The Association of Misleading Studies

Last week I pointed out several flaws in how researchers gathered data for the NIH-AARP Diet and Health Study, which has generated a slew of scary headlines such as Animal Fat Linked to Pancreatic Cancer.  

I also mentioned that even without those flaws, observational studies can at best only produce statistical associations. They don’t prove cause and effect … although you wouldn’t always know that from the headlines.

When people mention that obesity is associated with Type II diabetes and therefore must cause diabetes, I’ll sometimes reply that gray hair is also associated with diabetes and suggest we start giving Grecian Formula to everyone to prevent it.  That usually generates a reply along the lines of, “Come on, that’s ridiculous.  A lot of people develop diabetes when they’re older and happen to have gray hair.”

That’s the good news:  people don’t confuse an association with a cause when it’s obviously ridiculous.  The bad news is that if an association isn’t ridiculous, researchers often do believe they’re seeing cause and effect – especially if the association confirms a pre-existing bias.

Since observational studies produce so many alarmist headlines, I thought it would be a worthwhile exercise to recall just how spectacularly wrong a theory based on a statistical association can be.  This is a real-world example that generated a lot of headlines back in the day.

For decades, heart-disease researchers have known what while women certainly do develop heart disease, they typically develop it later in life than men … usually after menopause.  Naturally, this got the white-coat crowd wondering if female hormones – particularly estrogen – might protect against heart disease.  The theory seemed to make sense:  men don’t produce as much estrogen as women, and women don’t produce as much after menopause.

In the 1960s, men were given estrogen as part of a large clinical trial called the Coronary Drug Project – but that arm of the trial was stopped early because the men taking estrogen began dying from heart disease at a higher rate than men in the control group.  So the theory was adjusted:  estrogen appears to protect women from heart disease, but not men.

Then a major observational study gave the estrogen theory some real traction.  For 15 years, the Harvard Nurses Health Study had been tracking the diets, health habits and disease rates of more than 120,000 nurses.  When researchers pored over the mountains of data produced by that study, they found a startling statistic:  women who took estrogen had a 40% lower rate of heart disease than women who didn’t.  And women who continued taking estrogen were less likely to suffer a heart attack than women who took it for awhile and then stopped.

You can imagine the research papers and the headlines that resulted.  There calls among researchers and doctors alike to start prescribing estrogen to all post-menopausal women who had risk factors for heart disease.  More cautious researchers called for a controlled clinical trial before estrogen was given out like heart-healthy candy, and were criticized for it.  How could they, in good conscience, deny this obvious wonder drug to millions of women while waiting for long clinical trials to play out? 

A  pharmaceutical company, Wyeth-Ayerst, eventually funded the clinical trials – hoping, of course, that estrogen would be shown to prevent heart disease.  More than 16,000 women were randomized and enrolled in the study.  For five years, half received estrogen and half received a placebo.

The results were hardly what Wyeth-Ayerst had expected:  The women taking estrogen developed heart disease at a higher rate – 30% higher, in fact.  They were also more likely to suffer a stroke … another cardiovascular disease.  Later clinical trials confirmed the bad news.

The experts were flabbergasted.  The statistical correlation in the Harvard Nurses Study couldn’t have been more convincing:  women who took estrogen were far less likely to have a heart attack.  And it couldn’t have been fluke –  there were too many subjects involved.

So what happened?  Nobody can say for sure, but some researchers at the time offered an explanation that makes perfect sense:  the women in the Harvard study who took estrogen were more concerned about their health.  That’s why they took a hormone replacement in the first place. 

In other words, estrogen didn’t create healthy nurses, but health-conscious nurses did take estrogen.  Meanwhile, the health-conscious nurses were less likely to develop heart disease … for any number of reasons.

This really isn’t all that surprising. In clinical trials, people who religiously take their pills tend to have better health outcomes than people who don’t.  And guess what?  It doesn’t matter if the pill they’re taking is the actual drug or the placebo.  The difference is in the people, not necessarily in the pill. 

