Research Alert! Omega-3 Index & COVID
OmegaMatters: Episode 24
Hosts: Drs. Bill Harris & Kristina Harris Jackson
Background and Key Takeaways:
In this episode, Drs. Harris and Jackson talk about a recently published study on the Omega-3 Index and COVID. Published in the American Journal of Clinical Nutrition on March 1st, the study showed those with a higher Omega-3 Index are more protected from severe COVID. For more information on OmegaMatters, visit: https://omegaquant.com/omegamatters-broadcasts/
Dr. Kristina Harris Jackson: Hello and welcome to OmegaMatters. I’m Kristina and this is Bill.
Dr. Bill Harris: Hi.
Dr. Kristina Harris Jackson: Today, we are talking about a new paper, sort of hot off the presses, but it’s a paper published in 2023 in the American Journal of Clinical Nutrition and we have a lead author right here. The title is Association Between Blood Omega-3 Fatty Acid Levels and the Risk of Coronavirus Disease 2019 in the UK Biobank. We did a little teaser in OmegaMatters Episode 19 on the UK Biobank data that we have access to now, but this is fully peer reviewed published paper that we are going to get into.
Dr. Bill Harris: Great. This was exciting because this was really the first paper that we’ve done with a disease outcome in the UK Biobank and it worked out very nicely, so I’m glad to see it. Let’s just start off with an interesting observation made by this group Vivar-Sierra et al a couple of years ago. They were just looking at the reported intake from a variety of data sources of Omega-3 intake around the world. They then looked at COVID-19 mortality and the countries that they looked at or the regions that they looked at are shown over here and the number of countries in each region, they just averaged the supposed Omega-3 intake along the bottom axis and they had the regional fatality rate here and they showed an interesting trend that the higher the Omega-3 level, the lower the risk of death from COVID.
Well, of course this could be due to a whole bunch of different things, but it was an interesting step in the right direction and suggested that there may be something that the Omega-3s are doing. Just to point out some old data, this is now from the Framingham study some years ago where we looked at correlation between the Omega-3 index and a lot of inflammatory markers just to make the point. There’s 10 different inflammatory markers that we looked at and we found in multi-variable analysis comparing the Omega-3 index with the levels of these markers, and this again, is not in sick people. This is not in COVID individuals. This is in healthy people. There’s a negative correlation, see the negative sign on these R values, for all and they’re all statistically significant.
The higher the Omega-3 level, the lower the inflammation is the message here. It fits that obviously, COVID is a hyper inflammatory response disease and so it makes sense that the Omega-3s might play a role. Just one more little paper for background. This is a meta-analysis of randomized trials where people were given Omega-3 or placebo and looking at C-reactive protein, IL-6, TNF-alpha. All in all, there’s a significant effect of Omega-3 in lowering CRP, significant reduction in IL-6 and certainly a strong trend, almost a significant reduction in TNF-alpha. Not only is it associated in observational studies, but intervention studies also show on writ large that these inflammatory markers are lowered by giving Omega-3. Let’s get to the paper. This is the one you want to talk about, right?
Dr. Kristina Harris Jackson: That’s the one.
Dr. Bill Harris: That’s the one. As Kristina mentioned, I think we announced the findings of this paper some four or five months ago when we originally got them and then we submitted them for publication and they were finally published on March 1st I believe, just a couple of days ago. Again, we’re looking at the association between blood Omega-3 fatty acid levels and the risk for coronavirus in the UK Biobank.
What did we do? As our exposure here, we use plasma DHA levels. Now in the UK Biobank, the data on plasma fatty acids is quite skimpy. There’s a lot of it, meaning a lot of people have been measured but they only give DHA and total Omega-3 and linoleic acid and total Omega-6 and a few total saturates, total monos, things like that. It’s quite a limited look, but DHA level is available and we know from other studies we’ve done that it correlates very well with the Omega-3 index. We did the study using just the DHA level. DHA really constitutes about 80% of the red cell Omega-3 EPA / DHA pair.
So it’s a good marker of Omega-3 status. I wish they’d done the Omega-3 index but they didn’t. We were looking at three outcomes with COVID — 1) death from COVID; 2) being hospitalized with COVID with a confirmed infection; and 3) a novel one I think, testing positive for COVID. Of course, we could only do that among people that were tested and not everybody had been tested in this cohort. We had that limitation. We did this for really just about 15 months of follow-up from January 1st, 2020, which we called the beginning of the COVID pandemic, and the data we had available to us was through March 23rd, 2021. That’s probably good because this was right at the beginning when vaccines were starting to come into play, so we want to look at this outside the context of vaccines and we didn’t have vaccine status on people in the UK Biobank anyway, so we couldn’t do it.
