From: email@example.com(Steven B. Harris)
Subject: Re: Vitamin Supplements
Date: Wed, 16 Jul 1997
In <firstname.lastname@example.org> email@example.com
(Larry DeLuca, EdM, CSCS) writes:
Here's how it works...
The requirements among individuals vary somewhat for a given
nutrient, and while one can produce an "average" of these, about
50% of the people will need some amount more than this, and 50%
will need somewhat less.
The shape produced by plotting the individual requirements is
called a bell curve with a normal distribution. Most people are
clustered very close to the mean, and there are some people not
so close to the mean, and some people are farther away from it
(in both directions).
By using a statistical technique called the standard devitation,
it's possible to figure out where large segments of the populati-
on should fall with respect to the mean. By using the second
standard deviation, about 90% of the population will lie within
that area, with about 5% way down below the mean and about 5% way
up above it.
Typically, nutritional requirements are set at the second
standard deviation above the mean, so 95% of the population will
receive adequate or more than adequate intake by taking in the
RDA alone. Thus, for most people, it's not a "minimum" at all
(since most people are much closer to the mean, and so for them
it's quite an amount in excess).
Comment: Well, I think you're being pretty optimistic about
the state of human knowledge in 1997. The above utopian
procedure would indeed work if we had perfect and agreed-on
markers for "vitamin deficiency", AND we knew what the normal
distribution looked like for human vitamin needs. Unfortunately,
we don't have ALL this information for ANY nutrient. I know of a
few studies in both humans and animals that have found coeffici-
ents of variation of about 20% in need for several vitamins,
looking at several markers of function, and that is the figure
often used by makers of RDI values. However, there's no guaran-
tee that this is true for all nutrients (it isn't true for
vitamin C in guinea pigs, for instance, where there is huge
variation in need), and there are large disagreements as to what
markers (biological actions) to select for the "mean" requirement
for any nutrient. In practice, RDAs and RDIs (the measure
accepted on a more worldwide basis) are selected with a large
amount of guesswork, looking at markers for which there is most
information (not necessarily those which are most important), and
also looking at things like body storage pools and balance
studies. And generally without having the slightest clue as to
what that bell curve of population need actually looks like, in
each individual nutrient case. Don't try to pretend otherwise.
Actually, some of the RDI numbers are pretty arbitrary, when
you get right down to it. For example, 100 IU of vitamin D a day is
enough to prevent rickets, starting at birth, and it is known
that 400 IU a day causes babies to grow faster than 100 IU. But do we
know that 400 is as good as 600 or 800? We do not. How did we pick
400, and not 500 or 600 IU for children under 6? Took a good,
educated guess. Nothing more.
And nobody looks at prevention of chronic diseases like cancer
and atherosclerosis in selecting RDIs. Is the RDI for B6 the
value which maximally lowers everybody's homocysteine levels?
No. Is the RDI for vitamin E defined as the best dose to prevent
atherosclerosis? Hardly. And so on.
>>While it is true, however, that for 5% of the population
their daily requirement may exceed the RDA, it's also important
to remember that in industrialized nations most peoples' intake
of food (and nutrients) usually far exceeds the RDA, even without
the help of vitamins.<<
Incorrect. N-HANES data shows that in the US, 30% of people
don't get the RDA of one or more nutrients, and in selected
populations (the elderly) it's more like 50%. And if we all ate
the number of calories we should to keep from being obese, those
numbers would be even worse.
>> Special populations like pregnant women and persons with
specific disease states aside, even the outliers are likely
meetin their needs through food in these nations. <<
>>In addition, peoples' vitamin requirements (like any other
metabolic variable) fluctuate somewhat, so it's not as cut-and--
dried as saying that everyone needs so many molecules of vitamin
C a day.<<
>>Deficiency diseases are easily recognized and treated.
However, when was the last time you've seen a case of scurvy (one
of the diseases mentioned in the post above)?<<
In my geriatric practice so far I've seen one case of pellagra
(a Mexican dialysis patient fond of home-made cornmeal products,
but not vitamin pills), and one case of wet beri-beri (an elderly
alcoholic with a nearby liquor store, but not a nearby grocery
store). As for the "easily recognized" part, at risk of
tooting my own horn, I don't think most
practitioners would have picked these up (and several other
doctors DID in fact miss each case above; the alcoholic came to
me overdosing on prescribed diuretics that he didn't need since
his ventricular function was fine, and the lady on dialysis was
not diagnosed by her nephrologist who she saw regularly).
However, the problem with nutrition is not the flagrant cases,
but the marginal cases. In most human experimental vitamin and
mineral deprivation studies, the group getting deficiency
developed totally non-specific symptoms like malaise, loss of
energy, loss of appetite, irritability, headache, depression,
sleep disorder, and so on, before coming down with any
pathognomonic signs of deficiency. In other words, marginal
nutrient deficiency in an astonishing number of cases looks just
like dozens of other problems an internist sees. Do you know
many internists who draw blood vitamin levels to find out what is
what, for every patient with a nonspecific complaint? Do we know
enough to say much, even if we did? No. Know many doctors who
spend a lot of time asking their patients what they eat? Me,
neither. Do you really think that most people with vitamin
deficiency (people whose health and feeling of well-being would
benefit from supplementation) are getting diagnosed by their
medical doctors? My best guess (and it's not an uneducated one)
>>Yes, there have been some changes. However, you'll find
that in many cases the "optimal" amount is just the same as the
RDA was, and the "maximal" amounts were provided to encourage
people to understand that more is not necessarily better.<<
And again, the main problem of what is the optimal amount of
each nutrient has not been addressed, because we simply don't
know enough to address it with any confidence (and we're not
doing the needed studies, either, much to our everlasting shame).
