Shananigans ~ How Not to Respond to Being Caught Faking Something

So ooooooooo .....


I wrote a post a couple days ago calling out Dr. Cate Shanahan for her blatant misrepresentation of Ancel Keys and his part in the advocacy of a low fat diet.  One could write volumes on the problems with Shanahan's "science" and ideas, but hopefully this overnight guru phenomenon -- because some minor celebrity shill kingmaker deems them so -- is on the outs.  One can only hope.  

The first image included in that post was this one below.

Figure in Deep Nutrition
I'm not even going to be distracted and "go there" as regards where Shanahan got her information charging Keys with selecting countries based on margarine use, except to say that she's awfully fast and loose with these sorts of assertions.  It's not enough to just make things up, and while it may not be technically legal libel, it sure as heck runs afoul of the spirit of conduct in this regard.

Clearly Shanahan had to create these two plots, as I've never before seen them with cutesy heart symbols for data points.  But how did she do that?  Did she get the original data and put it into a graphing program?  I've done that with small data sets -- I C&P the original to good old Paint, turn on the grid lines, play with the aspect/size a bit and its pretty easy to pick off reasonably accurate numbers.    Heck, I just used the better quality (presumably accurate) version from Teicholz's book and C&P'd little pink hearts over the dots.  I'm quite sure it's not perfect, but it's close. 

No, Shanahan didn't do that, she just made up a bunch of points and put the graph in her book.    I mean she didn't even accurately portray Keys' supposedly impeccable cherry-picking!   There is no amount of aspect ratio and scaling that can turn the plot on the left into the one on the right; she's got Japan and Italy closer together on the horizontal axis, while Australia and Canada are now separated considerably.

Folks, this is actually a more serious issue than the misrepresentations of Taubes and his various plagiarizers.   Shanahan put these figures in a book.  Not some promo flyer or a blog post, or even some free ebook that you can download by signing up for someone's newsletter.  No.  A book that she explicitly copyrighted and sold/sells.  The first issue is that there is NO attribution whatsoever for her plots.  At the very least, there should be an "adapted from" and the original sources provided.  But it should be clear to everyone that she didn't copy the plots and she didn't adapt them.  No, she made them up.  Her handiwork just happens to be so, so, very, very bad and obviously erroneous that there can be no doubt about this.   Here again is her plot of the 22 (sic) countries juxtaposed with the original plot of the data from Yerushalmy & Hilleboe.  The red vertical line is at approximately 25% fat for reference/comparison.  I later numbered the countries (arbitrarily) on Shanahan's plot so as to make it easier for folks to see that her plot included a mystery 23rd country.


The second issue is that she usedthis fakery to level some pretty slanderous charges against Ancel Keys.  If you're going to level charges of this nature, you should damned sure be able to back them up with solid evidence.

And now, before I bring you Deep Hole Digging by Dr. Cate Shananigan, I took the liberty of using the (again, presumed-to-be-accurate) version of the Y&H data from TBFS, and superimposing it on Shanahan's.

This should be not only embarrassing to the doctor, but should have her thinking about how best to correct the record while saving face.


Dr. Cate, as she likes to go by, took a different tack.


Our exchange is perhaps best followed on Twitter from this link (that's the problem with Twitter when others get involved and/or new "tweet streams" are created", I only include a link for posterity, I'll screenshot everything pertinent here).


Yep ... bring up Teicholz's nonsense and demonstrate off the bat that you haven't even read the criticism to which you are responding.  Good start Dr. Cate!!


