Calling all you educated types

[andrewducker]

There's an article here looking at bodyfat, Vitamin D and Autism/ADHD and I'm wondering if it's junk science, or it's something I should be paying attention to.

It's already passed my basic "It's not written by a raving madman" filter, but there are people on this list who are much better educated than I about biology.

Anyone got any thoughts?

Comments

[endless-psych.livejournal.com]

Strikes me as utter bollocks on the face of it.

I'll ask the good folks on badscience forums later.

[endless-psych.livejournal.com]

In fact I'm not sure it even warrants further investigation "Sunscreen causes Autism"...

[andrewducker]

Ok. You're currently sounding _less_ convincing than he is. He at least backs up his statements with what look like causal links. In order to demonstrate that his argument is nonsense you have to either show his biology is nonsense or that his chain of logic is wrong.

[erindubitably.livejournal.com]

She. :P

[andrewducker]

Thank you! :->

[erindubitably.livejournal.com]

I made the same mistake 'til Mo pointed it out.

I'm reading the article now - it seems like a really tenuous connection based on other hypothetical conclusions, but I can't break it down more than that yet. Hopefully a chink will appear soon or someone more savvy will explain it, because it seems too odd to be true!

[kurosau.livejournal.com]

I'm not sure, but the previous poster may have been referring to the bit that was going around a while back about how autism can be caused by absorption of mercury from various products, the prime of which I believe was a vaccine of some sort. If a sunscreen had metals in it, or unusual elements, I could see people believing such a thing, and that might be an easy crap detection method.

[endless-psych.livejournal.com]

Thats only just spread across the Atlantic.

Indeed the MMR "controversy" or "scare" is as peculiarly a British thing as Creationism is mainly an American thing.

I was pretty much referring to the fact that it seemed unlikely that reading more then "sunscreen causes autism" in a an article published on t'internet and not in a peer reviewed journal specifically dealing with Autism and with references that were links to nutritionist woo was highly unlikely not to be bollocks.

I think I've shown more why this is the case below now however.

[heron61.livejournal.com]

Indeed the MMR "controversy" or "scare" is as peculiarly a British thing as Creationism is mainly an American thing.

Would that this was true. Over here (in the US) there is no shortage of health nutjobs who now fear all vaccines because they think they cause autism. If you point out that almost all modern vaccines don't contain mercury, most will then assert that the entire process of vaccination itself is bad and causes autism.

[endless-psych.livejournal.com]

Actually one theory is that a "stressful biological event" could concievably bring on the onset of autism (if a genetic prediliction already exists I think) and a vaccine falls into this category. Which was one of the competing theories put forward against the rather random split vaccine idea IIRC.

[endless-psych.livejournal.com]

A few things that brought me to the "it's bollocks" conclusion.

"In fact, I have now formed a mental profile of the prototypical mother of an autistic child:"

With no mention how this was achieved. Given this is pretty much the cornerstone of the argument we can assume its unfounded.

One of the earliest signs of the onset of puberty in girls is the development of shapely curves. Men find such curves to be sexually attractive, an inate instinct that plays into mother nature's goal of assuring a healthy supply of fat for a potential pregnancy. Biological processes intentionally program men to be turned on by curvaceous women in order to assure that women with inadequate body fat (thin and/or muscular women) will be less likely to become impregnated.

...Intentionally condition men? Biological processes? Hows that then? Certainly hip to waist ratio is often commented on as a universal indicator of human attraction however what the hell would constitute an operational definition of "shapely curves"? Another black mark there.

"Human biology wants fat not just on the person, but also in the diet, if a pregnancy is in the wings. This fact has been proven quite conclusively by a recent analysis of data from the Nurses' Health Study"

Follow the link, worry that the author has linked to a story published in the press and not the journal it appears in, or indeed on of the papers authors. Also read the comments there to get an idea of why the study might be flawed.

consider this: chicken eggs are now considered "unhealthy" due to their high concentration of cholesterol. They are also one of the best food sources of vitamin D.

Relies on the cornerstone of the argument being valid (health concious mothers are more likely to have autistic children) and given point one is an argument from authority...

The section entitled "The Rise of Autism in America" is particularly galling. (Not as insulting as http://www.peta.org/feat_autism.asp this granted but still pretty bad - see here for a better idea:- http://www.ipetitions.com/petition/autismPETA/).

The following remarks are emotive and have no place in any serious work of science:-
"the tragedy represented by the birth of a single autistic child"
But here I want to focus on autism, one of the biggest tragedies of the 21st Century, which I believe could be averted with very minor changes in lifestyle."

