to come here sans liquids, and
hopefully caffeinated to the extent that
you need to. [laughter] So, it’s always
fun to be here. You know I did this
talk last year. It was really good
to hear a lot of the questions I was getting
actually from all of you after it. Because a lot of what you
heard thus far is focused on kind of supplements
definitions, some of the legal issues,
marketing, labeling. But I am coming in from a different perspective here. I am a — I am
a researcher in epidemiologist
by training. I am involved with a lot
of the large-scale trials that have been conducted
over the last decade or two, looking at vitamin
supplements and health. And there is always a
question that comes up with all these trials
about whether or not, what we’re getting out
of these trials reflects either efficacy
or effectiveness. So, I like to kind of
walk you through at least my take on these two
terms, what they mean, as well as how they apply
to the types of research that you might be
conducting, and especially in the context
of clinical trials. Okay, so, we know that if
you take the traditional routes of nutrition
epidemiology we have seen a consistent association
between different foods and lower rates of
cardiovascular disease, cancer and many other
chronic diseases. We see these studies
come out all the time. And, the challenge has
always been how do we translate these findings. And we also — we also
know that if you go beyond the foods, beyond
the dietary patterns, and start delving more into
some of the specific vitamins, minerals
and other nutrients. That we see that there is
a strong justification for this potential
benefit. It’s why a lot of us
are in this room today. And what we have always
asked as researchers is the simple question of,
you know, if the foods are good for you, are
there specific components in those foods that might
explain or drive this potential benefit
or a fact. And inevitably for better
or for worse, we evolved into a question of
dietary supplements and what their role is. Can we somehow bottle, or
create a supplement that captures these benefits
in a way that we can have meaningful improvements. Not just in terms of
health individually, but in terms of
public health. Now that might be a
little too up high in the sky, but at the same time
it’s extremely important to remember that we
always want to continue to emphasize all the
healthy dietary patterns. So, it’s always important
to emphasize all the dietary patterns. But inevitably people
still turn to supplements for many different
reasons which we’ll start to walk through today. And what’s always
interesting is that we always hear about
different kind of vitamins and minerals of
the day, and different dietary patterns. What tends to happen
is that the dietary supplements use tends to
increase ahead of the definitive clinical
trials that help to really determine whether
there is any degree of efficacy or effectiveness
on those outcomes that people think that
they are taking these supplements for and
hoping to get benefits from. So, I think one thing
that we need to first do is just have an
understanding of the distinction between
efficacy and effectiveness. It sounds similar, but
they actually denote two different concepts. So, when we think about
the interpretation of studies, and particularly
clinical trials, in up for dietary supplements. So, when you think about
a clinical trial, it could be done in one
of two different ways. It could be done in a
kind of a real-world setting, where it’s
not as controlled. We’re not monitoring
patients quite as closely. And alternatively, it
could be done in a more optimized environment. Not necessarily treating
subjects as lab rats per se, but having a much
better sense of what people are taking, how
they are taking it, and being able to capture
and measure that very carefully over the course
of the clinical trial. And this is a nice
diagram that kind of asks two basic questions that
you want to link with either efficacy
or effectiveness. So, if you’re thinking
about efficacy, you always think of efficacy
as kind of the ideal perfect world scenario,
can this treatment or this dietary supplement
operate well or be beneficial under
ideal circumstances. On the other hand, we
know that, in the real world, sometimes people
take pills, sometimes people don’t. Sometimes people are
taking them all the time, they’re taking twice as
much or not as much. So, effectiveness
reflects the idea of whether or not a
particular treatment, or in this case, a dietary
supplement, reflects ordinary, or real-world
circumstances. So, I think this is one
— as you can imagine as we walk through this,
there are strengths and weaknesses to each
concept and how we either conduct the trial or
interpret the results of a trial that’s testing
a dietary supplement. And if anything, my hope
is that you’ll walk away from my little piece here
this morning being able to ask these questions
when you read about different studies or
hear — or if you’re conducting the
research yourself. So again, just to kind of
reiterate this, when we think of efficacy trials,
we’re asking the question whether that treatment
helps under ideal circumstances. And there is always
particular aspects of efficacy trials that you
want to highlight or look for, if you’re reading a
paper or trying to design something. So, first of all, do the
patients or subjects accept the interventions
that are offered to them? It’s a basic question
of compliance. So, it’s not just whether
they’re going to take those pills but also are
they going to follow the instructions faithfully
and take those pills, whether it’s daily, two
times a day, three times a day, four pills, one
pill, whatever the intervention
might consist of. Efficacy trials also
assume that you’re getting the best possible
care and you do not have care for other diseases. So, it’s always focused
on the exposure of interest as well as the
outcome that you’re interested in in
that clinical trial. So, most trials — if
you’re a trialist, and you’re trying to put
whether it’s a small trial, or a large trial,
you’re usually designing trials this way. You want it to be nicely
controlled conditions, everyone takes the
intervention, everyone takes the
matching placebo. In reality, we know
it’s not that easy. Depending on the type
of trial and the study design that you conduct. So, this is where
effectiveness trials come in. And again, it’s asking
the question, does that treatment, or dietary
supplement in this case, or herbal, whatever
you’re testing, help under real world
circumstances? So now it’s trying to
introduce kind of the reality of conducting
clinical trials whether in a smaller setting, or
in a much wider setting. You’ve got the usual
elements of patient care. You’ve differential
access. Some subjects might not
take their assignment. We want compliance to
be a 100 percent but we can’t necessarily
guarantee that in the real world. It’s no different
than antihypertensive medications, and other
types of prescription medications that we give
out in the doctor’s office. People might take
them, they might not. The key is to try and see
if we can understand what that compliance is. Patients might also
drop out of the study. Now of course, this is
the function of the type of trial. If you’re doing a trial
that might be one or two weeks long, it’s a very
carefully controlled environment. That might be
less of an issue. But for a larger scale
prevention trials, this becomes a greater issue
with lots to follow-up. The other thing is that,
subjects are pretty creative. You’d be surprised how
many ways that they can find the treatment that
they were not assigned and take that instead. Many times, especially
in trials with dietary supplements, if it’s
say a vitamin D trial, someone feels very
strongly about taking vitamin D, they become
uncomfortable with the idea of being randomly
allocated to either take a vitamin D supplement
or a placebo. They might just start
going to the store and buying a 1000 IU’s to
take daily anyway. And then you have
differential access to healthcare. So, whether it’s the
doctors that you see at the facilities
that you go to. And so, the outcomes that
you might be interested in in a trial are harder
to detect when you have issues of effectiveness. Sometimes what’s called
an MI in one part of the world could be different
from another part of the world, or by a different
treating position. But ideally in an
effectiveness trial, you want it to strive those
results as most patients would experience them. So, there’s been a —
many people actually define what we call an
implementation gap. And this is the
difference in the results that you would see, had
you conducted trial under the optimal circumstances
of efficacy versus the real world of
effectiveness. Now, it’s also kind of
the gap between ideal care and ordinary care. But I think the hard part
with this is that you really never know exactly
what that implementation gap consists of. Because it’s very
difficult in most circumstances to create
the “perfect trial.” Right? Complete 100 percent
compliance, no loss to follow up. No issues with pill
taking or whatever. Typically, you start with
the efficacy trials. You try and create the
ideal circumstance, the very protected fish
bowl, and then try and translate that to a
broader population in more usual care or
effectiveness trials. And the assumption is
that, if a dietary supplement under the best
circumstances, has no impact on your outcome,
then the likelihood of that extending into an
effectiveness trial is probably quiet low. In other words, if
you control all the circumstances of that
trial, and you can’t see in affect, or you see a
lack of effect, to be able to suggest that
you can lower your compliance, lower loss of
follow up, expect to see that same effect, is
actually quite low. At the same time,
sometimes you will start with the
effectiveness trials. And when those are done
first, and they showed no effect, you can argue
that the lack of effect might be due to the fact
that you didn’t have the ideal circumstances. That maybe the lack
of effect and that effectiveness trial might
be attributable to the fact that compliance was
low, or loss of follow-up was high, or other
circumstances that start to cut away at the
potential internal validity of that
clinical trial. So, it’s often seen as a
bit of an excuse, which I don’t think is
necessarily well warranted. But to me, if you see
poor compliance in a trial in the first place,
in what would be an effectiveness trial,
that to me is usually a signal. You’ve to ask yourself
right, why was compliance low? Was it something inherent
in the study design? Maybe the study was just
too long and it was just difficult to maintain it? But maybe there’s
different — there’s side effects or other adverse
effects that are associated with that
supplement, or other medical events, other
intermediary outcomes are happening along the way. So, what’s nice about
dietary supplements is that in the context
of clinical research, epidemiological research
and in trials, they’re really well suited, for
the evaluation of both efficacy and or
effectiveness. Most, not all, but most
dietary supplements are very much amenable
to placebo control. See if they have a very
clean comparison between treatment, and control,
or active — or a placebo. Excuse me. Most dietary
interventions, especially with supplements tend
to be fairly simple. You’re either taking a
pill, or a capsule, or a tablet; versus something
that’s the same but not the same, it doesn’t have
