Eating things that say "hydrogenated" on the label may piss you off.

Questionable merely means we still have lots of questions about it. Hence the 'Question' in the word 'Questionable'. We don't know it's bad. We don't know its good. Even if the experiment is sound, it may have too small a data set to draw conclusions from, or the data set may have been biased (as in the Millikan incident I mentioned), and so on.

 

If I may then paraphrase your definition of questionable as the experiment is sound but the data set is inconclusive, the experiment being discussed in this thread does not meet those parameters. The methodology used could not result in meaningful results towards establishing a causal relationship. The experiment was in short, bad, and the conclusion was irresponsible.

Considering the methodology I can't say with certainty that the individuals behind this report weren't simply making intentionally wild claims just to get published.

Let's be frank here. The experiment being discussed in this thread made a wild claim backed by an experiment that could not possibly verify the claim. If the people behind this were doing so intentionally for attention, how would they have conducted things differently?

"Bad" is also subjective. When someone claims that there is bad and questionable science, both subjective terms, it's helpful for the sake of argument to get them to actually concisely define what each of those are in a mutually exclusive manor.

 

Similar Article

Oh yeah, that's fucking wild scaremongering right there. 11 year trial? Yeah, that shit's for fun.



And lets not forget that they had insanely few individuals - 12000? Pfft, that's the sort of study people use all the time to make dramatic claims with minimal evidence. Worse still, they censored their data set, removing 1700 people because they didn't respond! Who were these people! Clearly dissenters against the steamroller that is Bad Popular Science. Worse - they took people out who were al-read-y depressed. No wonder they only had 12000 people! How are you meant to find out if something increases incidence of depression if you exclude people who are already depressed!

Are you telling me that they were between 50% and 75% sure of how much fat the people had eaten! That's insanely bad, I mean, I could tell you exactly how much fat you ate every time you cooked something! And lets not forget this is what they claimed their accuracy is. These scoundrels are probably hiding a 20% certainty in some of those 1700 dissenters they brutally axe-murdered.

And what about the people who HID their depression, huh? What about those poor souls. *someone comes through a side door, whispers* You mean, that'd make their evidence stronger? *more whispering*

And look at this self serving ass-covering. I mean, anyone would be incited to think that they were - heavens - responsible scientists conducting a reasonable trial in an attempt to understand how saturation in fats (including, mind you, hydrogenated oils, which are saturated fats in essence by the end of it) affects incidence of depression.

 

Now, lets look at the specific study that essence linked, having established that similar studies have shown a correlation - a priori, as well - and that the similar studies have been performed well.

Wow, look at this crap, they only had 1000 people! That's again hardly anything - except lets look at how it actually is, and realise that having 1k people for a trial is actually decent enough to show results, given that you probably only need 5-10 random people to show that most people aren't allergic to peanuts - we want a general trend, not a specific outlier.

Would you believe that? They actually used the results from another trial that measured results. So. Not only are they not actually conducting this trial (which is advantageous, because they're not influencing study results) but they're even using research that might not turn up the outcome they want? It must have been fake data. Except it probably wasn't.

These behavior scales, my goodness, they're exactly like filling in the missing letter of a word. Wait, they're not. My bad.\

Edit: post split for size.

 

You'll notice something insane here - they actually had data both from the control group, and a whole 'nother sample who were taking statins. This gives more context, and the analysis here is fairly solid, showing a non-trivial tendency towards aggression as a result of hydrogenated oils.

Now, lets look at their predictably overconfident conclusion.

Wait, "If the association is causal", what is this, a copout? No, its evidence that the authors consider it to be just one study.

Furthermore, they reference biological plausibility and coherence with other literature, showing that their (not extraordinary, but solid) evidence doesn't back up extraordinary claims, but backs up relatively unexciting, but still relevant claims.

 

Hardly a wild claim. Is it any wilder than me claiming that any number of substances will cause you to hallucinate?

And yet, mere 'surveys' can accurately gauge public perception with a high degree of accuracy - further, actual experiments aren't nearly as strong as you'd hope - how do you quantify aggression based on physical measurements.

And the scientists, funnily enough, agree.

Except the whole study is providing the evidence, lol.

You should.

The study (especially the first one) is emphatically not a joke. The issue here is less that the paper is weak, and more that there exist weak studies, and hence you assign those weak studies a greater percentage of total studies than is true, and allocate the label of weak to studies that you want to see as weak, for whatever reason.


I feel fairly confident that my review of two articles discussing the affect of fatty acids on the brain should provide sufficient evidence (quoted) and far too little analysis (I don't want to write a thesis on this or anything) that these studies aren't complete crap. Has anyone else actually read through the whole thing and looked at the published data? I mean, sure, its plausible that they're lying - but its just as plausible as the entropy implies the universe literally just came into existence just then theory.

 

I liked the part where you brought in another entirely different study, lol.

Anyway, neither study actually discusses the effect fatty acids have on the brain as that would require establishing a causal relationship. The problem with human minds is we confuse correlation and causation, hence this thread.

There are so many explanations that could explain a weak correlation here beyond fatty acids causing depression or aggression that it boggles the mind. The use of the word "may" is so generous it's like as you said the universe may have just come in to existence.