It's a fair bet that in many cases optimal intake of nutrients is
quite a lot above the RDA, however, for the average American
(consider the cases of folate in pregnant women, B6 and vitamin E
in those who have risk factors for heart disease, and selenium in
the case of cancer).
>>By and large, the intake of vitamins has never been linked
to disease prevention.<<
This is outrageously incorrect. I urge you to peruse the
NCI's last edition of "Nutrition and Cancer," just for example.
This isn't just a book about whole foods-- there is a lot of
information about vitamin and mineral intake epidemiology.
Gladys Block, for instance, has looked at dozens of epidemiologic
studies of food intake and cancer, and concluded that vitamin C
intake is the one link that most of the most preventive foods
share. There are cancer prevention studies using vitamin C in
animals also, and encouraging C supplement epidemiology studies
(people eating C rich foods live longer, people taking C supplem-
ents live longer, and animal studies giving C corroborate these).
There is better evidence that vitamin C prevents cancer than
there is that cigarettes cause it, so if you demand prospective
studies before you believe something, you ought to consider that
There *is* a nice prospective placebo-controlled study on
200 mcg selenium supplementation and cancer prevention, now, and
it nicely corroborates a very large animal and epidemiological
literature (as well as some specific prospective supplementation
trials in China). As for vitamin E, there are now both animal
supplementation trials and human epidemiology studies of supple-
ment users which suggest that large doses are anti-atherogenic.
We know also that E supplements prevent many heart attacks in
those who have severe coronary disease, from the Cambridge Anti
Oxidant Study (CAOS) trial. They work as well as aspirin, but
(oddly) are not recommended as universally for those at risk. A
little mystery, there, but perhaps explained by the existence of
diehard skeptics such as yourself.
>> Linus Pauling produced some interesting research involving
vitamin C, but that still remains controversial, and there are a
large number of studies showing no benefit from ingesting vitamin
C in this regard.<<
In regard to disease prevention? I'm sorry, but there are
NONE with regard to any important disease, if you mean
prospective trials. Just because vitamin C doesn't prevent
colds, it's foolish to think it therefore doesn't act to prevent
a worthwhile fraction of cancer or heart disease. All available
evidence points to the answer that it does, with very little
evidence pointing the other way. The case hasn't been proved
(only a prospective study in humans will do so), but there is
really good circumstantial evidence just the same.
>>Similarly, there's little scientific evidence that ingesting
vitamins has a definite anti-aging benefit.<<
Here I have to agree, while noting that the number of studies
is small. And the fact that vitamins won't slow aging (as
defined by maximum lifespan extension) doesn't mean they won't
help you live a longer and happier life (as defined by mean
lifespan extension, and mean active-life extension).
>> While it's understood that some hormones (like DHEA) drop
off as we age, it's not clear that this is the cause of aging (it
could be an unrelated effect). For example, consider a correlat-
ion between the number of speeding tickets a person receives and
the color of their car that shows that peole who drive red cars
get more speeding tickets. Did the color of the car
cause the ticket? Are drivers who like to speed more likely to
buy a red car? Do police officers tend not to pay much attention
and only notice the red cars on the road when they look up after
the radar beeps? Who knows...<<
Nobody until we do the studies (in the case of DHEA,
pregnenolone, melatonin, etc). Unfortunately, the NIH is not
doing the studies. Nobody's doing the studies. Apparently the
Feds have more interesting things to spend their money on, though
I can't imagine what. Meanwhile we know from one good study
that DHEA makes people feel good. It's sort of like exercise,
that way. So how can we recommend one and not the other? The
DHEA study, at least, was placebo controlled. <g>.
>>In all seriousness, however, we are continuing to learn more
about nutrition. In the anti-oxidant example, there are many
respected authorities who recommend taking anti-oxidants, and
many others who do not. However, the doses recommended by most
authorities are relatively small compared to what many people
And rather large compared with what a lot of other people
ingest. Since the risk and cost are small for supplemental
nutrients, I vote with the answers that the human epidemiology
and the animal intervention studies suggest. You can be conserv-
ative if you like. By the time the answers are in, however,
you're likely going to be dead (as are we all). I take those
pills in case the most conservative experts are wrong, knowing
that if they are right, I'll have lost very little (even Victor
Herbert doesn't suggest that 200 IU a day of vitamin E or 25 mg
of B6 is *dangerous*). Think of it as kind of a Pascal's wager.
I'm betting a few cents a day against many years of healthy life.
Fess up, Larry: do you take vitamin E or not? Jane Brody
does. Half of cardiologists now do, as well. What do these
people know about nutrition and disease that you don't? Are you
so conservative about the rest of what you spend your money and
time and effort on? Do you exercise? Think it will make you
live longer or be healtier? Again, do you know of any placebo
controlled prospective exercise studies (I just had to mention
this, seeing your police/red-car example)?
Steve Harris, M.D.