Sex differences observed in effects: Taking vitamins 'cuts risk of Covid infection by 13 per cent'

[ChristinaXYZ]

www.telegraph.co.uk/news/2021/04/19/taking-vitamins-cuts-risk-covid-infection-13-per-cent/

"Taking multivitamins may reduce the chance of getting a Covid-19 infection by 13 per cent, at least for women, a new study looking at over-the-counter remedies suggests.

Researchers from King’s College London, who have been using an app to track people in Britain with symptoms of coronavirus since the start of the pandemic, found that simple supplements appeared to play a role in the risk of testing positive.

As well as multivitamins, taking a probiotic was found to lessen the risk of an infection by 14 per cent while regular omega-3 lowered the chance of coronavirus by 12 per cent. Vitamin D also lowered the risk by nine per cent.

However Vitamin C, zinc and garlic supplements were not linked to a lower risk.

When the researchers looked specifically at sex, age and weight (BMI), the protective effects were observed in women of all ages and weights but there was no clear association for men.

Researchers said that the sex difference may be explained by biological differences in the immune system of men and women.

The same overall patterns were mirrored in both the US and Swedish subscribers of the app although the effects were greater, possibly because of greater underlying deficiencies in the populations.

In Sweden multivitamins were found to reduce risk by 22 per cent, probiotics by 37 per cent, omega-3 by 16 per cent and vitamin D, 19 per cent.

In the US multivitamins were found to reduce risk by 12 per cent, probiotics by 18 per cent, omega-3 by 21 per cent and vitamin D, 24 per cent.

The findings were published in the journal BMJ Nutrition Prevention & Health, however the researchers warned that trials were needed before any recommendations would be made...." Article continues discussing process of investigation etc.

It could be that those who took these suppliments were otherwise move health conscious or more likely to be limiting social contact?

I take probiotics and I'm a mess

I agree that you'd need randomly assigned groups to see if it's causation or correlation.

Yes, it's like studies on diets, it's almost impossible to sort out what is causation and what isn't. But it's interesting for further research.

I wonder if the difference might be that women are more likely to be deficient in various areas?

I dislike much of Taubes' recent work (reasons) but he wrote an NYT piece years ago with an excellent section on the 'bias of compliance' with a fine discussion of 'healthy user bias.' It is a very satisfying narrative account of why some people always do better than others in studies (and, general, I'd expect a disproportionate number to be women in default of any plausible mechanism of harm).

www.nytimes.com/2007/09/16/magazine/16epidemiology-t.html?partner=permalink&exprod=permalink

It's an Open Access journal so if anyone would like to look at the study it's not behind a paywall:

nutrition.bmj.com/content/early/2021/03/22/bmjnph-2021-000250

Presumably epidemiologists are going to be busy for decades studying COVID 19.

Good to see clear statements about biological differences in men and women, and their potential impact on healthcare. Ridiculous that this isn't the default of course.

I've some sympathy with this response:

Commenting for the Science Media Centre, Kevin McConway, emeritus professor of applied statistics at The Open University, said setting up such a trial would be expensive as well as difficult, as infection rates fall due to vaccine rollout. "I wouldn't expect to see clear results from randomised controlled trials of these or other supplements any time soon, if ever," he said.

Interesting. Agree about Taubes, embarrassing, I also have reasons

I too would expect that people who take supplements are much more likely to be health conscious, so the findings may all be down to people reducing their exposure risk.

However, it is really heartening to see researchers recording and looking at the difference in results for males and females (and by age and BMI).

In Sweden multivitamins were found to reduce risk by 22 per cent, probiotics by 37 per cent, omega-3 by 16 per cent and vitamin D, 19 per cent.

So if you take all of the above you can reduce the risk of catching Covid by 94%.

Nice.

@doublehalo

In Sweden multivitamins were found to reduce risk by 22 per cent, probiotics by 37 per cent, omega-3 by 16 per cent and vitamin D, 19 per cent.

So if you take all of the above you can reduce the risk of catching Covid by 94%.

Nice.

I don't think that's how numbers work.

Fascinating article, embarrassingadmissions, thank you

I don't think that's how numbers work.

Actually, I think that's exactly how numbers work.

No, those are probabilities, which are not usually additive.

If it were as you say = 94%, then it would be that if they found that say mask wearing reduced the risk of infection by another 6%, you would then reach 100% and you could be certain of not getting the virus! And if they found also that taking let's say B12 reduced it another 7%, you would have 107% where perhaps contact with you would be enough to rid someone else of the virus a la the idea of the "king's touch"? Consider.

If you have 100 people and the number are reduced by 20% you end up with 80, reduce by 20% again and you have 64, a futher 20% reduction get you to about 51. Take way another 20% and you have 41 and one last 20% and you are down to 33. So 5x20% reductions does not equal 100% reduction.

I'd imagine that those taking multivitamins probably wouldn't then also take vitamin D on top (you're not supposed to if the multivit contains vitamin D which they usually do in my experience).

Good to see sex differences spelled out. They are important.

Also the immune system is complex and I'd agree that these probabilities will not be additive.

@IloveJKRowling

I'd imagine that those taking multivitamins probably wouldn't then also take vitamin D on top (you're not supposed to if the multivit contains vitamin D which they usually do in my experience).

Good to see sex differences spelled out. They are important.

Exactly - those variables won't be independent of each other. Those who take vitamins are more likely to take probiotics.

I've met people with unhealthy habits (shite food, boozing, late nights etc) who popped vitamin pills "to make up for it". Lifestyle data must be a nightmare to unpick. I'm also minded of the bit in a Morse novel where the Inspector is asked by his doctor for his average weekly booze intake; Morse estimates it, then gives half the number....

I wonder if the recent Women's Health survey in England will reveal any COVID discrepancies?

I'd imagine that those taking multivitamins probably wouldn't then also take vitamin D on top (you're not supposed to if the multivit contains vitamin D which they usually do in my experience).

The amount of vitamin D in a multivitamin is generally 5 micrograms which may not be enough for everyone in the winter.

@doublehalo

I don't think that's how numbers work.

Actually, I think that's exactly how numbers work.

I have two coins. I need to get one head. I flip both. I can get tails head, head tails, head head, tails tails. My chance of getting a at least head is 75% not 100% even though each coin has a 50% chance.

Just to say that authors describe an adjustment in the statistical analysis (Bonferroni-correction) to account for some of the above discussion about potential confounding etc:

All p values presented were two-sided, with statistical significance determined by the Bonferroni-corrected threshold of significance (p=0.05/7=0.007). Statistical analysis was performed using Stata V.12 and ExeTera, a Python library developed at KCL to clean and process the raw dataset

nutrition.bmj.com/content/early/2021/04/20/bmjnph-2021-000250