Some people care about their health.  Some people are lackadaisical about health.  Researchers call them “adherers” and “non-adherers.”  I have my own, more colorful labels.  The point is, we’re talking about different kinds of people, and that difference can produce statistical correlations in observational studies that have little if anything to do with the true cause and effect.

Think about the estrogen studies again for a moment: we now know that estrogen doesn’t prevent heart disease and in fact can make it worse.  And yet in a large, observational study, taking estrogen was associated with a steep reduction in heart disease – almost certainly because health-conscious women were more likely to take it.

Now think about some of the alarmist headlines and health-nanny propaganda you’ve read over the years, and ask yourself what’s really going on.  Here a few examples I came up with:

Does a diet high in saturated fat cause cancer and heart disease?   Nope.  But since saturated fat has been demonized for 30 years, health-conscious people probably eat less of it.

Does giving up meat make you healthier?  Nope.  But most people who become vegetarians are probably health conscious.

Do whole grains prevent diabetes and cancer?  Hell, no.  But they’re less likely to cause those diseases than white-flour products, and health-conscious people are more likely to choose them.

Does watching Fat Head at least three times give you a high IQ?  Uh … no.  But I’d like to think there’s a strong statistical correlation.


If you enjoy my posts, please consider a small donation to the Fat Head Kids GoFundMe campaign.
Share

18 thoughts on “The Association of Misleading Studies

  1. Kim

    I no longer know how many times I’ve watched Fat Head – I watched it at least three times with my husband, and countless others with friends and family. I should qualify as a genius by now! 😉

    If you’re eating high-quality natural fats while watching, your IQ may soon exceed Einstein’s.

    Reply
  2. Laurie

    I can’t stop thinking about the idea that to advise people to eat low-calorie diets is akin to handing out the advice to go forth and starve themselves. How did we ever get to this point and why on earth would a doc advise his/her patients to actively starve!*!!*!? How about a car and fuel analogy? If the mechanic advises putting diesel into a gasoline-powered engine, and not enough diesel to boot, how is that good? Cars can’t run on alternate fuel sources within the vehicle when their tanks are empty but what if the low-fueled, wrong-fueled engine could start to cannibalize its own tires for fuel? Tires are made of polymers and they transport the rig (the car) around just like our muscles are made of amino acid polymers (proteins) that help us to roll around. Eating away at the tires to keep the car running just can’t be a great use of resources and energy for the car as a whole. I just get all worked up when I think about this and I can’t stop thinking about it. And the whole low-fat dogma is just like the difference, let’s say, between diesel and gasoline. A gasoline engine needs plenty (or at least enough) of good quality, ‘high-fat’ gasoline and not a misery limited quant. of ‘low-fat’ diesel!!!!!!!!!!

    When I think of all the people out there punishing themselves with a diet they don’t really like and being rewarded with damaged metabolisms and nutrient deficiencies — not to mention increased risk of type II diabetes and autoimmune disorders if they’re loading up with “whole grain goodness” — it does make me want to scream. (I mean, uh, scream like a ferocious leader-of-the-pack chimp … not scream like my daughters.)

    Reply
  3. Ms. X

    I’ve watched Fathead twice… my IQ is rising, no? 🙂 Acutally, what happened was the first disc Amazon sent me was warped, and in order to get it to play at all, I had to balance it (a little electrical tape did the trick). But naturally I returned it and when they sent the replacement, well I had to watch that disc to to make sure it played properly 😀

    Of course, if it hadn’t been a great movie I would never have watched it twice. I expect to get a third viewing in as soon as we have visitors.

    (Btw, the second disc is better, it will play without the redneck engineering, but it is not perfect. I brought it up just in case it is not a known issue).

    Yikes, I didn’t know anything about bad discs! When did you order them? I know our U.S. distributor stopped using one of the DVD-printing firms because they couldn’t seem to ship on time … or to the right place half the time.