Quick rundown of the flow chart, how we did this study. There were around 500,000 people enrolled in the UK Biobank in the whole suit to match. Basically 75% of them don’t have any fatty acid data because a random 25% were analyzed. We have about 18,000 data points with fatty acids. We have about 7,000 people who died before the pandemic, so they don’t count. They’re out. That leaves us 111,000. Then we had a few missing some relevant covariants. The 110,688 is what we really used for our analysis of hospitalization and death from COVID. About 84,000 of those people never got tested for COVID, so we couldn’t do the tested positive question with them. We took them out and that left 26,000 people, still a huge sample size to ask the question about whether you tested positive for COVID or not. That’s the way it flowed here. Here’s the big picture, the group of patients. Here’s 110,000 total that we did most of our outcomes on about 50% male, almost all of them white in the UK Biobank.
Interestingly, 15% of people who were tested for COVID tested positive within that window. Though out of the 110,000, 838 were hospitalized, so that’s 0.8% and 0.2% of people out of the 110,000 died from COVID. Only 235, so it really is not a very large sample size for death as an outcome. Pretty good for hospitalization. All the variables in yellow were used in the statistical model. We adjusted for all of these things.
The Townsend Deprivation Index is something people don’t typically know. That’s basically a measure of how deprived you are. It’s a combined metric. It’s a socioeconomic status metric. If it’s a negative number, you’re less deprived than average. It’s a little bit weird to think about it. Negative number means you’re better off socioeconomically. Positive numbers mean you’re more deprived than the average person in the UK. All these things were included. Any questions, Kristina, I’m doing all the talking here.
Dr. Kristina Harris Jackson: For the positive tests, could there be more than one positive test per person or is that known?
Dr. Bill Harris: I think all you had to have is one positive test to be considered positive and by that time, there really wasn’t that much testing going on. That came on a little bit slower. So, want to get to the outcomes fairly quickly here. Again, as far as testing positive, and all of the figures are going to look like this, we have the absolute percent testing positive and the average was about 15%. We’re broken down by quintiles of DHA levels, and so the quintiles of plasma DHA, Q1 through Q5, and then in parentheses, we have the median estimated Omega-3 index for each of those quintiles, and we have this from a different study where we’ve analyze the blood and so, the lowest quintile, we call it the estimated Omega-3 index, the eO3I, was 3.5% in this lowest quintile and was about 8% in this upper quintile, which fits very nicely with our 8% target for being the thing to aim for.
Dr. Kristina Harris Jackson: So at quintile four at about 6%, a little more than 6%, that’s where it’s first significantly different than the first quartile?
Dr. Bill Harris: Oh, 6% Omega-3 index?
Dr. Kristina Harris Jackson: Yeah.
Dr. Bill Harris: Yes.
Dr. Kristina Harris Jackson: Okay, but it gets even more significant the higher the number goes.
Dr. Bill Harris: You see because the more significant there, and if you look per standard deviation of DHA, each standard deviation was associated with 8% lower risk, so if you went out two standard deviations, you’re 16% lower risk.
Dr. Kristina Harris Jackson: Just for testing positive.
Dr. Bill Harris: Just for testing positive, which is something we’ll come back to and say not something I necessarily would’ve predicted. There’s no mechanism where Omega-3 would necessarily prevent you from just actually getting sick per se. Hospitalization, the same story. Very much a stair-step dose-related relationship between Omega-3 levels, DHA levels here in this case, and risk for being hospitalized. Again, there’s only 838 people here, so it’s harder to get statistically significant when you split them into five groups, but when you look across the whole spectrum, it is statistically significant 0.01 trend across this — highest to lowest across the quintiles. For dead from any cause or dead from… Basically dead with COVID. We don’t really know if they’re dead because of COVID. We know they’re dead with a COVID diagnosis. That’s all you can really get out of these databases. Still, certainly for the first four quintiles there were very nice monotonic drop in risk for death till you hit the fifth quintile and then it goes back up, same as it was in the third quintile.
