The World Turned Upside Down: The Second Low-Carbohydrate Revolution (29 page)

PART
II - POLICY AND THE
MESS IN NUTRITION

Chapter
16

The
medical
literature. A guide to Flawed Studies

Doctor
:
Therein the patient Must minister to himself.

Macbeth
:
Throw physic [medicine] to the dogs; I'll none of it.
– William Shakespeare,
Macbeth

Nutrition in crisis. Almost every day
"a new study" shows
that you are at risk for diabetes or cardiovascular disease or
all-cause
mortality brought on by a newly appreciated toxin which turns out to be
something that you just had for lunch. It is not clear that any of
these studies
were subject to any kind of serious critical peer review and, for the
curious,
they are frequently dismembered by the bloggers. The continuous cycle
of weak
studies and their deconstruction goes beyond time wasting. People are
hurt
because bad recommendations are left out there even when research shows
that
they are inappropriate. And science takes a big hit. Peer review by
technical
and medical journals is supposed to be the gate-keeper on scientific
evidence
but papers showing very weak associations or even ones that are grossly
misleading are accepted. And the media, which might be expected to help
us,
makes it worse. Its not really their fault. A science reporter could
not
reasonably have the time to read the original in detail and must accept
the
conclusions in the abstract and so the message is transmitted through
mass
media. And when you do explain to the reporter how misleading these
reports are
and how people will be hurt, they are truly concerned and sympathetic
but they
don't always have complete editorial control. In any case, they would
like to
help but tomorrow they have to cover a story that may be even worse. It
is
really hard for the consumer. This post from FaceBook probably tells
the story:

So epically confused about diet.
Everything I read is
contradictory on epic proportions. About the only consistencies are
low-sugar
raw veggies and water. How in the world is a girl to sort it out, other
than
try everything and see what works for me?

And I hesitate to even ask, as diet
has become as
controversial as religion and politics - but wouldn't you think that
all of
this would be easily testable and provable, in a way that religion and
political opinions are not? People are different, but shouldn't it be
possible
to come up with a system that takes inputs of body stats and genetic
history,
and outputs a general reasonable diet to follow? Any insight on getting
clarity
here?

It is likely that the population at
large is not any more
comfortable. I wrote to her off-list and re-iterated the
first
three rules
: 1. If you are okay, you are
okay, 2. If you want to lose weight don't eat; if you have to eat,
don't eat
carbs, if you have to eat carbs, eat low glycemic index carbs, and 3.
If you
have diabetes or metabolic syndrome, you have to try a low carb diet
first.

It is likely that many people wind up
believing nothing or
believing that everything is exaggerated although I think that there is
a
popular notion that "maybe fat is bad and maybe I should not put so
much salt
on my food." And then there is the progression of raspberry-ketones,
resveratrol and
trans
-fats
and methylglyoxal, each of which will either kill you or save you from
being
killed by the others.

Most discouraging are the health
agencies. The American Diabetes
Association (ADA) wants people with diabetes to have a lot of
carbohydrates.
They keep saying that they don't have a diet and, that they're not
opposed to
low-carb diets (for weight loss) and, as described above, they
continually
stress "individualization" but there are no indications as to which
individuals
benefit from which intervention. What criteria for what diet? Despite
the
disclaimers, it is for sure that the ADA is
perceived
as opposing low-carbohydrate.  And it seems clear they are
responsible for that
perception. The evidence that weight loss is not required for
improvement in
diabetes, from the work of Nuttal and Gannon
[78]
,
for
example, is not mentioned. They know about that evidence. I know
because I have
told members of the committee personally. The ADA guidelines do not
cite
important scientific work showing that weight loss is not required for
improvement in diabetes. People who are not scientists ask me: "Can you
do
that? Are you allowed not to cite other people's work?"

The ADA websites are clear on the
necessity for including
carbohydrate in your diet. All of the contradictory evidence does not
seem to
make any real impact.
Figure 16-1
,
from
The
Food Navigator
, the
food industry house organ, says it all: "Meanwhile, a growing body of
research
has suggested that replacing fat with carbohydrates could increase the
risk of
heart disease but consumers are still focused on low-fat food and
beverage
products...."

There is a daily progression of
sweeping statements that go
way beyond the published data. There what is perceived as an inability
or one
doesn
'
t
know the motivation
--
an unwillingness to zero in on real factors and, again, the unwritten
rule to
avoid mentioning the value of carbohydrate restriction.

The literature, especially major
medical publications are
still subscription based, that is, not directly accessible to many who
are
interested. The results are then fed downstream to the media, who take
at face
value anything it is fed and will pass it on to the general public.

Rigid dogma has reached Galilean
proportions. Fructose and
sugar are bad (unless you try to lump them in with all carbohydrates).
If you
want your paper on fructose to be published, begin with: "Because of
the
deleterious effects of dietary fructose, we hypothesized that..." Never
start
with "Whether dietary fructose has a deleterious effect..." I know. Our
paper on
fructose (
[79]
was published
with "whether.." as the
opening sentence but only after a rebuttal of reviewers' criticisms
that turned
out to be 15 pages long. And if you even mention low-carbohydrate, you
are
guaranteed real grief. When JAMA published George Bray's
"calorie-is-a-calorie"
paper
[80]
and I pointed
out that the study more
accurately supported the importance of carbohydrate as a controlling
variable,
the editor refused to publish my letter.  In this, the blogs
have
performed a valuable service in providing an alternative POV but if
unreliability is a problem in the scientific literature, that problem
is
multiplied in internet sources. In the end, the consumer may feel that
they are
pretty much on their own.

