Live data from Hacker News

Long-Awaited Study Results on Vitamin D and Fish Oil Supplements

npr.org

51–60 of 248 posts

Re: Long-Awaited Study Results on Vitamin D and Fish Oil Supplements

#51
post #11

This headline is quite misleading (and arguably very biased toward white readers). It describes a marked improvement in outcomes for African American study participants. So perhaps NPR views this as "disappointing" but the headline suggests broadly that everyone should stop taking those supplements, when in fact the results are far more nuanced. HN's absurd title policy is using a clickbait (and misleading) title cho…

I don't know, the linked editorial in NEJM summarises it pretty well:

>However, these “positive” results need to be interpreted with caution. First, there was no correction for multiple comparisons, as would be required to attenuate the chance that these are spurious results, owing, in part, to the number of secondary end points. Second, these putative effects have not been consistently observed across other large, randomized trials of n−3 fatty acids.5,6 However, in one of the trials of vitamin D supplementation, post hoc analysis showed a lower incidence of cancer among women who received active supplements than among those who received placebo, excluding women who had withdrawn or in whom cancer had developed in the first year of the trial.13 Finally, the medical literature is replete with exciting secondary end points that have failed when they were subsequently formally tested as primary end points in adequately powered randomized trials. Thus, in the absence of additional compelling data, it is prudent to conclude that the strategy of dietary supplementation with either n−3 fatty acids or vitamin D as protection against cardiovascular events or cancer suffers from deteriorating VITAL signs.

https://www.nejm.org/doi/full/10.1056/NEJMe1814933

Re: Long-Awaited Study Results on Vitamin D and Fish Oil Supplements

#53
post #39

Why do we keep looking for some magic pill and ignore diet and exercise, especially when we talk about cancer and cardiovascular disease? Expecting a pill to counteract a sedentary lifestyle and a diet high in sugar and carbs is lunacy.

Show me a randomized control study of healthy people with lower all cause mortality from either diet or exercise, and I'll gladly change my ways.

This comes across like the people who say, show me incontrovertible proof God does not exist and I will happily become an atheist.

Knowing, of course, it cannot be proven.

If you are actually of open mind, there are some natural experiments that have been done, e.g. see Blue Zones.

Re: Long-Awaited Study Results on Vitamin D and Fish Oil Supplements

#54
post #11

This headline is quite misleading (and arguably very biased toward white readers). It describes a marked improvement in outcomes for African American study participants. So perhaps NPR views this as "disappointing" but the headline suggests broadly that everyone should stop taking those supplements, when in fact the results are far more nuanced. HN's absurd title policy is using a clickbait (and misleading) title cho…

From https://news.ycombinator.com/newsguidelines.html

>Otherwise please use the original title, unless it is misleading or linkbait; don't editorialize.

Re: Long-Awaited Study Results on Vitamin D and Fish Oil Supplements

#55
post #48
post #26

Earlier quoted context omitted.

Does not matter. The effects on African American populations were probably not part of the pre defined endpoints, so yes, it is perfectly acceptable to say the study is negative. Looking at subgroups a posteriori is essentially what leads to p hacking and this is the behavior that we want to discourage in sciences.

Shouldn't one consider what the p-value is in weighing whether it is a fluke or not?

In theory, you can do some more statistical tests to study if the subgroup analysis has statistically significant results.

In practice, those tests are probably going to be skipped or done improperly, and it's difficult to tease out if the subgroup analysis suffers from the statistical fraud of reanalyzing until you get a positive result. Redoing the test from the start with the subgroup as the primary endpoint would always get you more accurate data, and history has pretty clearly shown that most instances of "we looked at the subgroups and found something where it works!" tend to not actually work when done properly anyways.

Re: Long-Awaited Study Results on Vitamin D and Fish Oil Supplements

#56
post #34

Given these studies. I will stop taking fish oil. However, in light of further evidence by gwern as presented below, I will continue to take Vitamin D, even though my confidence in it is lessened. https://www.gwern.net/Longevity

It's one study, and it was only done on people over 50. But yeah more science needs to be done.

Re: Long-Awaited Study Results on Vitamin D and Fish Oil Supplements

#57
Study linked was not the biggest cardiovascular health news of today--

This was: https://www.nejm.org/doi/full/10.1056/NEJMoa1812792

A single purified component of fish oil - EPA, did lead to a 25% relative risk reduction on top of statin treatment. This is a huge result with off the charts p values.

Re: Long-Awaited Study Results on Vitamin D and Fish Oil Supplements

#58

Earlier quoted context omitted.

Fairly recent Cochrane review concludes: > ... we do not have sufficient high quality evidence to determine the effects of [omega-3] as a treatment for [major depression]. Our primary analyses suggest a small-to-modest, non-clinically beneficial effect ... compared to placebo; however the estimate is imprecise, and we judged the quality of the evidence on which this result is based to be low/very low. https://www.coc…

> If the answer was easy we'd have found it, and we'd tell you about it. Like washing hands before dealing with women in labor? smh Also, enough with the "we", student doctor . You're not part of some privileged class of geniuses out to educate the rubes, on HN of all places.

We've come a long way since Dr Semmelweis and his discoveries prior to germ theory. I'm studying at the university that Barry Marshal was at when he discovered Helicobacter pylori - he was also initially mocked.

I don't claim to be a genius, and I don't see how using the collective pronoun is a problem. I'm just moderately persistent, and I do actually care about people and their health, and advancing our understanding. I'm also a scientist, with a not-bad-for-a-student publication record: https://scholar.google.com/citations?user=x-VF3w4AAAAJ

Re: Long-Awaited Study Results on Vitamin D and Fish Oil Supplements

#60
post #30

"At this point, if one is thinking about supplementation, either omega-3s or vitamin D, talking to your physician or healthcare provider is the next step" Does anyone actually talk to their doctor before taking a vitamin? What information could they possibly provide

I have a terrific new naturopathic doctor who actually cares about my health and will run blood panels frequently and let me know what lifestyle changes and supplements would help me. I never knew before that I was deficient in zinc. Before her though, most of my doctors were of the bland "if you aren't in extreme pain then it's not a problem" camp.

Naturopathic "medicine" is complete nonsense.

You are literally bloodletting your money.

Good work. I thought HN readers might be smarter than that.

Post reply on HN