    Reply
  4. Dave Dixon

    If I’m not mistaken, the estrogen used in the studies above is derived from horse urine. Gross factor aside, horse estrogens are not the same as human, and apparently have the annoying habit of hanging around the body for much longer than bioidentical estrogens. It’s a good example of another common fallacy in scientific reasoning, where the “givens” were forgotten. All of the results above are qualified by the statement “given that the patients took horse urine derived estrogens”, which you can’t just drop and then turn into a statement that “estrogen does X”. This happens all the time (another good recent example is the business with LDL lab results and the Friedewald equation).

    Yee-uk. From what I can tell, however, the nurses in the Harvard observational study were taking the same stuff. Premarin was named specifically on the Harvard questionnaire. So we’re still looking at an apparent reduction in heart disease if you go by statistical association, but an increase in heart disease in a controlled trial.

    I haven’t found any references to studies of bio-identical estrogen and heart disease. Has anyone conducted one that you know of? It would be interesting to see if the results held up.

    Reply
  5. Ellen

    Dave’s comment reminds of the forgotten “given” that started the whole crazy “cholesterol causes heart disease” propaganda in the first place.

    Scientists fed a high cholesterol, animal fat laden diet to rabbits, and found that the rabbits developed heart disease. In my mind, all that these scientists revealed was that rabbits aren’t meant to eat animal flesh, as their bodies obviously can’t digest it.

    Now, any person with the sense to avoid playing in the street would understand that extrapolating those results to human beings is just ridiculous. But not those scientists.. nope, they assumed if the rabbits got heart disease from eating animal flesh, humans must be in the same danger, ’cause anyone can obviously see that the human body structure and natural diet are the same as a rabbit’s..?? The leap of logic there is just mind boggling to me..

    And to this day, the whole “cholesterol causes heart disease” theory is still rooted in that nonsense. And even more amazing, many scientists and physicians don’t question that association.

    Yup. Other than in “Monty Python and the Holy Grail,” you don’t see a lot of killer rabbits.

    Reply
  6. Ms. X

    The first disc was ordered on June 30th.

    I’ll check with our distributor and see if they know what’s going on. I just a new box of DVDs to send out; I’ll check one.

    Reply
  7. TonyNZ

    @Ellen

    They probably fed the rabbits 2000 kcal as well, just to make sure that it was the same amount of fat that a human eats for a true comparison.

    Incidentally, talking about feeding animals too much, foie gras is fatty duck liver, made especially by force feeding them…you guessed it, corn. Just another in the myriad of examples where grains are used to fatten.

    For the record, I think PETA is as nuts as most others here would think, but I think foie gras is a bit unnecessary on the humaneness side of things…

    Indeed. So are feed lots for cows, when we could be raising them on grass.

    Reply
  8. Laurie

    Because Oprah is one of the most visible and influential people on the planet, I sent the following note and ‘show suggestion’ to her producers.

    Anti-aging, teenage obesity
    Please read his book and interview Gary Taubes. He wrote “Good Calories, Bad Calories”. The book might have been better titled Good Science, Bad Science. It’s not a diet book by any means. The teens you interviewed on a recent show are obese NOT primarily because they are over-eating,(over-nourished)- but because they are MAL-nourished. They are eating foods that make and keep them hungry- perpetually. Taubes explains that just like small children don’t grow taller because they are overeating, they’re overeating because they’re growing taller. The same holds true for obese folks. They’re not getting fatter because they are overeating, they’re overeating because they’re getting fat. This didn’t make sense to me the first time I read his book, but it does now.
    Aging. Insulin has only three functions. It is the hunger, aging and fat-storage hormone. Our ancestors produced lots of insulin if food was available and thus it was an appropriate time to reproduce. So there was a polar difference between high insulinemia and it being a great time to reproduce and inadequate food and it not being a good time to feed the extra mouths of offspring. The trade-off and by-product effect of insulin is hurry up and reproduce, but it will cause you to age faster, or low-insulin will keep you young until such time as conditions improve and you can then reproduce. I’m not making this up. The effect insulin has on the human body when it’s in too much supply is to create Advanced Glycation End-products. The acronym is “AGE” for a reason.
    Please, I implore you to read Taubes’ book and Malcolm Kendrick’s “The Great Cholesterol Con” and please watch Tom Naughton’s DVD documentary, “FAT HEAD”.