Note:  The picture in the initial tweet in this part of the exchange is the side-by-side where I numbered her hearts, that I included above.  Seriously??  If I were to have done something so half-assed (to be blunt) as Shanahan did with her plots, I'd take a pause and get my ducks in a row before responding further.  Because she's claiming basically that she must have mistakenly plotted the two low points (between 20 & 30% on the Y&H plot) on her version and if she hadn't her case for "no correlation" is even stronger!!   As you can see, I then pointed out to her that her plot contained 23 points -- to the best of my knowledge, there is no 23 Countries Study or data set out there.   So yeah, if you migrated two of the hearts on her plot south to the locations of countries 8 and 14 on the Y&H plot, you'd get even more faked scatter.  This is not a valid argument.   Undeterred, Cate shot back with:


I had not as yet addressed Shanahan's "fit line".   If anyone, not knowing what they were looking at, were to see the scatter at right and not see an upward trend, they have greater problems than I can address here.   The scatter is obviously considerable, but quite often it is and it can appear far worse than this and still be statistically significant.   Even Shanahan's plot as drawn does not justify her ridiculous horizontal arch there.   Ultimately, she can't have it both ways.  Either there is no correlation, in which case that line of hers should not be there, or there is one.  Her inclusion of that trend line only adds to her deception because it implies that there is a correlation that is essentially a horizontal line.

Ahhh, but here it comes!  Ya gotta love someone too stupid and full of themselves to even bother to vet their opponent!


Ahhh yes, I need to study statistics!  LOL.  I should call back when I understand linear regression.   This is super rich coming from someone who seemingly can't even understand the basics of a scatter plot or the difference between no correlation and drawing a fit-line on a scatter plot.    But she digs deeper.  It's not like I didn't telegraph my position there -- the source of the data, that Yerushalmy & Hilleboe paper -- included extensive regression analyses.   Clearly she didn't even read The Big Fat Surprise, even though referencing it was her first response.  How stupid do you have to be to not take a breath when numbers don't add up?  I don't care what the late Seth Roberts found regarding butter consumption and mathematical ability, of late there have been way more demonstrations of buttery ketones impairing the ability to add, and even the ability to count!    

What Shanahan is claiming is that in his Seven Country's (sic) Study, Keys only picked Seven out of the now 23 countries for which data were available.  This is a whole new level of imagination we're dealing with here.   I don't know but if someone were challenging me, I'd be going back to double check my references and sources ... you know, just in case I might just be remembering incorrectly?  No, not Cate!  


The link is to the Y&H paper, and below is the image in the second tweet.
Keys only used B-26 "arteriosclerotic and degenerative heart disease" for his Six Country Plot, so the first column above is the relevant one.

A quick stats lesson on regression and correlation coefficient (r):
  • A positive correlation slopes up and to the right, as the x-variable increases, the y-variable increases as well
  • A negative correlation slopes down and to the right, as the x-variable increases, the y-variable decreases
  • The correlation coefficient, r, is a measure of the "fit" of a trend line and is calculated using the deviations of the data points from the value predicted by the line.  
  • If r = ±1, this is a perfect correlation and each data point falls exactly on the line connecting them.
  • If r = 0, there is no correlation and no line is drawn
  • Both of the previous two situations rarely occur, but the closer the |r| is to 1, the better the fit.
  • The critical values (bottom right of the table) are determined for the degree of significance and the number of data points.  
  • An r-value either greater than a positive r-crit, or less than a negative r-crit (e.g. closer to +1 or -1 respectively) denotes a statistically significant correlation for a given significance level (α).
Y&H had 22 values for their analysis of B-26 mortality rate vs. % calories from fat data.  The correlation coefficient for this regression was 0.587 while the r-crit for α = 0.02 (98% confidence level) was 0.508.    Therefore, the correlation between the proportion of calories as fat in the diet and death from ateriosclerotic and degenerative heart disease reached the level of statistical significance.  I'd encourage reading that paper, it's interesting even if such correlations do not imply causality.  Moreso than demolishing whatever correlation Keys sought to demonstrate in his plot, Y&H stressed that there were even more tightly correlated -- "specific" -- variables than % calories from fat.   The percent calories as animal protein had an r = 0.643, and % as animal fat is even worse at r = 0.677 (though n=21 so not directly comparable).    It gets worse for the Shanahans of this world.  Notice the negative signs next to the values for percents of calories from veggie fat, veggie protein and carbohydrate.  This denotes a decrease in mortality vs. increasing percentage of dietary calories, and the latter two reach statistical significance for the negative correlation.   To repeat, percent carbohydrate in the diet was significantly correlated in negative fashion to B-26 heart disease in the 22 countries.   The greater the percentage of calories as carbohydrate, the lower the B-26 mortality.


