Hacker News Reader

Last paragraph:

The University of California, Los Angeles, owns a U.S. patent (6,274,119) entitled “Methods for Labeling ß-Amyloid Plaques and Neurofibrillary Tangles”, which has been licensed to TauMark, LLC. Drs. Small, Satyamurthy, Huang, and Barrio are among the inventors and have financial interest in TauMark, LLC. Dr. Small also reports having served as an advisor to and/or having received lecture fees from Allergan, Argentum, Axovant, Cogniciti, Forum Pharmaceuticals, Herbalife, Janssen, Lundbeck, Lilly, Novartis, Otsuka, and Pfizer. Dr. Heber reports receiving consulting fees from Herbalife, and the McCormick Science Institute. The manufacturer of Theracurmin, Theravalues Corporation, provided the Theracurmin and placebo for the trial, funds for laboratory testing of blood curcumin levels, and funds for Dr. Small's travel to the 2017 Alzheimer's Association International Conference for presentation of the findings.

If that were the gp's argument, I'd have no problem with it (heck--I'd agree with most of it).

>> Small study

>Almost entirely irrelevant, provided it's properly powered--and based on the effect size and p-values, it is.

Sorry to state this so strongly but I feel it's important to point out that this is misinformation of the most egregious kind.

There is no examination of statistical power in this article whatsoever so there is no basis for your claim that this study is properly powered.

Large effect sizes are not evidence of adequate statistical power. Quite the opposite is true. E.g. See this article: https://www.nature.com/articles/nrn3475/figures/5

'Effect inflation is worst for small, low-powered studies, which can only detect effects that happen to be large.'

Am I right in thinking that statistical power is the true positive rate, so a test that always rejected the null hypothesis (whether correctly or incorrectly) would have the highest possible statistical power?

No, statistical power as a concept is independent of the observed result of a test.

It is widely misunderstood though, so often people try to use an observed result to justify low power.

http://andrewgelman.com/2017/08/17/just-google-despite-limit...

Is the wikipedia page on statistical power plain wrong then?

"The power of a binary hypothesis test is the probability that the test correctly rejects the null hypothesis (H0) when a specific alternative hypothesis (H1) is true. The statistical power ranges from 0 to 1, and as statistical power increases, the probability of making a type 2 error decreases. For a type 2 error probability of β, the corresponding statistical power is 1-β. "

So according to that, if I always reject H0 whether rightly or wrongly, that means β=0 and power=1.

  Sorry to state this so strongly but I feel it's 
  important to point out that this is misinformation of the 
  most egregious kind.
  
  There is no examination of statistical power in this article whatsoever so there is no basis for your 
  claim that this study is properly powered.

The power is only relevant when you fail to detect an effect. This is a positive study, so the power is simply not relevant to criticizing that particular study.

The nature article you reference talks about literature in aggregate, and the problem created by small studies in general, but it does not allow you to say anything about a single positive study. In fact, the place where those very authors draw their primary conclusion, they state:

  Inflated effect estimates make it difficult to determine an adequate sample size for replication 
  studies, increasing the probability of type II errors

Which means: when you try to replicate it, you might fail because you thought the real effect size was bigger than it is, based on your estimated effect size with small n.

That's well and good, but is a problem for the replicating authors, not for these.

An even more direct point related to the current article: it was preregistered, so you know the authors have not been farming well-designed but small curcurmin studies until they did one than that was positive by chance, and then published it. The act of pre-registration goes a very long way toward addressing the issues raised by Button et al.

> The power is only relevant when you fail to detect an effect.

This is simply not true. It is sometimes called the fallacy of "what does not kill my statistical significance makes it stronger”.

See http://andrewgelman.com/2017/08/17/just-google-despite-limit...

  The power is only relevant when you fail to detect an effect.

This is absolutely false. Small, low-powered studies are more likely to produce false results than larger studies. This was well-documented by John Ioannidis over ten years ago, in work that should be required reading for all publishing scientists:

http://journals.plos.org/plosmedicine/article?id=10.1371/jou...

This is addressed by pre-trial registration. You can't have publication bias (his argument) when you register the study and see that this is the only one.

I know this is a dead thread: I'm answering because if you are a clinician you need to know that your understanding of the statistics is incorrect. This statistical effect has nothing to do with publication bias. If you want to continue the discussion please email me and I can explain and give copious references which will show you why.

it's not turmeric, it's curcumin, which is about 2% of turmeric by mass (so, you'd need to eat ~4.5g turmeric daily to match study doses)

Typical Indian cooking uses turmeric in pretty much everything. I wouldn't be surprised if daily consumption of turmeric is very close to the figures you mentioned.