    Can’t hurt to try. We sent copies of the film to a research director there who happens to be a relative of my wife’s. If Oprah would stop the low-fat rollercoaster ride and tell the world to kick the carbs and embrace natural fats, she’d do more to improve the nation’s health than anything being kicked around on Capitol Hill.

    Reply
  9. Deborah

    Have you seen this?-

    http://www.healthday.com/Article.asp?AID=629014

    “Lower IQ Linked to Higher Risk of Heart Deaths
    Public health messages should be simplified, study finds”

    Apparently everyone is dying from heart disease because they’re too stupid to understand the solution.

    Yes, it takes a highly intelligent person to understand “Eat a low-fat diet.” That must be it.

    Good grief, they’ll never admit they were just plain wrong about this. Lower IQ people are generally poorer, and poorer people consume more cheap carbohydrates. But they’ll never look at that possibility.

    Reply
  10. Ms. X

    @TonyNZ – remember ducks/geese are birds. Just because you assume gavage would be horrible, doesn’t necessarily mean they feel the same way.

    http://www.proteinpower.com/drmike/uncategorized/foie-gras-cest-moi/
    http://www.nytimes.com/2005/06/29/opinion/29iht-eddownes.html?_r=1

    Just because something *can* be done cruelly doesn’t mean it is. Almost everything we do with our treatment of animals can be done cruelly. Very little of it actually is.

    http://everydogsblog.blogspot.com/2009/07/tied-up-in-nots.html

    Reply
  11. TonyNZ

    Ms. X, I work in agriculture, I agree with you 100% on the cruelty issue. Stress hormones give you less and worse product at the end, so you want your animals (in my case, cows) to be as content and happy as possible. These people that think that farmers are horrible exploitative creature-beaters couldn’t be further from the truth.

    And as for foie gras, I’m sure they don’t make it violent for kicks for the above reasons (and they are probably even more anxious to make a point of no unnecessary mistreatment given the stigma associated with it). I have nothing against farmers or eaters of foie gras and have no problem with them continuing with it. I just don’t like foie gras that much and given the procedures of it’s production, would much prefer a marbled steak (which is achieved by similar means). Someone who actually likes foie gras would obviously be much more interested in keeping it available.

    Reply
  12. gallier2

    Stressed ducks and geese do not produce foie gras. I’ve seen several time how the birds allign to get their shot of maize, if they hated it they probably wouldn’t.
    You have to see that foie gras is not really unnatural, it’s their way of accumumating fat for the long migration flights. If the birds put it subcuteanously like us, they wouldn’t be able to fly anymore so they developped this strategy which allows then the liver to break it down as necessary.
    Foie gras is only the best source of Vitamin K2 also it is order of magnitude better than anything else for that (this may explain why the south-western french have so much less cardiac problems than the rest of France).
    And it tastes so incredibly good (when done right)…

    Reply
  13. Isabella

    Anyone who is interested in foie gras production (pro or anti) really should watch this clip from Anthony Bourdain’s show “No Reservatons:”

    http://www.youtube.com/watch?v=ABeWlY0KFv8

    The ducks appear to be very well treated. I imagine that this question applies to all animal products. Some animal farms treat their animals very well indeed; others, not so much. Choose sources wisely, and everyone (including the animals) benefit greatly.

    Reply
  14. TonyNZ

    OMG! Nuts are fatty but people who eat nuts have less heart disease, diabetes and obesity! WTF!? There must be something unidentified in nuts that is so good for you it counteracts the effect of the fatty horribleness!!!