the active ingredients. So, it’s a very straight
forward comparison. This is especially
important in the context of dietary modification. There’ve been a lot of
really important trials that have been done, that
involved, you know, could be adhering to different
types of dietary interventions, a
Mediterranean diet, you name it. Those are very
difficult to conduct. Given the fact that
you’re not just changing things on the basis
of a pill, but rather behavior. And it just involves
a lot more effort. Not that it doesn’t
undermine — or still — these trials are still
truly important, but they’re just
harder to conduct. And the other thing
that’s nice about dietary supplements, is that they
are very much conducive to large scale trials. You don’t have the same
type of potential adverse effects that we believe
are associated with those dietary supplements. So, they’re much more
conducive to large scale trials. Again, it’s hard to
achieve ideal efficacy. Even in the dietary
supplement trials that have been conducted. Compliance will never
be a 100 percent. You want it to be close,
but it’ll never be a 100. There’ll always be some
degree of loss to follow up. And it’s always a bit of
a slippery slope between efficacy and
effectiveness. There’s really no single
demarcation line based on different specs of a
trial, whether it’s compliance, or loss to
follow up, or adverse event. There’s no way to clearly
distinguish between the two. But I think it’s
important to understand at what point do you
think you might hit a threshold beyond which,
you can go beyond effectiveness and argue
that you see efficacy for a particular
dietary supplement. There’s another
consideration that’s important to think about
when we — when we are trying to determine
whether or not something might be a trial that’s
determining efficacy versus effectiveness. And these are data from
the Physician’s Health Study, which is a trial
that was initiated back in the 1980’s, they’re
now a part of, in terms of a second iteration
of this trial. What this simple diagram
captures is, just because you’ve randomized 22,000
people into a trial, the representation of those
people to the general population, no matter
whether you’re testing it in physicians or even
a trial in a general population. Also affects the degree
to which you might be able to determine whether
something is efficacious versus effective. So, even though we’ve
randomized 22,000 people into the trial, we had
to start by mailing out questionnaires to more
than a quarter of a million physicians around
the country to even get them to the point where
they can be randomized then. So, by the time people
decided they even want to return the questionnaire
that we sent, they’re willing to be in the
trial, they’re eligible based on the criteria
that we have. They get through a run-in
phase where they can demonstrate consistent
pill taking, and then they’re randomized. This is not necessarily
representative of the general population, so
it’s also difficult in clinical trials. It’s important to keep in
mind the volunteer bias, or the selection that —
that comes into play here too. And in fact, this
selection has an impact on outcomes that you
might be interested in. So, if you are to look
at either, people — and this is a, kind of a
half-hearted joke to be often say when we
do these trials. Is that, we want to
encourage people to actually respond to the
questionnaire’s that we send, because merely
responding to the initial enrollment questionnaire,
regardless of eligibility, your
mortality rates are lower. [laughter] Whether you’re willing to
be in the trial you’re going to have [laughs]
lower mortality rates. It’s not just for total
mortality, it’s for cardiovascular and cancer
mortality as well. And not surprisingly by
the time you get to those who are randomized, and
you see the absolute age adjusted mortality rates
here, at the very right of this graph, very, very
different, and much lower than where
people started. So, it’s another
consideration I think, if not just in terms of the
design of the trials, but also in thinking
about efficacy versus effectiveness. This is looking at it
from a slightly different angle. So, this is a
meta-analysis that was — that was published in
BMJ, earlier this year, looking at, I think
it was behavioral interventions
and diabetes. And this is again,
getting at the same idea from a different angle. This is when they were
combining the data, these are all trials that were
focused on pre-diabetic populations. So obviously those who
— you have the general population, but you get
it in smaller and smaller circles. Smaller slice of that
pie by the time you get people who are either
randomized or who actually complete
the trial. So, when you’re
conducting these trials, whether they’re
particularly large or particularly small. They’re all important. But the way to which you
can generalize those findings in the context
of that potential efficacy, is something
that also needs to be carefully considered. So, study design also
comes into play. You’d be very surprised,
particularly in dietary supplement research, to
the extent that the study design will impact how
people will interpret something that might
be seen as efficacy — efficacious vs effective. This is not to belittle
the importance of any of these studies. When we think about
particular dietary supplements we need a
mechanism, we need to be able to demonstrate
everything from the entire range of studies. But by and large, we
usually would like to assume that clinical
trials are at least towards the top of that
heap, if it’s well conducted in terms of
determining not just validity internally,
hopefully with some external validity, but
also, hoping to identify some degree of efficacy
with that dietary supplement. So, shifting