No further word yet from Shanahan today.  


Perhaps someone finally did take away her shovel.

Her appearance at AHS12 was an utter embarrassment.  She's slotted to speak again this year.  Any impression that AHS is the academic arm or the more scholarly symposium in that community is a total farce at this point.  


One can listen for a couple of minutes from about the 29:30 mark of this video.  Just before this Chris Kresser cites peer review literature outlining the content of the Okinawan diet.  She just doesn't believe it.  "I'm sure you have a good reason to say what you said", she says with a smile.  

I would hope you have good reason to say what you said in your book Dr. Shanahan.  I just can't seem to find it.


Comments

David Pete said…
Unfortunately she is only trying to sell books the "the faithful" She will probably be successful..
charles grashow said…
http://zerodisease.com/archive/Dietary_Goals_For_The_United_States.pdf

U.S. DIETARY GOALS
1. Increase carbohydrate consumption to account for55 to 60 percent of the energy(caloric) intake.
2. Reduce overall fat consumption from approximately 40 to 30percen energy intake.
3. Reduce saturated fat consumption to account for about 10 percent of total energy intake;and balance that with poly-unsaturated and mono-unsaturated fats, which should account for about 10 percent of energy intake each.
4. Reduce cholesterol consumption to about 300mg.a day.
5. Reduce sugar consumption by about 40 percent to account for about 15 percent of total energy intake.
6. Reduce salt consumption by about 50 to 85 percent to approximately 3 grams a day.

The Goals Suggest the Following Changes in Food Selection and Preparation

1. Increase consumption of fruits and vegetables and whole grains.
2. Decrease consumption of meat and increase consumption of poultry and fish.
3. Decrease consumption of foods high in fat and partially substitute poly-unsaturated fat for saturated fat.
4. Substitute non-fat milk for whole milk.
5. Decrease consumption of butterfat, eggs and other high cholesterol sources.
6. Decrease consumption of sugar and foods high in sugar content.
7. Decrease consumption of salt and foods high in salt content


OMG - McGovern wanted us ALL TO DIE!!
BTW - see ANY mention of margarine??
charles grashow said…
http://news.google.com/newspapers?nid=1299&dat=19730419&id=kNRHAAAAIBAJ&sjid=_YsDAAAAIBAJ&pg=6312,1170888
Dr. Aklins Goes to Washington
Karin said…
I've always wondered at people who said there is no correlation in the 22 countries study. There is definitely an upward slope there. It's not the tightest correlation ever, but it's there. But, geez, her graphs are so obviously made up. It's hard to look at them.


Well, if you every figure out a way to get the delusional to listen to facts, let me know the secret.
hmavros said…
Reading that Twitter exchange was infuriating. Not so much that she didn't take your points, but that she made no real attempt to engage with you, falling back instead on quick and easy debating tactics.


It seems to me that the biggest problem here is that people (like the good Dr) get an idea about the diet, test it out a little on themselves and their friends/clients, cherry-pick some research/blogs that confirms their new bias, and then (here comes the doozy) launch into writing a definitive book about it that leverages off their anecdotal experience and their other qualifications (i.e. not on the scientific rigour of their hypothesis).


When they are (inevitably) challenged by real science, they are "in too deep" (having a book to sell, and a reputation to protect), and so they resort to defensive debating points and counter-attack, rather than engaging with the critique in a good-faith way.