    And as for the foie gras can of worms, I never expected this sort of response, but there’s been some good posts come from it.

    Don’t worry, the experts will come with an ad-hoc theory to cover themselves. Well, uh, it’s not saturated animal fat you see, so, uh … we’ll get back to you on that.

    Reply
  15. Gita

    “Does watching Fat Head at least three times give you a high IQ? Uh … no. But I’d like to think there’s a strong statistical correlation.”

    I have noticed that people who don’t get the point after seeing the movie are not too bright. Would you believe that a couple of people that my mom showed the movie to concluded that you must have been paid by McDonalds to make the movie as a promotion.

    As a side note, the DVD I sent to my stepson ran for an hour and then had problems. Amazon is sending a replacement, so it is not a problem. It may just be a fluke.

    That’s the second time I’ve heard about a faulty DVD, so I’m going to alert the distributors. All the ones I’ve played have been fine, but the printing plant may be having quality-control issues.

    I’ve had a few people comment in print that I must’ve been paid by McDonald’s — one was suspicious because there was so much McDonald’s “imagery” in the film. Let’s see, I declare from the start that I’m going to eat all fast food for month, mostly McDonald’s, and wouldn’t you know it, their images show up in the film. Shocking.

    My accountant probably wishes McDonald’s had paid me, but other than granting me permission to shoot in their restaurants (which took some doing), they had diddly to do with it. If they’d paid for this thing, they’d probably be handing out copies with Happy Meals right about now.

    Reply
  16. Dana

    Ha, I wrote about this (although I’m reworking the blog), and my take on it was of course you are going to find pancreatic cancer in a bunch of old people–it was an AARP study! Older people are THE afflicted group for the vast majority of PC cases.

    Re: gray hair and diabetes–Your joke is apt, and your friend’s remark about diabetes appearing in older people is somewhat valid (though less and less so–at the rate we’re going diabetes will be a disease of twentysomethings and we’ll be seeing a lot more older people without their feet), but diabetes is also associated with *accelerated* aging, so someone who might have gone gray in their sixties is doing it in their forties instead.

    My dad has been slow to gray for most of his adult life despite having jet-black hair which statistically begins graying in one’s teens for most white folks with that hair color. So he’s always had at least a few strands. But up until the early 2000s his hair was still mostly black. Then I went to stay with him for a while at the beginning of 2006 and I swear, in five years it looked like he had aged ten to twenty. He looked more brittle and he had a LOT more gray. I saw him once more in 2007 and haven’t seen him since, but he had an episode of blood loss last year and my brother tells me he’s even worse since then. He was born in 1951. He’s not *that* old. He was diagnosed with type 2 diabetes in late 2005.

    So there you go, I know what anecdotal evidence is worth but anyway. And yeah, Grecian formula would have done him about as much good as the anemic salad veggies and margarine his doctor told him to eat.

    Reply
  17. Dana

    Oh, and, I’m 35. Another piece of anecdotal data. I did not have a single solitary gray hair until I was 31. At the same time my gray began appearing, I had put on a huge amount of weight and suddenly looked haggard. Creases had appeared under my left eye (yes, just my left eye–I sleep mostly on my left side since my daughter was born) and I had a lot more discoloration in my face than previously.

    Makes no sense that it would all hit me so fast but most of the weight I gained was abdominal. I have no health insurance but I don’t think I am diabetic yet–however, I’m likely prediabetic. This stuff is taking its toll on me already. I come from a line of women who age late and tend to age well, IF they take care of themselves. Of course that’s been hard to do on a low income–my life’s recently taken a turn for the better and hopefully I’ll get a handle on it now. I don’t mind the gray hair, I just mind what it portends. My mom wasn’t graying yet in her mid-thirties.

    Reply

Leave a Reply to Laurie Cancel reply

Your email address will not be published.