[endless-psych.livejournal.com]

All of the following is almost entirely factually innaccurate:-
"An autistic child will never be able to function independently as an adult. Their IQ typically tests well below the IQ of a Down's syndrome child. They live in self-imposed isolation, and anyone who tries to reach through their thick fog is only left frustrated by an inability to communicate with them, an inability to teach them anything of consequence. Asperger's is a much milder form of the disease. Asperger's children often test to have IQ's that are above average, but they lack intuitive social skills and come off as odd and awkward in social settings."

Autistic children often perform better then Downs children at IQ tests.
Living in "self imposed isolation" and "inability to communicate" are only ever really true at extreme ends of the spectrum and even then I would argue against the "inability to communicate" thing. I work with individuals who have very severe ASDs and trust me they can communicate far more effectively then you might think... ...even if they are non-verbal. "An inability to teach them anything of consequence." is pretty much a non-sequiter. What things are counted as being of consequence? We've taught our clients not to cross the road without looking first - is that inconsequential? Should we have left them to it? How to wash, how to cook and clean - are these inconsequential? If so what is consequential? Author/definition FAIL.
Asperger's is a much milder form of the disease This totally ignores the controversy surrounding aspergers/autism - are they the same developmental issue or are they two seperate problems? Are those with aspergers just "high functioning" individuals on the Autistic spectrum. I feel this betrays the authors total lack of the specialist knowledge required to make grand claims about autism.

I'm sure there is more in the paper thats more amazingly woo tastic - but that will surely do for now.

Causal links my arse ;)

[henriksdal.livejournal.com]

It has wikipedia as a reference? Seriously, if I was writing an article to be taken seriously, most of the refs would be peer-reviewed journals, not URLs, but I've not had a proper look. I'll wedge some time in and investigat-o, if you like.

[ninox.livejournal.com]

I'm a medical librarian who has already offered. I agree with the wikipedia comment in this context. Sometimes it may have a place. Peer reviews are not the be all and end all though, especially as it take so long to publish and competition is heating up for the main titles.

[nancylebov.livejournal.com]

I've just skimmed the article, but it looks as though it implies that going low-fat is just a random choice that a pregnant woman might or might not make.

Here's a history of the recommendations for women to limit their weight gain during pregnancy or even lose a little if they're very fat.

Eating low-fat food would be a likely way of trying to comply with the recommendations.

As for melanoma rates, it looks as though they're at least not dropping.

And here's one about vitamin D deficiency and multiple sclerosis. Note that it doesn't mention the recommendations to avoid sunlight.

[robhu.livejournal.com]

"Stephanie Seneff is a Principal Research Scientist in the Spoken Language Systems Group at the MIT Laboratory for Computer Science. She received the B.S. degree in Biophysics from MIT in 1968"

Seems like she's out of her field (although she has a BS in Biophysics).

I think the acid test for claims like this is whether they've been published in a reputable journal. AFAICT this claim has not been (I didn't read the article, just skimmed it). That's the first step. At that point it might be a reputable idea, but could still be wrong.

[akicif.livejournal.com]

Seems like she's out of her field

Quite. Cargo cult science at its best - along with a healthy dollop of Engineer Syndrome.

I think the acid test for claims like this is whether they've been published in a reputable journal

Or google for '"Stephanie Seneff" autism inurl:.edu' or '"Stephanie Seneff" autism inurl:.ac.uk' and see how few hits you get....

[ninox.livejournal.com]

Ah but you can find papers to back up most statements - it is finding valid ones that count. For future reference here is a tool kit for the evaluation of research studies. Nice easy step by step guide at what you should be looking at when critically reading a topic.

http://www.phru.nhs.uk/Pages/PHD/resources.htm

This is essentially an untested hypothesis not a scientific paper. There are some published papers to back up a few of the elements (very few), but currently it has not been put through any proper clinical trial.

I am really not impressed by any of the referencing included here. I have done a quick search and discovered no effective studies that can back this claim. Currently I wouldn't take any credence in this article, but it is a new hypothesis just mooted this year and you may find that in 5-10 years actual research may be published to prove one way or another.

I certainly wouldn't recommend this as a paper or include in any topic digest.

[henriksdal.livejournal.com]

1. "Stephanie Seneff has a B.S. degree in Biophysics and M.S., E.E., and Ph.D. degrees in Electrical Engineering and Computer Science from MIT"
   
   Her background is varied, but is not biological or medical



2. She's referencing personal projects and using a lot of secondary references that aren't very scientific. In particular, lots seem to lead back to http://www.cholesterol-and-health.com which in turn references something called "Wise Traditons". Googling this reveals that it is the journal for The Weston A. Price Foundation In Wise Traditions For Food, Farming, and the Healing Arts. I don't think Mr Goldacre would be impressed..



3. I strongly think she's formed a theory based on personal opinions and occasional use of a select few accredited papers that back up her theory. Her final paragraph brags about her convincing argument which just makes me less convinced... I don't see the point of this essay other than to crow about some conclusions leapt that have been leapt to. See what Ben Goldacre says about fish oils or mmr and autism. As a highly qualified academic (although in a totally different field) she should be speaking to other highly qualified academics in the field of this sort of study, and persuading them to take on the study.



4. In conclusion: she's not done any actual research other than google lots of things.

I'd love to hear what other people think of this essay!

[akicif.livejournal.com]

Her background is varied, but is not biological

There's the BS in biophysics....

Admittedly biophysics is a broad church, which can mean pretty much anything: from macro-scale mechanics of bones and joints, physics of biological materials, spectroscopic and structural studies of complex biomolecules, all the way down to sequence analysis, you name it.

(I spent not quite a decade in a biophysics department, and there was an unusually large number (well, compared with the chemistry and pharmacology departments I've been in) of folk with not quite testable ideas just outside their areas of competence. To be fair, there were also some dead smart folk, too.)

[despotliz.livejournal.com]

That is a long essay and I'm not sure I can be bothered hunting down all the references necessary to assess the madness of it, so let's go for one paragraph:

"Human biology wants fat not just on the person, but also in the diet, if a pregnancy is in the wings. This fact has been proven quite conclusively by a recent analysis of data from the Nurses' Health Study, an ambitious long-term study involving over 18,000 nurses, which has yielded a wealth of data on issues related to womens' health. Dr. Jorge Chavarro at the Harvard School of Public Health analyzed data on their dietary practices over an eight year period, and looked for correlations with various health issues . Of all the dietary associations that were investigated, the one associating fat in dairy consumption with fertility gave the most striking and statistically significant results. Women who said they ate low-fat diary (e.g., skim milk and low-fat yogurt) increased their risk of infertility by 85%, whereas women who consistently ate high-fat dairy (whole milk and ice cream) decreased their risk by 27%."

The study is this one. This is not my area of expertise, but I can see that the study has controlled for a number of known risk factors for infertility like BMI, smoking, alcohol, exercise, etc, so one point there. A possible minus point is that all the dietary information comes from self-reported questionnaires, but they reference another study which shows them to be reasonably accurate. The impression I get is that this is a reasonable study which has tried to avoid the obvious pitfalls.

The problem is with the 85% and 27% numbers quoted above. The risk of infertility does rise by 85% in women who eat low-fat, if you compare the group of women who eat two or more servings of low-fat per day to those who eat one or less per week, so the two extremes of the range. The 27% figure is for women eating 1+ high fat per day versus 1 or less per week, and comes with a 95% confidence interval of 0.52 to 1.01 and a P value of 0.001, so it's not as strong a result as for low-fat intake, and it's possible the effect is smaller than 27%. Everything in the paragraph I quoted is correct, and there does appear to be a correlation between fat intake and infertility, but I don't think it gives the whole picture.

The other problem is that they assume the infertility is entirely due to the lack of fat in the diet, while the paper mentions several possible options they couldn't rule out, such as the composition of skim milk producing androgenic effects, and conversely the presence of estrogen and possibly other fat-soluble substances in whole milk etc can promote ovulation. In fact, the authors consider this a more likely explanation than fatty acids being responsible for increased fertility: "Although particular fatty acids in dairy products could potentially have a beneficial effect on ovulatory function, a more likely explanation is that a fat-soluble substance present in dairy products may exert this action. Whole milk and other high-fat dairy products have a higher estrogen concentration than their low-fat counterparts."

I suspect that if I were to read the whole thing and fact-check every paper, it might be similar in each case - on the face of it they have backed up each of the steps in their chain of logic with an appropriate reference, but when you look a bit deeper than the reports in the media and maybe an abstract or two, the paper does not actually say what they claim it says.

Oh, and on a further quick skim of the essay they claim that autism has increased alarmingly over the last thirty years, which is thoroughly debunked here.

(edited to fix the bits where I messed up what the paper says)

[endless-psych.livejournal.com]

It strikes me as an update of the debunked idea of "refridgerator parents".

[heron61.livejournal.com]

It's a mixture of total idiocy and likely truth. The autism connection is utter nonsense (as is the entire so-called autism epidemic) [[1]], and the cholesterol info looks little better. Also, too much sunlight is clearly bad for you. OTOH, the info on vitamin D seems to be correct. I've seen evidence from actual medical studies (I don't have any links) of increased Vitamin D intake being useful for reducing cancer rates and similar things. I take vitamin D supplements.

[[1]] In general, any non-standard article about autism that treats it from a nutrition or alternative health perspective is complete junk.

[mouflonmatt.livejournal.com]

It's utter wank from start to finish. I can't be arsed to back up my statements because I am in bad mood. Suffice it to say, it's right up there with pastafarianisms connection of global warming with pirate decline for intellectual validity.

[drjon.livejournal.com]

YARRR!!!!