Bottom line, people get too certain about things far too quickly...and when they do, they invest themselves in their new certainty with such commitment as to strongly discourage revision (and certainly, recanting).
Donnie Reilly said…
That's such a bad job that it's funny..... but it's not funny in the very real sense that people will buy into that $##!
Mike Victor said…
The Bible's a bestselling book too. Doesn't mean it's true or accurate.
Mike Victor said…
Could you delete this double post? Thanks, Carbsane.
Wuchtamsel said…
Ahrgh, this is really sickening. I think I even had more productive discussions with my kitchen table... Vicatious embarrassment for the win.
David Pete said…
What book is that from. Please tell me it's Jimmy Moore's. If that uncritical mess of dog shiz was written by someone with a medical degree I may sit and cry.
carbsane said…
LOL ... "just there for the refreshments" Jimmy Moore. Oh my!
carbsane said…
Hey Steve, any way you can shoot me an email? If not I understand, but anonymity will be guarded if you are at all worried about that.
carbsane said…
Wonder if he knows Bill Lagakos of Calories Proper :p
carbsane said…
Hi Harry! The back is back :-)

http://www.marksdailyapple.com/primal-peak-performance-breaking-through-to-the-next-level/#axzz35glmWhxW

>>> A doctor lacking nutritional/metabolic expertise is not a criticism – it’s simply not part of their formal training. You could just as easily assume they know nothing about sailing or skydiving. When I started my medical practice, I didn’t have a clue about nutrition or metabolism either. The only reason I have a clue now is thanks to time spent – about five years – updating my education, accomplished by reading a small library’s worth of textbooks and articles from Medline and attending medical meetings held by relatively enlightened societies like the American Academy of Bariatric Physicians and the Nutrition and Metabolism Society. Much of what I learned is now compiled into the two books I’ve co-authored with my husband Luke. <<<



Since she's now famous for being the nutritionist who turned around the LA Lakers (cough cough cough) , I don't see her turning back!
carbsane said…
Thanks for saving me the time of investigating the diet of the Himba. Why am I not surprised?
charles grashow said…
The Search, 1953

The ENTIRE SHOW

http://www.youtube.com/watch?v=SY7ASzu-X1U
charles grashow said…
http://www.youtube.com/watch?v=qGU_NX8Ao5E

Italy and Crete Pilot Survey 1957
carbsane said…
She's quite a scientist that one.
carbsane said…
This is from Deep Nutrition. Small nit pick in her photo credits, she identifies the man on the right as a "Maasai Man, photo courtesy of John Hanley"
MacSmiley said…
Of course not. Telling the real truth, versus the real fake truth that manipulates emotions, just doesn't make for NYT/ Amazon best-sellers.
MacSmiley said…
Kevin is incredibly talented but Matt is not? Dontcha just love manipulative language?
Trae Palmer said…
This is such a nonsensical comparison between the two photos: they're not taken from the same perspective, lit the same, nor are the individuals depicted even making the same facial expression. You know how they require ID photos to be taken within certain, standardized requirements...yeah, there's a reason for that.
Steve said…
I shot you an email under the name "Stevie Wonder." Let me know if you don't get it.
Steve said…
You're just getting shallow nutrition, Mike.
MacSmiley said…
Interesting potshot against the AMA in the closing paragraphs. Forget the lack of data on Atkins's part. Not even case studies. Now we've got Westman. Don't know which is worse. No science or Atkins-funded science.
carbsane said…
See I don't see what this woman does. Matt is seemingly ageless, and this is actually not the most flattering photo of him, but Kevin looks younger to me if I had to choose. Ditto many of her other sibling comparisons on her website, etc. I do not know anything about Kevin but Matt got discovered by some random talent scout if memory serves, and soon after was in Little Darlings. Perhaps if Kevin is an actor with the same talents and break, he is the guy who gets all the romantic leads. SO DUMB.
Mark said…
You don't think replacing saturated fat with polyunsaturated meant margarine and vegetable oil?
MacSmiley said…
Is that the entire 36 min video? I can't get it to play on my iPhone, not even with my neighbor's Wi-Fi.
charles grashow said…
Just click on the watch video icon - it's the whole 36 minute video
MacSmiley said…
Wow! Gotta love the last few min with Eric Sevareid saying the team was cautious and not ready to draw any flat conclusions. Thanks for this post, Charles. Will watch others later.
charles grashow said…
You tell me.
MacSmiley said…
Yeah. Did that. Didn't work. Guess I'll have to try it on a PC at the library.
Paleo Nouveau said…
It seems that more than one thing occurred around this time that probably influenced the detrimental health outcome that followed.
More women in the workforce meant less time to cook.
Intro processed foods & Madison Ave touting "low fat" prepackaged foodstuffs.
Along with more work we started to become more sedentary.
Stress levels for most increased.
TV became more & more a part of our lives.
Less fat & sugar was not consumed as suggested, nor were lean meats. More processed foodstuff was along with more salt, sugar & bad fats.