That bump up, it’s about 30-ish people in each column, so it’s not very many people. Part of the problem. Not sure why this bumped up, but it made the whole curve not statistically significant, but Q4, quintile four, was significantly different than quintile one at a 0.1% versus 0.3, over 50% lower risk for dying if you were in this quintile. Certainly, a strong relationship between higher Omega-3 levels. It’s good to remember we’re talking about Omega-3 levels, DHA plasma levels that were measured at baseline or drawn at baseline, which is roughly 10 years before the pandemic came. As in all observational biomarker based research, we assume that the Omega-3 level at baseline was maintained over those 10 years and that was the Omega-3 level at the time of the pandemic.
We don’t know that for sure, but that’s just one of the assumptions and one of the weaknesses of this design of studies, but that’s something you live with. I guess you would say if there were major changes in the Omega-3 levels over time, then that would simply add a lot of noise to this analysis and we would be less likely to find any effect. The fact that we find a strong effect suggests that even if we knew the plasma Omega-3 levels the day before, December 31st of 2020 or 2019, whenever, we get a stronger relationship. We even see it more powerful.
Dr. Kristina Harris Jackson: Right. So these are showing up despite the change.
Dr. Bill Harris: So again, back to this testing positive, why in the world would higher Omega-3 keep you from testing positive? Well, it would have to be something to do either with higher Omega-3 levels are associated with people that are more vigilant in preventing exposure. That could be, but there is an interesting hypothesis that again, the Vivar-Sierra group proposed and there’s some good evidence for this that there’s something about having a high Omega-3 level in the cell membranes that actually causes the spike protein on the virus to be in a closed configuration, what’s called a locked configuration, and when it’s in a locked configuration, biochemically, it doesn’t interact very well with the ACE2 receptor, which is the receptor which pulls the virus into the cell and where all the replication of the virus happens. If you got low Omega-3, the spike proteins are more open and it can come in and replicate and you can become positive, but in this scenario, it would reduce your risk for positivity. It makes some sense. That could be it, but this is conjectural at this point.
Dr. Kristina Harris Jackson: Yeah, and the cell membrane examples of having one cell is all arachidonic and the other is all EPA and DHA is not accurate, but…
Dr. Bill Harris: Oh, did they actually… Oh, yeah. Oh, wow.
Dr. Kristina Harris Jackson: I just noticed that.
Dr. Bill Harris: No, I never noticed that. Never noticed that. That is…
Dr. Kristina Harris Jackson: But there is definitely evidence that arachidonic acid is replaced by EPA and DHA in the membrane, but they’re more like… We’re talking about 8% of the fatty acids and what, 10 to 15, 12?
Dr. Bill Harris: 12. Yeah.
Dr. Kristina Harris Jackson: For arachidonic, but-
Dr. Bill Harris: That’s interesting.
Dr. Kristina Harris Jackson: There, when they also are comparing the… If the virus does get into the cell, then the inflammation piece comes into play and [inflammation is out of control or not can be affected by what access it has to which fatty acids in the cell membrane. That is an arachidonic, EPA, DHA story a bit.
Dr. Bill Harris: Yeah. The higher Omega-3s are… Even if the virus got in the cell, you wouldn’t have the hyper inflammatory response.
Dr. Kristina Harris Jackson: Yeah. More likely not to, I think. That’s the idea.
Dr. Bill Harris: Yeah, so just to conclude, Omega-3 status at baseline was associated with increased… Low status associated with increased risk for hospitalization and testing positive, and certainly, there was some evidence for increased risk for death with the low Omega-3 levels. The cut points of under 4%, over 8% were supported by this study as originally proposed. The results suggest that Omega-3 from oily fish or supplements should be encouraged to raise Omega-3 levels before you get sick. This is not a treatment study, this is a status study and people at higher levels before the virus did better, so that’s the end of that. We can maybe stop sharing my screen.
Dr. Bill Harris: There it is right there. Okay. Great. Thoughts?
Dr. Kristina Harris Jackson: It’s just another pretty striking example of how Omega-3s are interacting in with the immune system. Prior to this, we would only really think of Omega-3s in relation to the inflammatory process and that you’d get sick but it wouldn’t be quite as bad, or if you have way too much Omega-3, you might not mount enough of an inflammatory response because you’ve knocked it down so much, but you have to be pretty out of the ballpark on that, but actually preventing a potential viral infection, is this pretty specific to COVID or are they thinking this… Most colds are the same kind of coronavirus, right?
Dr. Bill Harris: Yeah.
