Hacker News new | past | comments | ask | show | jobs | submit login
[flagged] Foods Linked to Better Brainpower (health.harvard.edu)
39 points by DoreenMichele 72 days ago | hide | past | web | favorite | 9 comments

This reads like a BuzzFeed listicle. Vague references to unlinked studies, with no critical analysis. It's about as genuinely empirical as asking the local witchdoctor.

Edit: And it seems they're trying to sell some ebook "nutrition" guide for $20 a copy. This is garbage.

Correct me if I'm wrong but it was a couple of Harvard "scientists" who were paid by Big Sugar to put a knife in the back of cholesterol.

Can these help you? Yeah, probably. However, what's doing more harm - than these might do go - is sugar, soda, processed food, and other assorted dietary garbage that so many people are fond of.

That is, it's bad food that doing the harm. Eating good food might help, but the better question is: how much kale must you eat to break even from that can of soda?

The canonical link on this web page is:


Rarebirdinc.com "specializes in internet marketing, website design, e- commerce development, strategic marketing..."

For people with interest in nutrition and the impact it has across both body and mind, I suggest checking out www.examine.com. It's the best site I have come across online when I need any science backed knowledge on nutrition.

Cocoa should be on the list. There's evidence that it enhances cognitive ability through multiple mechanisms. It even appears to reverse some of the effects of sleep deprivation. See https://www.frontiersin.org/articles/10.3389/fnut.2017.00019....

> A 2015 study from UCLA linked higher walnut consumption to improved cognitive test scores.

Never trust "a study", especially without a specific citation. Here's the study in question[0].

FWIW, the study was partially funded by the California Walnut Commission[0]. This doesn't mean the science is bad, but probably worth double-checking the methodology. I have no evidence that it's bad. I just don't trust any single study.

Did they just examine NES2 outcomes, or did they try dozens and publish the ones with significant effects? Did they control for income? Is this the only study of walnuts and cognitive outcomes, or have we done dozens--published or otherwise--and are just picking the one with an interesting/desirable outcome?

This isn't just the walnut study. The "Berries" section talks about "a study" (still no citaiton, but Google knows[1]). Was this a study of berries, or a study of hundreds of foods and they published the one that had interesting outcomes?

We do this stuff wrong all the time. Here's a nice video about one Very Bad study that made the news lately[3]. TL;DR: They looked at numerous outcomes from thousands of variables and published one that was statistically significant.

Relevant xkcd: https://m.xkcd.com/882/

[0] https://www.prnewswire.com/news-releases/new-ucla-research-s...

[1] https://www.ncbi.nlm.nih.gov/pubmed/25732213 -- Sci-hub is timing out for me when I try, but this is the URL: https://sci-hub.tw/10.1007/s12603-014-0569-2

[2] This one: https://www.ncbi.nlm.nih.gov/pubmed/22535616

[3] https://www.youtube.com/watch?v=yQrBblRuQFo

Surprisingly short list, there should be a lot more here.


How is this pseudoscience / anecdote loaded page relevant here?

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact