Hacker News Reader

Comment by niteshpant | Hacker News Reader

I have thoughts on this

1. Yes, vitamin D actually controls a lot of bodily functions it’s very easily set aside as not a “main” factor when in reality it actually controls a lot

2. This study was done on women in Denmark only which isn’t a great study subject considering Denmark doesn’t get a lot of sun to begin with so most of these women would already start at very low levels

3. This doesn’t directly correlate to women of color because WOC need higher dosage of vitamin D than white women do. The general range of “good” level of vitamin D that doctors tend to use is related to studies results gotten from white people when in reality brown and black people need way more for their range to be at a normal place.

View on HN

Study seem to have addressed point #2 in several ways.

1. They measured maternal vitamin D before supplementation began. They explicitly adjusted for these preintervention levels.

2. the two groups started at essentially the same vitamin D levels.

3. They specifically tested whether baseline vit D status changed the effect of supplementation

For point 2: I don't think there's really good or bad places to study it, it might not generalize to sunnier places but the reverse is also true. Presumably the scientists working on this can understand these things (I know in my field I'm aware that studies in the tropics will find different things than studies in Canada).

For my own point: in this study they have like 22 test values but still use the 95% confidence interval. Even on random data there will be a significant result like a third of the time so I think it's easy to interpret these result as more definitive than they are. Not that it's a bad study though (no study will be everything, baby steps like this are important in science).

I just skimmed the summary but in the results section it reports three significant results and two after “multiple-test correction”. I’m not sure how they did that correction; I would expect that measures of cognitive performance are correlated for each child and so the standard Bonferroni correction might be too harsh to apply here.

> Covariate-adjusted analyses of standardized scores (mean [SD], 0 [1]; higher values indicating better performance) showed positive associations of high-dose vitamin D3 with verbal memory (β = 0.17 SD; 95% CI, 0.03-0.32 SD; P = .02), visual memory (β = 0.24 SD; 95% CI, 0.06-0.42 SD; P = .01), and flexibility or set shift (β = 0.19 SD; 95% CI, 0.01-0.37 SD; P = .04); however, high-dose vitamin D3 was no longer associated with flexibility or set shift after multiple test correction.

My statistics isn't great but reading the study more it looks like they control the rate of false positives via the q values so my initial concern may be unwarranted. I'm surprised that it keeps so many barely significant results with so many hypotheses. I'll have to look it up when I get time.

> Denmark doesn’t get a lot of sun to begin

First, that's only true for about 4 months of the year. Second, people cooped up in offices in China, India, and the US don't get a lot of light either. In fact I'd bet the better work-life balance in Denmark means people actually do get more light there because they spend more of their evenings and weekends outside instead of in the office. Office buildings in Denmark also tend to have much better sunlight by design.

>Office buildings in Denmark also tend to have much better sunlight by design.

Just on this note, you don't make vitamin D if you're indoors even if there's direct sun hitting your skin, because commercial glass filters out UVB.

> This study was done on women in Denmark only which isn’t a great study subject considering Denmark doesn’t get a lot of sun to begin with so most of these women would already start at very low levels

Generally, when a study is done in the US - no one will ever question the location. The moment the study is outside the US, "not US so not generalisable" questions always arise.

to be fair, the US has a lot more variability in climate and population than most countries

> when a study is done in the US - no one will ever question the location

Studies everywhere are now being scrutinized for the participant cohorts because it is now widely recognized that biological differences exist between different groups. Some medications for example aren’t sufficiently studied for effects on women vs men and are being reviewed.

Plus, studies in US are less scrutinized because researchers are aware of the need for a diverse cohort and you are more likely to get one in the US vs elsewhere.

The grandparent explained exactly why it is an issue though. It isn't because US is somehow just magically more legitimate than Denmark.

As they stated, it is because the population of Denmark is very homogenous, as opposed to the US. If you are trying to make a generalization that applies to a range beyond just white people, having Denmark as your sole sample is clearly flawed.

Along the same lines, picking Japan for the purpose of generalizing to wider racial/ethnic groups would also be a bad idea. Not because their research is untrusted/considered non-reputable (it is quite the exact opposite), but because their population is too homogenous.

I don't think it invalidates a study as long as you do things on relative terms and have a control group. Another study can see if the same delta effect is reproducible in an e.g. homogeneous Asian population and report on it.

It is probably a logistical nightmare to do a study of this sort in multiple countries and regulatory systems simultaneously.

It doesn't invalidate the study at all! On the contrary, if you're measuring vitamin d levels from blood tests, it is easy to adjust the dosage to match.

It's just an important factor - if you live much further south or spend a lot of time outdoors, your target dosage will be different than someone in _Denmark_.

> considering Denmark doesn’t get a lot of sun

> As they stated, it is because the population of Denmark is very homogenous

If you know about vitamin D, you'll note that sun exposure is one of the primary reasons location matters for this study. It would be similarly relevant if they only studied students in Miami or southern California.

Essentially: sun exposure helps you create vitamin D, and so you shouldn't naively generalize this study to other lines of latitude

Do they need more vitamin D? I thought they just needed more sun to get the same vitamin D

At these latitudes they just get almost none naturally, that's why you need to supplement more

Ahh, more supplement. I must have misread as more absolute amount

Apparently the pale, semi-translucent skin is a mutation that allowed hominids to live at high latitudes, where darker-skinned hominids would whither due to the lack of vitamins of the D group [1].

That same mutation made them vulnerable to the levels of sunlight at lower latitudes, susceptible to sunburns, etc.

[1]: https://pmc.ncbi.nlm.nih.gov/articles/PMC8359960/

Correction in 2. Contrary to popular believe, sunnier countries in Europe have higher deficiency in vitamin d.

Spain have lower levels of vitamin d than Denmark.

The countries in the northern parts of Europe know this is an issue and thus fortify some commonly consumed food items (milk) with vitamin d.

In places like Spain only some “premium” milk gets this treatment.

Women in the study weren't all starting from very low vitamin D levels

2. Good for me as I live in Sweden. Started only in my 30s to take Vitamin D and multivitamin. My 20s where wasted really. I studied and worked but felt even more shit. Less shit now I guess.

2. Good for us who live in Northern Europe! Or does everything have to be tested in California or MENA or India?

You can probably have the same results in the New England area of the USA, no? Even NY with 10mil people?

3. Okay, and?

Reply on news.ycombinator.com