[jrauser]

Did you even read the headline? “Supplements won’t prevent dementia...”

The first half of the article is all about various dietary supplements and how there is no good evidence for them. But there is decent evidence for exercise, controlling blood pressure, and congnitive training.

[andrewla]

I mention the supplements as a comparison. Like the nonsense peddled here, the techniques (increased physical activity, blood pressure management, and cognitive training) have shown preliminary promise. But all three (I boldly predict) will fail on further testing, and even now, as the article itself notes, have failed to produce results that exceed statistical significance.

This is irresponsible reporting that is falling for the exact same magical thinking that led to the hype around supplements. Because attempts to prevent dementia have been previously shown to be without merit, we should report on new ideas with a great deal of skepticism until the evidence becomes very strong.

[hmahncke]

The field of cognitive training has moved way beyond preliminary promise.

Here's a meta-analysis of 97 published randomized controlled trials, showing reliable improvements in cognitive function and transfer to real-world function [1]

Here's an NIH-funded study of more then 2,800 older adults followed for 10 years showing that speed-of-processing training reduces the risk of dementia by 29% [2]

Disclosure: I work at a cognitive training company.

[1] PDF link: https://www.researchgate.net/profile/Catherine_Mewborn/publi...

[2] https://www.sciencedirect.com/science/article/pii/S235287371...

[andrewla]

From [2], above: (hazard ratio [HR] 0.71, 95% confidence interval [CI] 0.50–0.998, P = .049)

Yowzers -- just squeaked by on that one, didn't they. Given the unknowns around dropouts, I would be frankly astonished if these results stood up. It mainly looks like the "speed" category dubiously "benefited" from a higher death rate; the ~20 excess deaths in that subset can neatly account for ~2 additional cases that drives significance to random-chance levels again.

In addition, the meta-analysis seems to indicate that only memory-based interventions work, while the individual study linked showed differently. Hopefully as pre-registration becomes mandatory it should become clearer how many of these studies were abandoned because of null results.

[bigmit37]

Can you recommend a book or course that can help me understand stats on clinical trials (if not a stats book with emphasis on clinical trials)?

I feel like I am of those that can easily be fooled by misleading claims.

[hmahncke]

From [1] (the meta-analysis), "Working memory interventions proved most effective (g = 0.479), though memory, processing speed, and multi-domain interventions also significantly improved cognition."

From [2] (the crash results), deaths and drop-outs were handled by adjusting crash rates to total person-year exposure times. In this study as a whole, non-completion rates were not significantly different across groups.

Yes, pre-registraton is an excellent idea.

[rajadigopula]

> we should report on new ideas with a great deal of skepticism until the evidence becomes very strong

A nutritionist with Ph.d with good credibility claims that he got good results with using black sesame seeds. The way he used it is - roast 2 spoons of black sesame seeds and eat once a week for 6 months (no need/should not take every day). He claims to have cured them in 6 months. If I have someone I know I would try as there is no harm in having a regular food item taken once a week. Don't know how it works or if any studies already done to back it up.

[gus_massa]

Great! Until he does a double-blind pre-registered randomized study (with a control group) and it is published in a serious peer review journal (and reproduced a few times by independent research groups), it should be treated like snake oil.

The harm is indirect, because they are mixing real medicine wit pseudoscience. There are a lot of miracle "cures" that appear every month. Many of them steal the patients or the family money, other steal just time and hope. And in some cases, people following the alternative cures avoid following the standard treatment that is better ot at leas reduces the suffering.

[rajadigopula]

May be you forgot the context I mentioned it in. We have to look for alternatives when there is no definitive one present for the time being.

He's been curing patients for past 20 years with diet changes alone. So, I am impressed with his results and shared his idea. Doing a study on it and not using it until you study is your problem. In other words, I sense your worry of how your medicine will react since it's posion (full of unnatural chemicals) compared to a natural food item in it's unaltered form which is being used in daily food consumption in many cultures. Are you suggesting people taking medicine to stop eating food on a daily basis? I never suggested to stop any medication.

I am more than willing to contact him, will you be able to fund the study?

[gus_massa]

> He's been curing patients for past 20 years with diet changes alone.

From your other comment:

> He claims to have cured them in 6 months.

I'm not sure how long it takes to transform a 6 month treatment to a paper. I guess 1 year before for designing, filling all the forms and preparations. And I guess 1 year after to process all the statistics, write the results in a nice form, fighting with the referee/editor. So in 2.5 years it could be published. Let's round it to 3 years because it's not my field of study. [1]

So in 20 years he has enough time to publish a serious paper on a cure an a popular illness that has a lot of people working on it. (Perhaps if the evil Big Pharma is blocking his study, he can replace the pre-registration to a single big announcement in a blog post an a hash of the post in the bitcoin blockchain. It's not official, but any good registration is enough. Most of the other steps, like double-blind, can be made anyway.)

[1] In Physics the typical time is 3 month, but there are very few paperwork because the study doesn't use humans. In Math the typical time is more than 1 year, perhaps 2, because math is forever and the review is proportional.

[rajadigopula]

I totally agree. It will be beneficial for everyone.

[pmiller2]

Where’s his paper in Nature or some other high profile journal? Even as a preliminary result in mice, this would have been big news. The lack of said news means something is wrong with the study.