Figure 16-1
.
The
Food-Navigator report on trends in 2012.

It does take some confidence,
especially for the lay person,
to feel that their intuitive understanding that the difference between
white
rice and brown rice is so small that it really doesn't matter what
Harvard's
computer says. Most researchers are very much disinclined to get into a
shooting match, or worse, whistle-blowing. The long blue line is a
strong force
in repressing investigation, not because the police think corruption is
okay,
but because scandal reflects badly on everybody. Whistle-blowing in
this field
is especially weird because sometimes the transgressions are right out
in the
open. Figures are published showing marginal associations and the text
says
that they are significant and therefore you should change your diet
accordingly. Contradiction is right out there in front of you. No
Wikileaks
needed.

Statistics: death of
the medical literature.

Many scientists believe that if you
do a good experiment,
you don't need statistics. David Colquohon, a well known neuroscientist
and
critic of poor scientific method, agrees. Colquohon is the author of an
excellent, if technical, statistics book (free download from his
website
DC's Improbable
Science
) and the introduction to his book points
out:

"the snag, of course, is that doing
good experiments is
difficult. Most people need all the help they can get to prevent them
making
fools of themselves by claiming that their favourite theory is
substantiated by
observations that do nothing of the sort."

This kind of circumspection is,
unfortunately, more common
among people who write the statistics books than those who use them. A
good
statistics book will have an introduction that says something like
"what we do
in statistics, is we try to put a number on our intuition." In other
words, it
is not really, by itself, a science. It is, or should be, a
tool
for the
experimenter's use and, like any tool, you have to know how to use it.
And
there is not always agreement on which tool. All statistics is
interpretation.
The problem is that many authors of papers in the medical literature
allow
statistics to become their master rather than their servant: numbers
are
plugged into a statistical program and the results are interpreted in a
cut-and-dried fashion. Statistical significance (that two sets of data
are not
from the same population) is confused with clinical significance (that
differences are sufficiently large to have a biological effect). Misuse
of
statistics is the subject of numerous papers
[81,
82]
and books
[83, 84]
but this
has had little effect.

The golden
rule of statistics (GRS).

Here's the Golden Rule for reading a
scientific paper, from
the book
PDQ
Statistics
by Norman & Streiner
[85]
:

"The important point...is that the
onus is on the author
to convey to the reader an accurate impression of what the data look
like,
using graphs or standard measures, before beginning the statistical
shenanigans. 
Any paper that doesn't do this should be viewed from the outset with
considerable suspicion."

In other words, explain things
clearly to the reader. There
are complicated ideas in science but the often quoted statement (only
once
before in this book) from Einstein that you should make it simple but
not too
simple is a reasonable demand for you the reader to make of the
scientific
literature.

Statistical shenanigans
vs Common sense

So, how do you deal with the reports
in
the press that tell you that
white rice will give you diabetes but brown rice won't? The principle
is that
biochemistry is not separate from anything else we do. It can have
subtleties
and it can rely on mathematics but it uses the same rules of logic as
daily
life. So, the first question you have to ask is whether it makes sense;
how is
it possible that white rice is so different from brown rice. A moment's
thought
suggests that the major difference is in stuff that isn't even
digested, the
fiber. On the other hand, they are always pushing fiber, whole grain
and all
that, so you might want to hear their case, but there is the
possibility that
the entire fiber thing itself is questionable or at least exaggerated.
And the
Asian societies that are always invoked to tell us how much we need
grain never
eat brown rice. Never. So, common sense makes you suspicious.

habeas corpus datorum

Science is an extension of common
sense
but there are
revolutionary ideas. In fact, the revolutionary ones are usually the
most
important. How do you deal with that? You want to be suspicious if it
violates
common sense but you can't throw out the idea just for that reason. The
solution is simple if not always easy to implement. If it is a
reasonable
conclusion, you can cut the author some slack. If the idea is far out,
you need
to see the data. All the data. Not the hazard ratio, not just the
conclusions
from the computer. My new grand principle of doing science:
habeas corpus
datorum
,
let's see the body of the data. If the conclusion is non-intuitive and
goes
against previous work or common sense, then the data must be strong and
all of
it must be clearly presented.

So, how should you read a scientific
paper? I usually want to see
the pictures first. In a scientific paper, they are called figures but
it's not
just saving a thousand words. (I get a thousand emails every week or
so). It's
about presentation of the data. It's about the GRS: "convey to the
reader an
accurate impression of what the data look like, using graphs or
standard
measures, before beginning the statistical shenanigans." In fact, it's
really
not "or." Figures are so much better than tables that a whole book
"Medical
Illuminations" has been written about the idea
[84]
.
Many of us write scientific papers the same way. We make the figures
first and
then try to explain what they say. The principle: a scientific paper is
supposed to explain. I tell graduate students that if you do an
experiment and
you don't explain it well, it is as if you had never done it. In
teaching
students how to present their work, I ask them: "Describe what you are
supposed
to do in a scientific seminar or other presentation." Sometimes they
try but I
usually make it worse by adding: "No. In one word. What are you
supposed to do?
One word." Having reached an appropriate level of annoyance they will
be
relieved to hear the answer: "teach." You want to explain things to
your
audience. The same is true of a scientific paper. Again, the GRS: "the
onus is
on the author."