Does anyone really believe that the above suggestions are harmful & the culprit of our current health challenges? I think most people just repeat the BS punchline that the internet gurus claim Keys science was responsible for which is "Eat less fat!" And obviously for their understanding, that did not work.
carbsane said…
>>> More women in the workforce ...<<<


It is not PC to discuss this lest one get accused of wanting to go back to the days before women could vote, but in addition to this you had an explosion of single parented households. Divorce was quite rare, as was having a child out of wedlock when I was growing up. Nowadays many have children deliberately as singles, and divorce is rampant. This makes for eating most meals together as a family at home a rare thing and that has pretty far reaching implications in a fast food world where food is available 24/7 "on the go".
charles grashow said…
http://sph.umn.edu/site/docs/epi/SPH%20Seven%20Countries%20Study.pdf
ON THE TRAIL OF HEART ATTACKS IN SEVEN COUNTRIES


By Henry Blackburn, MD
MacSmiley said…
Rats! I'm sitting in the Public Library at a PC and the video's not loading here either! Grrrrr!
StellaBarbone said…
I think that two working parents is a big part of the problem. I don't think it's not PC to note that a full-time homemaker has a lot more time to make a home and that cooking from scratch requires time. An employed mom who is also doing the majority of the housework and the childcare will have to cut corners somewhere.

Buying food ingredients rather than prepared food makes you a lot more aware of portion sizes, too.

We've also seen a lot of cultural changes in eating between meals. When I was a kid, my mom did not provide after-school snacks, soda was a dessert item, desserts were a weekend treat, desserts were usually fruit or ice cream, and hunger before a meal was considered a positive sensation. "You'll ruin your appetite" was a frequent admonition. When I first started working, people used to bring their cup of coffee (and cigarettes) to meetings, but there were no bagels or donuts, much less lunch, routinely served. People did not walk around in stores or down the street while carrying food and drink. Water did not even come in bottles. A lot more people smoked back in the pre-obesity epidemic days too.
MacSmiley said…
Fascinating. Certain things stuck out for me.

"Nefarious activities"
"Lead poisonings"
The guy from Pasteur Institute in FRANCE going to Greece to figure out their longevity
Fly poop throwing off serum cholesterol measurements
;-P

No mention of "low fat diets". Keys was not so keen on the marketing of vegetable oils, either. In his estimation, biggest factors contributing to longevity: Not smoking and blood pressure.
charles grashow said…
http://blog.americanhistory.si.edu/osaycanyousee/2013/05/what-did-soldiers-eat-during-the-revolutionary-war.html?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+OSayCanYouSee+%28O+Say+Can+You+See%3F-+National+Museum+of+American+History+Blog

Even before a food supply system was organized, on June 10, 1775, the Massachusetts Provincial Council set the daily allowance or ration for its troops in Boston as:

One pound of bread

Half a pound of beef and half a pound of pork; and if pork cannot be had, one pound and a quarter of beef; and one day in seven they shall have one pound and one quarter of salt fish, instead of one day's allowance of meat

One pint of milk, or if milk cannot be had, one gill [half a cup] of rice
One quart of good spruce or malt beer
One gill of peas or beans, or other sauce equivalent
Six ounces of good butter per week
One pound of good common soap for six men per week
Half a pint of vinegar per week per man, if it can be had.
charles grashow said…
http://www.fasebj.org/content/28/1_Supplement/LB536.shor