 

Well, I was hardly going to argue the merits of a theory based on one piece of evidence, was I.

That'd be because it was well beyond the purview of any individual study to conclude a causal relationship. Given the immense complexity of the human body, that's something we likely will never be able to prove, but we can at least use it as something we're X% confident about. To consider this, lets look at Bayes' theorem:

Say, I was initially ambivalent about the chance of transfats affecting behavior: Say I figured aggression was random with respect to transfat consumption.

I would say that my initial estimate of the probability of someone eating transfats becoming more aggressive was 50%.

Now, having read this article, the probability of them finding this evidence if transfats have a role in causing aggression is, say, 90% - there will be correlation. In contrast, if transfats had no role in causing aggression, finding correlation is still feasible - lets say 60%. More often than not, but not as often as if it were.

My new estimate for the chance of transfats affecting behavior is (0.5*0.9)/(0.5*0.9+0.6*(1-0.5)) = 0.6.
So, I now estimate that the chance of eating transfats affecting behavior is now 60%. While this study doesn't improve it to the nearly 95%+ or 99%+ that we would like, its a start. If we than had more and more studies like this, each time the probability will gradually increase (or decrease) until we find that it has 1:1 causation (plausible), 1:1 correlation (highly plausible), no correlation (unlikely) or 0 correlation (doubtful in the highest.)

Except they both discuss the effects of fatty acids on the brain - in one case, its a candidate for causing depression; in the other it's a candidate for causing aggression.

A proposed cause for this would be

"One mechanism may be trans-fats' substitution for docosahexaenoic acid (DHA) levels in the orbitofrontal cortex (OFC)."

Note: The correlation is not all that weak. It's sufficiently strong (especially when various variables, e.g. calorific intake are controlled) to indicate a correllation between two variables. If you actually read the conclusion:

You will most likely note that they specifically mention it's an if its causal. There is some plausible biological mechanism, however there would need to be more studies building on the results of these to show that it is causal. However, in particular, in the first study: By considering 12k individuals,

 

Wow. Mining. Exceptional walls of text there. I can hardly read it all without getting dizzy. I think I may spontaneously combust if I were to try writing that stuff. Good job.

If I had a Badass list in a thread somewhere, you may well be the first on it. I still have not come up with a viable argument for the last few posts you made against my position on various topics.

 

^^ This. Very impressive, Mining, and thank you.

 

No, just no.

Hydrogenated oils are fucked up unsaturated oils that are out of alignment, the only thing they are reasonably close to saturated fats in is their firmness/rigidity(staying solid at higher temperatures). They replace other oils in cells and thereby overly harden the cell membrane and interfere with passing of nutrients. They cause arterial hardening, aka atherosclerosis, and are also linked to cancer(because they're usually ionized aka free radicals).

 

Hrmm, didn't know the biological context. I was mostly thinking in terms of chemically - a fully hydrogenated oil is equivalent to a saturated fat.

 

Still no. Hydrogenation is nothing like saturation. Hydrogenation takes a kink in the structure of an unsaturated fat and uses artificial means to straighten it out, but the only way that hydrogenated fats are "like" saturated fats is that they're both straight -- in every other respect, a hydrogenated fat is still unsaturated.

 

Transfats aren't truly straight, they're zigzag.

Top is unsaturated, left is saturated, right is transfat.

That's for a monounsaturated transfat. Polyunsaturated transfat zigzag for each out of alignment bond.

 

Thanks, Aegho!

 

Ok, right, so transfat is when its not actually hydrogenated, but undergoes a change as the result of the process, raising its melting point but not its level of saturation.

 

Except that when they say on a food label that an oil is "hydrogenated", they are always referring to the process Aegho is referring to. There's only one naturally occurring trans fat; it's in milk and we've already studied it and found out that the body knows what to do with it so it's not a threat like the artificially hydrogenated oils are.

Its kind of like how the cops call all drugs 'narcotics' even though doctors know that only opiates are narcotics.

 

I was under the impression (clearly mistaken, as it seems) that hydrogenated oils were fully hydrogenated.

 

Correction: You're right, Mining. Turns out that there are some products that specifically list "fully hydrogenated" fats -- those are what you're thinking of. But unless it actually uses the word 'fully', its a trans fat. Partially or just plain hydrogenated are equally bad.

 

Also note that the body has specific uses for each naturally occuring fat. Even fully hydrogenated fats(aka unsaturated fats that have been artificially saturated) might not be that good for you, the body doesn't necessarily have a normal use for them. Also ALL processed unsaturated fats are rancid, as a result of the process(mainly the heat, but the hydrogenation process also does this), which is why they also undergo a deodorizing chemical process, that they are rancid means they are ionized/free radicals and thus carcinogenic.

To be clear: every margarine, and every cooking oil you see in stores is rancid unless they are expressly marked as either extra virgin(olive oil), or expeller pressed(linseed oil for example). Because all other extraction methods employ high heat, which makes fat rancid(particularly unsaturated fats, and particularly polyunsaturated). Oh and the worst offender is canola oil.

The best cooking grease is coconut oil(98% saturated), followed by palm oil, lard, and butter(lard is the best for high temperature cooking because it has a higher flashpoint).