gears, you know. One of the other
challenges that we see with dietary supplement
research, is also trying to understand how the
dietary factor gets translated to
the supplement. And for no particular
reason, I’ve decided to pick on a strawberry,
literally. If you think about it. So, we — if there are
studies that come out, that suggest that
strawberries might be beneficial in various
health outcomes, we always are asking
different questions. It could be the food. So, maybe it’s something
about the food matrix. You know for the whole
food people out there, that’s the argument
that they would take. The other argument, what
could be — that it could be a function of
dietary patterns. People who eat
strawberries are eating other components of a
heart healthy diet, and that might explain
those benefits. It might be that there’s
something specific to the strawberry. Maybe it’s the folic acid
content, maybe it’s the fiber content. Or it could be simply
confounding, that there could be other factors
that are related to strawberry intake that
could otherwise explain those effects. And without wanting to
completely rehash some of these points, as I know
that a lot of this came out from the discussions
from yesterday. You know, dietary
supplements, it’s been a very interesting history,
if you see it not just in terms of the history, but
also looking at store shelves over time. They’re occupying greater
and greater chunks of our store shelves each year. And we know that the
research had really kind of started to put the
supplements onto the shelves, back
in the 1970’s. And the question has
always been whether the supplements provide
equivalent benefits as the food source. And these data start from
the basic science of observational studies. But these don’t really
have bearing on clinical efficacy or
effectiveness. And so, it’s really the
clinical trials that have ultimately been the judge
of whether a lot of these dietary supplements,
do or do not have real effects on a lot of the
health outcomes that have been of interest. And just as one
example, The U.S. Preventative Services
task force put out a statement back in 2014. Where they are — and
this is a common issue, usually they’ll say the
evidence is insufficient. This was in particular,
with regard to multi-vitamin,
multi-mineral use. They came to the
conclusion that, “current evidence is insufficient
on the basis particularly of clinical trials.” Lots
of people take these supplements, but we
have very few trials to necessarily see whether
these potential benefits or risks, might be
real effects or not. So again, you know, many,
most people are taking them. They might help specific
nutritional deficiencies, or insufficiencies. But they’re still are
perceived unproven benefits. And whether that nutrient
in the food, is the same as the nutrient in the
supplement, is always an ongoing debate. Eventually these
nutrients are encapsulated, but this is
where the clinical trials start to become helpful. And, we always talk about
the high rates of dietary supplement use in
the United States. But it also extends
out globally. It’s kind of hard to read
this at this hour in the morning. But the bottom line is,
and this is a review that a postdoc of mine had
conducted last year. It nicely summarizes
the fact that, “while certainly the supplement
use rates are highest in the United States,
they’re actually still quiet high in other parts
of the world as well.” So, this has not just
national implications, but global
implications too. So, what typically
happens when we do research for a
dietary supplement? We always think that
there are potential health benefits. And we start with
these, specially these small-scale trials that
start to show potential reductions between
dietary supplements, related biomarkers,
remediators of chronic disease. So, kind of giving it the
mechanisms of effect. So, it could be oxidative
stress, inflammation, hyperglycemia, insulin
resistance, you name it. Insert your favorite
mechanism of effect there. Eventually that starts
to lead to some of the larger perspective cohort
studies that will try and look to see whether
dietary supplement use is also linked, not just
with the immediate intermediate outcomes,
but also looking at some of the harder
clinical outcomes. And that could be type II
diabetes, hypertension, hypercholesterolemia,
some of the site-specific cancers, and other
outcomes of interest. But the hard part has
always been whether these observational studies,
really do demonstrate an efficacious effect. You know, are these
observed findings real or not? We know that residual
confounding — so explanations by other
factors come into play. The small-scale trials
are certainly instructed, but those are usually
getting at mechanism rather than looking at
real outcomes that help to determine efficacy
or effectiveness. We also know that
people who take dietary supplements are
inherently different. In ways that we don’t
still fully comprehend. We know they
tend to be older. They’re women. They tend to have a
healthier diet and have other beneficial
lifestyle patterns broadly. They’ll exercise more. And in some instances,
they’ll tend to be diagnosed with
conditions. So sometimes dietary
supplements use can be a bit reactionary. “I just went to the
doctor; my blood pressure was a little high. I was called
pre-hypertensive. I don’t want to start
taking medications. Maybe I’ll take some
supplements that I believe that might reduce
my blood pressure and make some other changes.” And just because the
cohort study is large, and this is kind of
pointing to the Nurse’s Health Studies, and the
Health Professionals, some of the other larger
cohorts that are out there. The size of the study
is nice, but, but that doesn’t necessarily
denote greater efficacy when you’re trying to