HDL from Patients with Inflammatory Disorders has High levels of Oxidized Lipids and is Dysfunctional

Population studies have shown that plasma levels of HDL inversely correlate with the prevalence of coronary heart disease.HDL levels in an individual does not necessarily correlate with the risk for coronary heart disease. Many patients with high HDL, present with coronary events and many with low HDL do not. We demonstrated that HDL composition and function is important than HDL levels. Under excessive inflammatory pressure unsaturated fatty acids can undergo oxidative modification. Protective enzymes including paraoxonase, LCAT and PLTP are inactivated by oxidative modification reducing HDL anti inflammatory properties. This results in HDL not being able to protect LDL from oxidative modification.


Methods: Fasting plasma from patients with coronary disease were analyzed for their ability to prevent LDL modification. Additionally the content of oxidized lipids in HDL was determined using LC-ESI-MS/MS. Results: Among coronary patients over 69 percent had non protective HDL while among healthy controls 4.5 percent had dysfunctionalHDL. The content of a variety of oxidized lipids in the HDL from patients was orders of magnitude higher than that found in HDL from normal Controls.


Conclusion: Until there are effective medications for raising HDL, it should be clear to the public that the best way of dealing with low HDL cholesterol is lowering LDL cholesterol.
Lighthouse Keeper said…
Thank you Charles for providing some splendid Ancel Keys material of late.
Lighthouse Keeper said…
Note the recommendation to decrease sugar consumption.
Paleo Nouveau said…
Maybe it's the pseudoscience sphere but until now that seems to be the multi million dollar question! It is pretty clear why he, Ancel Benjamin Keys & his peers selected the 7 countries. Why all this selected & bias BS that it was "selective" to prove his hypothesis? Hypothesis may be a stretch!!! After reading more from his peers & other articles it appears that Ancel Benjamin Keys had more questions that answers. Isn't that the best starting point of learning?
So much of this crap really comes down to chance as well. It is ludicrous to assume that his brother's career should in any way mirror his own, and one the age front, neither one looks older or younger. They're not only in the exact same age group, but just half-a-year shy of being Irish twins.

*SMFH*
NS said…
As always, wonderful, Charles. Thank you.
Yeah, and I mean, if it's all about 'teh PUFAS and veggie oils', then I can tell you of a diet that is lower in processed PUFAs and veggie oils, even more so than some of the more popular variants of low carb and paleo.

A low fat diet. OOOPS!

Ironic. Sam Feltham used actual packets of nuts as his source of fat with whole meat during his 'high fat low carb - mega high protein' 5000 calorie diet. Then when it came to his atrocious 80-10-10 vegan implementation of a 5000 calorie diet, he used 100 grams of olive oil for the fat portion of the diet.

Ahahahahahahahahahahahaha!

And then these clowns get huffy, snarky and even self-righteous when they aren't taken seriously.
charles grashow said…
http://tobaccodocuments.org/ti/TIMN0210997-1095.pdf
Select Committee on Nutrition and Human Needs
Vol No ll
Diet Related to Killer Diseases
28 July 1976
This is what I just find laughable about the Paleo-tards.

Mostly plants/plant-based/vegetarian with a rounder face, which is actually a rarity. <- Oh, teh fatty no collagen!

Most people of plant-centred diets with lean chiselled faces. <- Oh, noes... too skinny and not enough subcutaneous fat in face!

Guess some idiots just can't get their story and standards straight.
charles grashow said…
http://sevencountriesstudy.com/about-the-study/countries
charles grashow said…
http://www.caloriegate.com/calories-in-calories-out/9-more-experts-lay-waste-to-the-calories-in-calories-out-cico-model-of-obesity/comment-page-1#comment-1886


EXPERTS!!! Give me a break!!
charles grashow said…
http://ajcn.nutrition.org/content/61/6/1321S.full.pdf+html
Mediterranean diet and public health
Ancel Keys
Blogger said…
EasyHits4U - Your Free Manual Traffic Exchange - 1:1 Manual Exchange Ratio, Business social network. FREE Advertising!