Dr. Kristina Harris Jackson: Has that been looked at?
Dr. Bill Harris: No. It’s not a huge leap to think this would apply to any respiratory virus, your susceptibility to adverse outcomes with it.
Dr. Kristina Harris Jackson: There’s pretty good vitamin D data on upper respiratory tract infections specifically and lung health, but for the Omega-3s, I don’t know if that has been studied as much.
Dr. Bill Harris: Not has been studied as much, not as vitamin D, but there’s certainly some evidence for it. We talked about it in the paper.
Dr. Kristina Harris Jackson: Yeah. The good news is if you did eat more fish, fish is also one of the best sources of dietary vitamin D besides the sun. The sun is not dietary, so fish are a good package of nutrients.
Dr. Bill Harris: Yes, it is. Right.
Dr. Kristina Harris Jackson: When we think about population studies and what their levels are and how they’re affected, it seems more likely that the dietary and genetic pieces are bigger factors than potential supplementation. Do you know from the other UK Biobank studies how much supplementation was happening in the overall cohort or in your cohort?
Dr. Bill Harris: About 31%.
Dr. Kristina Harris Jackson: For the overall or yours?
Dr. Bill Harris: Yeah, well, just both.
Dr. Kristina Harris Jackson: Which is pretty-
Dr. Bill Harris: Yeah, which struck me as quite a high number, but then I talked to people in Great Britain and they’re not surprised by it. A lot of the stories started in Great Britain. They picked up the Danish stuff and they ran with it. Early researchers in the eighties and nineties were English.
Dr. Kristina Harris Jackson: Okay. So it’s more accepted.
Dr. Bill Harris: Calder’s over there beating the drum all the time.
Dr. Kristina Harris Jackson: Yeah, yeah. That also helps, to some extent, explain having that bigger range. Sometimes our issue is when we’re looking at population studies, the lowest quintiles, 4% and the highest is 6 or something, because naturally, people aren’t having…
Dr. Bill Harris: That’s a real advantage of the UK Biobank because there is a bigger spread of Omega-3 levels.
Dr. Kristina Harris Jackson: In other studies, if you don’t have as big of a spread, you won’t see as big of a difference, and then the statistical challenge is a lot greater. What else are you guys looking at in the UK Biobank?
Dr. Bill Harris: Looking at total mortality, which is following the same trend as we’ve seen for everything else. Looking at stroke per se as part of a FORCE project and lower risk for stroke, higher Omega-3 in UK Biobank, and we then want to look at other outcomes. Dementia is a low hanging fruit, so we want to look at that as well. A lot of things. My blackboard’s full of diseases I want to study in the UK Biobank.
Dr. Kristina Harris Jackson: Yeah. The fatty acids we have access to are really just the long chain and Omega-3 and Omega-6 or did they also have total saturates and monos?
Dr. Bill Harris: Total saturates, total monos, total polys, total Omega-6, total Omega-3, linoleic and DHA.
Dr. Kristina Harris Jackson: I don’t know if we talked about this last time we did the UK Biobank overview, but there might be some real interesting things happening in the saturates and monounsaturates compared to the PUFAs, so there’s..
Dr. Bill Harris: Yes.
Dr. Kristina Harris Jackson: Probably, the DHA is a really good marker of the Omega-3 index, but with this big of a dataset, being able to look at larger patterns in fatty acids, there seems to be a lot signaling around that.
Dr. Bill Harris: Yeah. Actually, one of the biomarkers that’s most associated with risk for disease in the UK Biobank is the monounsaturated to polyunsaturated ratio, which we’ve not even begun to scratch the surface on that one. Omega-6 is favorably associated as usual with most outcomes. Linoleic acid is. That’s, again, a good thing for that whole discussion about Omega-6, Omega-3s.
Dr. Kristina Harris Jackson: There’s always more there.
Dr. Bill Harris: We’re not going to go there right now.
Dr. Kristina Harris Jackson: Yeah, no, but this is a pretty great paper. It’s great to have it in AJCN. That’s one of our lead journals in nutrition, so it does take a decent amount of work and diligence to get published in AJCN. Great to see this data set and these statistical methods are passing snuff and showing some real trends here.
Dr. Bill Harris: Yeah. Good.
Dr. Kristina Harris Jackson: All right, well, that’s all we have for today and we will see you next time.
Dr. Bill Harris: Bye-bye.
Dr. Kristina Harris Jackson: Bye.