evaluate whether that dietary supplement has
a real effect or not. And just to underscore
some of the — I wouldn’t quite call it the healthy
user effect, because it’s not necessarily a
healthy part of it. It’s more of a user
effect when we think about supplement use. These are data from the
Physician’s Health Study where they had
reported use of either multi-vitamins,
multi-vitamins and other supplements, or other
supplements only. And you can see that
there are distinct patterns. Whether it’s, you know,
older men tend to use supplements more
frequently. They tend to be more frequently exercising vigorously. They’re using Aspirin. But they’ve also
had a diagnosis of hypercholesterolemia. To a certain extent,
hypertension as well. But they’re also eating
their fruits and vegetables. So, it’s a — it’s a bit
of a mixed bag when you think about the impact of
residual confounding and who uses supplements,
when you’re trying to interpret the findings
and the literature. Particularly from
observational studies, and the smaller
cross-sectional studies, on dietary supplements
and health. And this is just as an
example of some of the one of the challenges of
trying to interpret this. These are data from the
Physician’s Health Study, observational data. In which we looked at the
duration of multi-vitamin use and risk of
cardiovascular disease. So, you know, it’s a
fairly — looks like a very encouraging finding. It suggests that those
men who were taking a multi-vitamin for more
than 20 years, which is a long time, would suggest
to have reductions, fairly consistently. And not just major
cardiovascular events, but also some of the
individual components of cardiovascular disease. The problem with this
is that, this is an observational study. So, those men, that are
taking the multi-vitamin supplement for longer
period of time, are going to be fundamentally
different from those who are. We try to adjust for
these things in the analysis, but in an
observational study, you can’t quite cut out
all that residual confounding. So, this is again where
the trials become particularly important. Another example. Let’s look at magnesium
intake and type II diabetes. This is a meta-analysis
that was done. So, if you look at
the farthest plot. These are based on
observational studies or perspective
cohort studies. I guess pretty strong
compelling evidence to suggest that those who
consume greater amounts of magnesium intake,
might lower their risk of type II diabetes. But again, while
promising, certainly should lead to
more research. Not going to deny that. But yeah, at the same
time, it’s important to think about how residual
confounding might otherwise be cutting into
these findings, and to what extent. Then extending us into
— looking at vitamin D supplementation. If you’re to look at
circulating-25 hydroxy D levels, combining
results from different observational studies,
looking at cardiovascular disease, same type
of question arises. Can these observational
data rise to the level of either efficacy or
effectiveness for vitamin D supplementation? I would argue that
for the most part, it doesn’t. And it doesn’t mean that
these are not important studies, but these are
really meant to be hypothesis generating,
through which we can use the clinical trials to
get at the real answers. So, what are the typical
dietary supplement trials? They might be a
relatively short duration of treatment
or follow-up. You’re comparing versus
placebo, it might be, as an example,
post-menopausal women, with pre-hypertension. You look at some
different outcomes. Now is this optimal to
determine efficacy or effectiveness? Probably not. But, if it’s
well-conducted, you can start to get
toward that end. And it’s always important
when you conduct these trials, either as a
trialist or interpreting it, in the literature, to
ask to what extent we can take those conclusions
from these smaller trials and extend them into some
degree of effectiveness or efficacy. They tend to have
good methodology. You’ll tend to have
pretty good compliance, right? If you were told to take
something for eight weeks, that’s a different
prospect, then being told to take a supplement
for eight years. It really helps to get at
the mechanism of effect. But you don’t know how
reproducible it might be, and how generalizable
it might be. And the lack of hard
endpoints becomes problematic ultimately. Oops. And that’s where the
large-scale trials, albeit the challenges
of conducting them and getting the funding,
and trying to get it coordinated still have
a very important role, particularly in the
context of dietary supplements and efficacy. And you know, one thing
that’s often missed out, because I think a lot of
the — one would look at a lot of the large-scale
dietary supplements and argue that a lot of them
had either find a lack of effect or even a
potential risk and some beneficial effects
in there too. All the results from
these trial tends to be informative, and they
drive public health and policy. A lack of — a null
effect, yes that means a lack of efficacy but that
tells people that there’s no reason there’s
necessarily be taking that supplement. If you see benefit, then
there might be potential efficacy. I mean that people should
at least consider taking that supplement. It doesn’t mean that
everyone should universally start taking
it, but at least we know what we’re seeing greater
evidence heading toward, that kind of, that
zenith of efficacy. Whatever you want
to find that as. [laughs] Now if there’s harm, we
see a lack of efficacy that we now know, that
people should probably stop taking that
supplement. And it’s always important
to consider the balance of risks and benefits
with these trials. They tend not to just
look at one endpoint but many different endpoints. And despite what we see
from these large-scale trials, people still tend
to take the supplements. There’s been a lot of
trials that looked at beta-carotene, and
vitamin C, and vitamin E, and others that have
largely found a lack of effect. But people will still
tend to take them, because there’s a belief
that — oops — too much of a hand waiver. It — there’s a belief
that it couldn’t hurt. And, based on the store
shelves, knowing that they’re just growing with
more and more dietary supplements, and herbals,
and other products, from week to week literally,
we know that there’ll always be new supplements
that will be coming out that will be emerging. That we need to make sure
we test in some way shape or form. But I would argue that
the clinical design, is probably more important
than the actual outcome that you get
from that trial. Because regardless of
what you test, if you don’t conduct that trial
well, and have the best methodology that you can
put into it, it’s going to undermine, the extent
to which you can make any claims for efficacy
or effectiveness. And, I just wanted to
walk through a few examples of some of the
type of trials that I’ve been involved
with up in Boston. One is the Physician’s
Health Study II. And this was a — a large
simple, I wouldn’t quite say simple, but it’s
simple in design and conceptually, a
randomized clinical trial. In which, it was
conducted entirely by mail. And this is one thing
that’s nice about dietary supplements. Is because you don’t
have the types of restrictions, that
you might have with prescription medications
per se, you can do these mail-based trials of
dietary supplements, not just very, in a large
scale, but also fairly cost-effectively as well. In addition to collect —
doing the trial, we also were able to collect
bloods, in about 70 percent of the
participants in this trial. We had nearly 15,000 men
that were age — mail physicians, age 15 up. And we tested a very
common multi-vitamin at the time, in the mid to
late 1990’s, Centrum Silver. We also in this trial
had tested, vitamin C, vitamin E, and
Beta-Carotene. I’m just going to focus
on the multi-vitamin results for today. Our primary end points —
basically it’s a larger, larger trial,
longer trial. Clinical endpoints:
so, we looked at cardiovascular
disease and cancer. And we had secondary
endpoints that included eye disease and
cognitive function. Now, compliance
was not perfect. It was still decent. I would argue that 67
percent compliance to take a multi-vitamin
daily for 11 years is actually pretty
darn good. But it still raises the
question of whether this is an efficacy trial or a
trial of effectiveness. Depends on where you
want to draw that line. Now, when we had — after
all this time when we looked at the results,
we actually found that taking a daily
multi-vitamin had no effect on one of the
primary outcomes which is major cardiovascular
events. If you looked at some of
the other components of cardiovascular disease,
we saw a similar lack of effect. One thing you always have
to be careful of, is that when you see these
individual things like, MI depth or a suggestion
of a benefit. It’s a suggestion that
when you see that in the context of all the other
results, it was something that we thought was
interesting but certainly not anything that we’re
going to be hanging our hats on. We also had looked at the
effect modifications. What’s nice about these
clinical trials is that, especially the larger
ones, is that you can start to slice the data
carefully, and in a pre-specified manner, to
see whether particular subgroups might benefit
more from taking a particular dietary
supplement versus others. And you have to be
careful about this. You know, in certain
respects, to suggest that there’s a one size fit
all — fits all capacity of a dietary supplement
to work in everyone in this room for example,
is a bit of a misnomer. We know that — and this
where this notion of personalized medicine,
personalized nutrition, how it’s going to be
defined over in the years to come we’ll see. We know that the results
tend to be a bit more nuanced. And if we looked at the
role of baseline age in the Physicians
Health Study II. What we saw is that
there are at least some suggestions that older
men might benefit a little bit more, although
it wasn’t a very strong effect. But in particular, what
we had seen is that, if you look at the cancer
endpoint, there was a potential, a very modest
but significant 8 percent reduction in the other
co-primary endpoint of total cancer. If you kicked out
prostate cancer, there was a lot of prostate
cancer in the Physicians Health Study II, because
these were physicians going through the 1990’s
and 2000’s when PSA screening was going
through the roof. In about half of the
cancer in the Physicians Health Study II was very
early stage prostate cancer. So, if you focus more on
the true epithelial cell cancers, you actually saw
an even stronger effect. What’s interesting is
that, if you looked in the older men, there was
an 18 percent reduction among those men ages 70
and up for the primary endpoint of total cancer. And one slide I don’t
have here is that, if you actually isolated, there
were about 1500 — wrong number — about 1400,
1500 men or so, who came into the Physicians
Health Study II where they — coming into the
baseline of history of cancer. In those men, we actually
saw a 27 percent reduction in subsequent
cancer among those men taking a multi-vitamin
versus those taking placebo. So, like any good trial,
a good trial will answer your initial question
that you have but will come up with multiple
questions after that you still want to
try and pursue. Another trial that’s
currently going on in our research group is
the VITAL trial. The vitamin D and