Interestingly, one of the recent fads I've been seeing has been for "turmeric lattes" - maybe can reach that level even if you're not keen on Indian cuisine?

"turmeric lattes" :)

My kitchen medicine for running nose, soar throat is hot milk with turmeric. Works for me most of the times.

Apparently, eating more than a teaspoon of tumeric a day increases the risk of kidney stones though, so don't overdo it.

I've been mixing it through my oatmeal, together with chilli-pepper and black pepper, a bit of cardamom, and a teaspoon of real cinnamon (the non-cumerin kind). While it feels like it works, I know that the placebo effect is probably much stronger than any real effect these foods have. Still, a placebo effect is still a real effect on my mood, so that's still a kind of health benefit I guess. If nothing else the spices help me wake up!

Tangent regarding health food fads: these days I mainly use nutritionfacts.org[0] to determine which of those are actually supported by the latest nutrition research. It is the only health food/diet website that I know of that directly cites nutrition research and continuously scours the latest papers for new findings - sources are always linked, and quotes are directly lifted from the papers with no modification (so less likely to suffer from "stronger or opposite of what the paper actually says"-shenanigans often seen elsewhere).

Some of the other things I tried out based on that website have too big of an effect to just be placebo: adding blue-berries to a meal really reduces sugar rushes/crashes in the hours after it[1]. Taking a table spoon of freshly broken flax seeds[2] does a lot to counter the rise in blood pressure due to my ADD medication (I measured it, plus I feel a lot less discomfort in my chest area and less jittery).

Last time I mentioned that website here, it was in a discussion of which diets are healthy. Ironically, it was the only comment in the discussion that attracted multiples downvotes without explanation, while everyone else was sharing their unsourced opinions.

[0] https://nutritionfacts.org/

[1] https://nutritionfacts.org/video/green-smoothies-what-does-t...

[2] https://nutritionfacts.org/video/flax-seeds-for-hypertension...

Also:

http://www.ijamhrjournal.org/article.asp?issn=2349-4220;year...

> Worldwide, esophageal cancer is the eighth most common cancer [...]. In India, it is the fourth most common cause of cancer-related deaths.

Granted it's probably due to alcohol and tobacco but curry probably won't save you from cancer.

This is spurious correlation at its best. Terrible science at it is worst.

chewing tobacco is more common, at least, on the uncles I watch.

What prevents researchers from running studies like these on the order of thousands of individuals, as opposed to 40 or 100?

Funding

A lot of things that basically boil down to budget, optimization and space available. If you have a drug that you believe you can show a strong effect with the effect group at ~15 people, maybe you run the trial with 40 (15 e.g., 15 control, plus wiggle room for drop-outs over the 18-month course and extra certainty).

Check the last few paragraphs--"The authors thank Shayna Greenberg, Dev Darshan Khalsa, and Anya Rosensteel for help in recruitment, data management, and study coordination; John Williams, University of California, Los Angeles, Nuclear Medicine Clinic, for performing the PET scans; and Vladimir Kepe, Cleveland Clinic, for developing the FDDNP-PET analysis procedures for proteinopathies in patients with neurodegenerative diseases." Add all that to the effort involved in the actual chemical extraction of the bioavailable curcumin analogue and it just doesn't make sense to add any more people than you need.

For the recruitment, data management, and study coordination piece, sounds like a great area to optimize with software. I'm sure big pharma will also happily shell out for something to help them manage their own clinical trials. Academia PIs can then get the same stuff at a lower rate.

Edit: Heck, you can even lease top grade medical space in all major metro areas and offer it as part of the package in a WeWork style arrangement. I'm sure they are tons of nuances to rolling something like this out, but it sounds like it can cut costs substantially.

Money.

If they're marketing their own drug (as implied by other comments - I don't know), you might also ask what stops them from running the studio on a 100 groups of 40 patients, and picking the "best" for publication... (other than ethics, obviously)...

Cheaper to run it on 100 and select the best 40 to report in your study.

As others have pointed out. Its about money. If the compound is expected to be very profitable, then it make sense to invest a lot money to do a large scale study. On the other hand, if it is a natural compound that you can't really patent, the profitability will be much lower. It doesn't make sense for a commercial entity to do a large scale study. If the compound is important enough, a government entity will sometimes invest the money to study it.

Thank you for your informed comment.

I don't' work in the field so I find it hard to interpret these studies. What's your opinion about the size of improvements in practice?

For example, How meaningful is the 20.3 point difference in Buschke SRT test?

Cohen's d (mean difference divided by SD) above 0.60 with the memory or attention seems good (73 % of the people who take the dose get above the mean test score). But difference that is detectable in a test is not telling me how big the effect would be in the daily performance unless I'm familiar with the test.