Omega 3 trial. So, this is testing
vitamin D supplementation at 2000 IU’s daily versus
placebo, and also looking at Omega 3 fatty acids,
essentially fish oil. It’s about a gram a day
with a mix of EPA and DHA. So, this trial, got
started about six or seven years ago now. And is actually coming
to a close at the end of this year. So, the randomized
treatment and follow up will be about five to six
years on average, will end at the end
of this year. And our hope is that
sometime next year we’ll be coming out with
those study results. But again, vitamin D is a
nice example of a dietary supplement that’s being
used with an increasing number of people
in the population. I feel like there is like
a study a week that comes out, that’s praising its
potential virtues, but yet there is very little
trial evidence that has supported it. Especially from the
larger-scale trials. And so VITAL will be an
important step forward to try and have a better
understanding of what the role is of vitamin D as
well as fish oil, on not just the primary
endpoints of cardiovascular
disease and cancer. But VITAL is a nice
example of a clinical trial that built in a lot
of different ancillary studies, looking at other
major endpoints, as well as intermediate endpoints
to try and delve further into the types of
mechanisms that might be in play. And then finally, based
on some of the results that we had seen, from
the Physicians Health Study II, we recently
initiated what’s called the COSMOS trial. And as you can imagine
with these, you always have to have a acronym
that sounds kind of cool. You can’t just call it
the blah trial, because no one’s going to be
interested in that. [laughter] But, what’s nice about
COSMOS is that we had seen this potential
effect of a multi-vitamin particularly for cancer
endpoints in the Physicians Health Study
II, to which we extended it to this new trial
called cosmos. We are in the midst
of recruiting for it. We have already
randomized about 11,000 people. We’ll eventually get to
18,000 people by the end of this year. And they’ll be either
taking a multi-vitamin or a placebo. Then we’re also testing
cocoa flavonols, which are viewed as the
bioactive component in the cocoa bean, and is
part of this whole dark chocolate craze. So, this will eventually,
in the years to come, provide some interesting
results too. But ultimately, you’re
always kind of at a catch 22. When you do these trials,
if you show favorable results, people will be
happy to highlight the strengths of the trial
design and everything you did well. No differently than if
you have unfavorable results that counter the
conventional wisdom, people will complain that
your trial is basically a load of, you know what. You didn’t do it right,
the design was wrong, and there were
inherent problems. So, doing these trials,
there’s a bit of a leap of faith. You have to assume that
these trials are going to at least represent steps
forward, it’s not the final word. And finally, you know,
when we think about dietary supplements, we
need trials of all sizes and shapes. Not just the large ones,
but we also need the small ones. And trying to patch these
together to come up with a cohesive understanding
of exactly whether or not dietary supplements does
or does not have any degree of efficacy
or effectiveness. And you know, I think
that there’ll always be new supplements that
are coming out that are viewed as promising, so
there’ll be no lack of opportunities hopefully
for new research projects to get involved in. But it’s also to think
about that methodology. Who is the representative
population? Do you have a way to
recruit effectively? And do you have an
intervention that you believe in? As well as intermediate
outcomes to help to capture the mechanism of
effect that might be in place. And while the large-scale
trials are certainly a big undertaking,
requiring a lot of coordination across all
sectors, they’re hard to initiate but they’re
still critically needed. And we’ll see in the
years to come what these trials will
ultimately show. So, thanks for your
attention and I’d be happy to take any
questions if we have time for it. [applause]>>Female Speaker: We
have time for a few questions.>>Howard Sesso: Yes?>>Female Speaker:
Great talk. Thank you. What about the
nocebo effect? People getting
harmed from placebo. Do you see that?>>Howard Sesso: Yeah,
it’s an interesting question. I’d like to say that the
placebo effect is real. It’s one thing that we
have not fully utilized our trials to see to what
extent the placebo effect takes hold. What’s interesting is
that, at least in the trials that we’ve done,
we’ve always had, what we called placebo run-in. Where people start to get
pills and we use a three to six-month placebo
run-in to ensure that the people who ultimately get
randomized are people who are compliant. A lot of times you can
start to take pills and then you realize
immediately, “I don’t want to do this,” or
“it’s too involved.” But what’s interesting about
it, and something that we haven’t really delved
into, is that, everyone’s taking placebo. So, we in essence have
a mini-trial testing a placebo in everyone. And you’d be surprised
what types of feedback we get from participants
claiming that “gosh I started taking these
pills, just in two weeks, and my skin is clear, and
I have more energy.” And so, it’s a real effect. We haven’t quite captured
it in our trial data yet. It’s something that
we would certainly be interested in doing.>>Female Speaker: Can I
just ask a follow up on that? Since you’re involved
with cardiovascular disease. I have been reading
about Statins and CoQ10. And just — I’m sorry,
it’s not coming to mind but, I know, do you
have a comment on that?>>Howard Sesso: Yeah. You know CoQ10 is a
supplement that a lot of cardiologists and
physicians will swear by for, kind of serving as
an alternative to Statin use. But the problem with
CoQ10, which is the case with many other dietary
supplements out in the market, is that there’ve
been a lot of very small focused trials that have
looked at the effects of CoQ10, but nothing that’s
really even come close in my view to coming a
point where we’re making conclusions about
true efficacy or effectiveness. It’s not really
happening. There’s really no long
term, larger scale trials. I’m not saying it has to
be 10 years in duration, but there’s been very few
definitive trials on that product. As well as potential
interactions and side effects might be with it. It’s a very expensive
supplement too. Yeah?>>Male Speaker: Our
center has some people with experience around,
but we might try a combination of a couple
dietary supplements for a condition. Is it common practice to
start with a case-series or case control study
before attempting a small-scale trial, or you
can — you go for it with the right mentorship?>>Howard Sesso: I would
always say; any data is better than no data. So, if you’re going to be
seeing, you know whether it’s through a case
series, or even a very small — you know, any
data is better than where we were before
particularly. And you know, I
think it’s specially challenging, given the
fact that no one tends to take just one supplement. People tend to take a
supplement in combination with other supplements. Now that makes it harder
for us to study it, whether it’s in a case
series, or in a small clinical trial,
or a larger one. Because you are trying
to isolate the effects, ideally of a
particular supplement. But to me I think,
there’s a general lack of data on dietary
supplement use and its health effects. Or even its mechanisms. So, I would certainly
encourage those types of work to be done.>>Male Speaker:
Great thank you.>>Howard Sesso: Yeah?>>Female Speaker: How
do you pick your doses?>>Howard Sesso: It’s a
really tricky process. You’d be surprised how
challenging it can be. You know just as an
example, for the VITAL trial, we ultimately
wound up at 2000 IU’s, but in truth that dose or
amount actually kind of crept up over time in
the design phase of it. There’s no perfect
answer to it. You can argue that if
you’re doing a vitamin D trial, the perfect world
will be one in which you test what people
are commonly using. So, there might be a
1000 IU’s, 2000 IU’s and perhaps three
or 4000 IU’s. Although that starts to
get close to the upper limits that the Institute
of Medicine that would necessarily recommend. The problem with that
type of an approach is that, it requires, more
people, longer follow-up, more outcomes. So ultimately you have to
kind of take a leap of faith at a certain amount
of a particular nutrient and go for it. Hoping that the science
won’t change tremendously by the time
you publish it. You know we could start
with a 1000 IU’s daily. Five years later,
everyone’s taking 2000 IU’s. You have a trial result
that’s interesting, but doesn’t really reflect
current clinical practice or practice. In the back?>>Female Speaker: In the
VITAL trial how are you planning to control for
geographic differences?>>Howard Sesso:
Ver good question. We know that obviously D
is a sunshine vitamin for a reason. Lower latitude, the
better off you are in terms of these getting
potential sun exposure. What’s nice about the
VITAL trial is when it’s a randomized clinical
trial, randomization inherently will control
for geography and latitude. So, if you’re to look
at the VITAL trial participants, and
separate it by latitude. Because we’re randomly
choosing who in the room is going to be taking the
active or the placebo, it’s going to be equally
— it is equally distributed,
I should say. [laughs] That being said,
what’s an interesting subgroup analysis that
ought to be considered, is one in which you look
among those people who are at higher latitudes,
whether it’s over the 37th parallel or not. And compare to people
who’re at lower latitudes to see the
supplementation might be perhaps more beneficial
among those who at least have a greater
opportunity for sun exposure or not. But of course,
it’s more factors. It’s not just whether
you’re living in Florida, whether you’re actually
outside playing tennis and walking and doing
other things too.>>Female Speaker:
Last question.>>Howard Sesso: Yeah.>>Female Speaker: How do
you measure compliance in these trials? The mail trials?>>Howard Sesso: Yeah. Good question. It’s a bit of a
leap of faith. So, because they’re mail
based trials, we’re not seeing patients directly. We’re sending them
calendar packs that they receive and then they
get questionnaires periodically sent to them
that will ask them to report, approximately,
how many pills on a percentage basis were, or
on a numerical basis are you taking each month. So, it’s not as good as
doing a smaller clinical trial, where you have
some of the pill bottles that you can open and close, and the chip counts it. But that still is not
the perfect measure of compliance either. So, it’s based
on self-report. We do compliment it. For example, in the VITAL
trial and the other trials that we do, we do
collect bloods both at baseline and follow up in
VITAL, as well as in the COSMOS trial. We can track, some of the
key biomarkers, and other markers of compliance
that we believe ought to be linked with taking
the active versus the placebo. So that’s a second way
— second layer through which we can verify
the compliance that they’re self-reporting. But we do believe that
it tends to be good. Not perfect
though, sorry.>>Female Speaker: Please
try and show appreciation to the speaker. [applause]

Posts Tagged with…

Write a Comment

Your email address will not be published. Required fields are marked *